Proceedings of the South Dakota Academy of …South Dakota Academy of Science Volume 83 2004...

Proceedingsof the

South Dakota Academy of Science

Volume 832004

Published by the South Dakota Academy of ScienceAcademy Founded November 22, 1915

EditorSteven R. Chipps

Associate EditorKenneth F. Higgins

Terri Symens, Wildlife & Fisheries, SDSUSecretarial Assistant

Tom Holmlund, Graphic Designer

TABLE OF CONTENTS

Consolidated Minutes of the Eighty-Ninth Annual Meeting of the South Dakota Academy of Science, Cedar Shore Resort, Oacoma, SD 2-3 April 2004 .................................................................................. 1

South Dakota Academy of Science 2003-2004 Executive Committee ........................... 42004 Regional Science Fairs .......................................................................................... 5South Dakota Academy of Science March 31, 2004 Proceedings

Disbursements/Receivables and Cash Balance in SDSU-SD Academy of Science Proceedings Account—Prepared By Di Drake ....................................... 8

Report of the Resolutions Committee ........................................................................... 8Executive Summary—Proceedings Editors .................................................................... 9SDAS Budget Summary For 2003 .............................................................................. 10The By-laws of the South Dakota Academy of Science ................................................ 10Presidential Address: In Search of a “Golden Age” For Science in South

Dakota. Andrew Detwiler .................................................................................... 19

Complete Senior Research PapersPresented at the 89th Annual Meeting of the


Use of Low Vacuum Electron Microscopy to Quickly Estimate Bacterial Populations on incubating Salmonid Eggs. Michael E. Barnes, Hans Stephenson and Mark Gabel ....................................................................... 27

Relative Abundance of Small Mammals in Native and Restored Tallgrass Prairie. Samuel J. Kezar and Jonathan A. Jenks .................................................... 33

Assessing Potential Transport of Tylosin in the Landscape. Jared K. Oswald, Todd P. Trooien, Sharon A. Clay, Zhuojing Liu and Robert Thaler ................................................................................................ 39

Fox Squirrel Weight and Age Structure in Mast and Non-Mast Forests. Steven L. Reighard, Aaron D. Bucholz, Justin A. Haahr and Jonathan A. Jenks ................................................................................................. 47

Ergodic theory and Its Application to the Analysis of Nucleic Acid Sequences. O. Michael Melko .............................................................................. 55

Accuracy of Home Soil Test Kits on South Dakota Soils. Rhoda Burrows .................. 65A Comparative Study of Seed Characteristics in the Chenopodiaceae and

Amaranthaceae. Elke Kuegle and Mark Gabel ..................................................... 73Naturally Occurring Acid Rock Drainage and Impacts to the Upper

Rapid Creek Watershed Near Rochford, SD. Scott L. Miller, Arden D. Davis, Scott J. Kenner, A.J. Silva .......................................................... 83

Hydrogeology of Lower Spearfish Canyon. Perry H. Rahn ....................................... 91Molecular Dynamics of Atomic Clusters: An Object Oriented Approach.

John M. Schneiderman and Brian G. Moore ..................................................... 101Percolation Partitioned Into Pore Size Classes. S.G. Wangemann and

R.A. Kohl ......................................................................................................... 115Road Culverts Across Streams With the Endangered Topeka Shiner,

Notropis topeka, in the James, Vermillion, and Big Sioux River Basins. Steven S. Wall and Charles R. Berry, Jr. ............................................................. 125

Habitat Use and Population Biology of the Northern Redbelly Snake at Oaklake Field Station, SD. Regina D. Cahoe and Nels H. Troelstrup, Jr. ........................................................................................ 137

Overview and Results of the Black Hills Hydrology Study. Daniel G. Driscoll and Janet M. Carter ............................................................. 149

Stratigraphy and Analytic Paleontology of the Lower Pierre Shale At Brown Ranch, Southwestern South Dakota. Marcus R. Ross ............................. 163

Environmental Effects on Xylem Cavitation in Seedless Cottonwood. Steven L. Matzner and Laura C. Gooch ............................................................. 183

Effects of Different Fire intensities on Understory Vegetation Diversity in the Jasper Burn Area of the Black Hills. Katie D. Derr, Steven L. Matzner and Lee A. Vierling .............................................................. 189

Field Evaluation of Pedotransfer Functions to Estimate Saturated Soil Hydraulic Conductivity. Darrell W. Deboer and R. G. Tekroney ....................... 197

Abstracts of Senior Research PapersPresented at the 89th Annual Meeting of the


Growth of Nano-Crystalline Hydrogenated Thin Films With Extremely Low Hydrogen Dilution. Yung M. Huh, R. Shinar and V. L. Dalal ................... 207

Development of A PCR Primer Set For the HMW Glutenin Genes Expressed in Chinese Spring Wheat. Leslie M. Baehr and Michael K. Wanous ........................................................................................... 209

Levels of Water-Extractable N-Nitroso Compound and N-Nitroso Compound Precursor in 6 Brands of Snuff. Michael Belling, Lin Zhou and Sidney S. Mirvish ........................................................................ 211

Synthesis of Deuterated Tris (2,2,6,-Tetramethyl-3,5-Heptanedionato) Europium (III) (Europium THD). Kathryn A. Henning, Annie Thompson, Allison Caster, Mary T. Berry, and Krisma D. Dewitt .............................................................................................. 213

Direct Observation and Tensile Strength of Some Polycarbonate Nanocomposites Containing Carbon Nanotubules. Josiah Reams, Tsvetanka Filipova, Guy Longbrake, and David A. Boyles ................................ 215

The Simultaneous Evolution of Defense and Competitiveness: Natural Selection. Riston Haugen and David Siemens ................................................... 217

Simultaneous Evolution of Competitiveness and Defense: Induced Switching in Arabis drummondii. Tessa Jones, Shannon Kulseth, Karl Mechtenberg, Charles Jorgenson, David H. Siemens, Michael Zehfus and Paul Brown ........................................................................ 219

Identifying Microsatellite Markers in the Genome of the Endemic Antiguan Ground Lizard, Ameiva griswoldi. Nathan T. Stephens, Brian E. Smith, Cynthia Anderson and Paul Colbert ......................................... 221

Generation of gE-EGFP and gI-EGFP Constructs to Study the Role of BHV-1 Glycoproteins gE and gI in BHV-1 Pathogenesis. Ihab Halaweish, Ehab Hassan, C. C. L. Chase and Lyle Braun .......................... 223

Non-Chloral-Based Synthetic Routes For the Production of BPC Monomers. Rachel Waltner, Josiah Reams, Guy Longbrake, Tsvetanka Filipova and David Boyles ................................................................. 225

Nanobio-Plastics and Composites From Linseed Oil and Saccharidic Source Materials. Michelie R. While, Annie M. Thompson, David A. Bovles, Jon J. Kellar and William M. Cross ........................................ 227

Synthesis of Novel Tetraaryl Copolycarbonates and Determination of Mark-Houwink Constants. Guy Longbrake, Tsvetanka Filipova and David A. Boyles ................................................................................................. 229

Some Effects of Land-use Management Practices on Seed Set of Cypripedium candidum. Carol Wake ................................................................... 231

Emergence Date Affects Growth and Fecundity of Amaranthus spp. E. Uscanga-Mortera, S.A. Clay, F. Forcella and J. Gunsolus ............................... 233

Development of A Method For Evaluating the Yield Goal Approach. K. Kim, D. E. Clay, C. G. Carlson, and S. A. Clay ........................................... 235

The Presence of Fluorapatite in Prismatic Cartilage From the Permian of Texas. Tylor B. Sampson and Gary D. Johnson ................................................. 237

Macrofungi Collected in the Black Hills of South Dakota and Bear Lodge Mountains of Wyoming From 1998-2003. A. C. Gabel , E. Ebbert, K. Lovett, S. Herrin, S. Mullen and D. Woolwine ............................................. 239

Analysis of Microsatellite Variation Within and Among Populations of the Topeka Shiner, Notropis topeka. Shane Sarver and Cynthia Anderson ................. 241

3-D Modeling and Computer-Generated Articulation of A Tyrannosaurus Rex Forelimb. Ellen Naito Starck ................................................. 243

The Effect of Calcium-Channel Blockers on the Closure of Traps of Venus Flytrap Plants (Dionaea muscipula). Brenda Simon, J.C. Neilson, Erin Talsma and James Sorenson .................................................. 245

QTL Associated With Maize Kernel Traits Among Illinois High Oil X B73 Backcross-Derived Lines. James Wassom, J. Wong and T. Rocheford ...................................................................................................... 247

Polymorphisms in the Agouti-Related Protein (AGRP) Gene in Pigs. Juanita Perera and Nels H. Granholm ................................................................ 249

Controls on Sedimentation in the Late Cambrian Deadwood Formation Near Lead, South Dakota. Melissa M. Campbell and Christopher J. Pellowski ..................................................................................... 251

Activity Period of Four Species of Carrion Beetle on the Pine Ridge Indian Reservation of South Western South Dakota. Louden Whirlwind Horse, Daniel G. Snethen, and William Wyatt Hoback ...................................................................................... 253

Population Density of Nicrophorus americanus the Federally Endangered American Beetle in the Nebraska Counties of Blaine, Brown, Loup, and Rock. Daniel G. Snethen and William Wyatt Hoback ................................ 255

The Effect of Three Types of Light: Halogen, Mecury Vapor and Ultra Violet on Nocturnal Species of Carrion Beetle including Nicrophorus americanus the Federally Endangered American Burying Beetle. Daniel G. Snethen and William Wyatt Hoback ..................................... 257

Dust Fraction in Catchment Sediments, Prairie Pothole Region, South Dakota. Richard Faflak and Jodie Ramsay ......................................................... 259

The Effects of pH Changes on Aerobic Denitrification Conducted By Pseudomonas aeruginosa in Hyperbaric and Normobaric Conditions. Ryan M. Klenner, Joshua J. Bathke, Sheena M. Benson, Jena R. Christianson, Ashley J. Hughes, Jessica A. Maschino, Amy M. Schmidt, Adam J. Smith, Angela L. Steams, Paul G. Van Heukelom and William J. Soeffing ................................................ 261

Are the Northern Great Plains An Inland “Coast”? – Fall Age Structure of Neotropical Migrant Birds in Southeastern South Dakota. Kurt L. Dean, Heather A. Carlisle and David L. Swanson ................................. 263

Fall Stopover Duration and Energetic Condition of Three Sandpiper Species in Western Minnesota and Eastern South Dakota. Nathan E. Thomas and David L. Swanson ........................................................ 265

Oak Lake Field Station As A Model For Ethnobotanical Research in the Prairie Pothole Region of South Dakota. Gretchen Mcclintock-Ames and R. Neil Reese .............................................................................................. 267

Method Development For the Determination of Neomycin in Animal Feed. Amanda Dupay, Nancy Thiex, Dave Ferris and Douglas E. Raynie ............................................................................................. 269

Primary Gustatory Centers in the Zebrafish Brain. Judy Locati, Matt Nehl and Charles Lamb ............................................................................ 271

Effect of A Spring-time Deworming Program on Strongyle Egg Output in Weaned Calves From Eastern South Dakota. A.F. Harmon, W.B. Epperson and M.B. Hildreth .................................................................... 273

Seasonal Fluctuations in Adult Mosquito Populations in Eastern South Dakota During the Summer of 2003. J.R. Bradley, K. Dahmash, R. Beyer and M.B. Hildreth ............................................................................... 275

Attempts to Locate Culex Tarsalis Larvae in Traditional Mosquito- Breeding Habitats From Brookings County, South Dakota. M.L. Hart, D.J. Thorpe and M.B. Hildreth ......................................... 277

Characterization and Comparison of Cells By Antigen Presentation, Microfilament Organization, and Phagocytic Properties. N. Harms, B. Tigabu, L.J. Braun, E.A.D. Hassan, G. Elmowalid and C.C.L. Chase ..................................................................................................... 279

Thyroid Hormone Distribution Associated With the Development of Hypertension and Congestive Heart Failure in Dilated Cardiomyopathic Hamsters. Laurie Keogh, A. Martin Gerdes and Patricia M. Tille ................................................................................................. 281

Polyarylate Co-Polycarbonates Containing the Monomer 2,2-bis[4-(4-Hydroxyphenyl)-Phenyl]Propane. Marci Medalen, Guy Longbrake, Tsvetanka Filipova and David Boyles ....................................... 283

Comparative Proteomic Analysis of Planktonic and Biofilm Pseudomonas Aeruginosa Cells: Identification of Differentially Expressed Proteins. Sébastien Vilain, Pascal Cosette, Guy-Alain Junter, Thierry Jouenne and Volker S. Brozël ........................................................................................... 285

Conversion of Lewis and Clark Lake and Lake Francis Case to Sustainable Systems. Howard Coker .................................................................. 287

Fecal Coliform Strain Identification to Facilitate Water Resource Management in South Dakota. Erick Jorgenson, Nels Troelstrup, Jr. and Bruce Bleaklev ............................................................................................. 289

The Hydroclimatology of the Central Territory of Conterminous U.S. and Stream Flow Regime in Upper Missouri River Basin. Boris A. Shmagin and Carol Johnston ............................................................... 291

Aerobic Enzyme Activities and Seasonal Acclimatization in Resident Passerine Birds. Eric T. Liknes and David L. Swanson ....................................... 293

SPECIAL SYMPOSIA:Following in the Footsteps of Lewis and Clark:

The Geology and Aleontology of the Missouri River

Fossil Fish From the Cretaceous of the Western Interior: Clarifying the Taxonomic Record. David C. Parris, William B. Gallagher and Barbara Smith Grandstaff ................................................................................... 299

Geology and Paleontology Recorded By the Corps of Discovery, Including the First Fossil Reptile From the American West. David C. Parris and Sally Y. Shelton .................................................................. 301

Bentonite Correlation of the Pierre Shale of South Dakota. Janet L. Bertog ............. 303Stratigraphy and Paleoecology of the Middle Pierre Shale Along the

Missouri River (Central South Dakota). Paul A. Hanczaryk and William B. Gallagher ......................................................................................... 305

Rare Earth Element Analysis (REE) of Fossil Vertebrates in the Pierre Shale: Paleoenvironmental Conditions. Doreena Patrick and David E. Grandstaff ........................................................................................... 307

Unusual Preservation of the Pelvic Region of Platecarpus, (Mosasauridae; Reptilia) From the Sharon Springs Member of the Pierre Shale. Bevin Rose O’Grady, John A. Pappas and Randolph J. Moses ........................... 309

Size Variation in Cranial Morphology of Late Cretaceous Toxochelys (Testudines; Cheloniidae) of South Dakota. Margaret Hart ............................... 311

The Largest Mosasaur (Squamata; Reptilia) From the Missouri River Area (Late Cretaceous; Pierre Shale) of South Dakota. Robert W. Meredith, James E. Martin and Paul N. Wegleitner ........................................................... 313

Meek and Hayden’s Nonmarine Paleontology of the Upper Missouri River Section. Joseph H. Hartman ..................................................................... 315

Distribution and Taphonomy of the Flying Reptile, Pteranodon, From the Campanian Lower Pierre Shale of Western South Dakota and Eastern Wyoming. Jennifer Roberts ................................................................... 317

Paleoecological Implications of the Fox Hills Formation (Maastrichtian) Reptilian Fauna From South-Central North Dakota. John W. Hoganson and Mark J. Erickson .......................................................... 319

Comparison of Gastroliths Within Plesiosaurs From the Late Cretaceous Marine Deposits of Vega Island, Antarctic Peninsula, and the Missouri River Area, South Dakota. Wayne A. Thompson, James E. Martin and Marcelo Reguero .............................................................. 321

Fossil Mammals of the Sentinel Butte Formation (Late Paleocene) of North Dakota. Allen J. Kihm, David W. Krause and Joseph H. Hartman ........................................................................................... 323

Geological Structures and Timing Constraints Along the Missouri River, Central South Dakota. J. Foster Sawyer and James E. Martin .................. 325

Revised Stratigraphy of the Lower Pierre Shale (Upper Cretaceous) of Central South Dakota. James E. Martin, Janet L. Bertog and David C. Parris .................................................................................................. 327

Nannofossils and Environment of the Upper Unit of the Crow Creek Member, Pierre Shale (Upper Cretaceous), Crow Creek Sioux Indian Reservation, Central South Dakota. Andrea Concheyro and James E. Martin ................................................................................................. 329

Molluscs in the Stomach Contents of Globidens, A Shell-Crushing Mosasaur, From the Late Cretaceous Pierre Shale, Big Bend Area of the Missouri River, Central South Dakota. James E. Martin and James E. Fox ...................................................................................................... 331

Rare Earth Element (REE) Analyses of Fossil Vertebrates in the Pierre Shale: Fossil Provenience. Doreena Patrick, James E. Martin, David C. Parris and David E. Grandstaff ........................................................... 333

The First Named Marine Reptile From the American West, A Mosasaur: Mosasaurus missouriensis (Harlan) 1834; Its History, Source, and Osteology. James E. Martin ................................................................................ 335

A New Species of the Diving Bird, Baptornis, From the Lower Pierre Shale (Upper Cretaceous) of Southwestern South Dakota. Amanda H. Person and James E. Martin ........................................................... 337

The Origin of James Ridge, Yankton County, South Dakota. Gary D. Johnson and Kelli Mccormick .............................................................. 339

PLENARY SESSION:Recovering Lewis and Clark’s River: Today’s Perspectives

on the History, Management and Ecology of the Missouri River

Fluvial Processes and Recreational Opportunities of the Lower Missouri River. Perry H. Rahn ......................................................................................... 343

CONSOLIDATED MINUTES OF THEEIGHTY-NINTH ANNUAL MEETING OF THE

SOUTH DAKOTA ACADEMY OF SCIENCE, CEDAR SHORE RESORT, OACOMA, SD 2-3 APRIL 2004

The Executive Council met at 7:00 a.m. Saturday 3 April 2004 for a final check of plans for the day. President Miles Koppang opened the executive committee meeting and noted that a quorum was present. A copy of the budget was distributed by Secretary Donna Hazelwood for Treasurer Kristel Bakker. Consensus of the Council was for Kristel to look into options for payment by credit card from the web site. The following individuals were selected as Fellows of the South Dakota Academy of Science. Perry H. Rahn, South Dakota School of Mines and Tech-nology and Robert Stoner, University of South Dakota. Nominations for Fellows for 2005 will involve a letter of nomination and brief biography, and will be due 1 October 2004. Discussion on the location of future meetings involved the following op-tions: 1) two campuses partner to host meetings; and 2) hosting at a central location such as Cedar Shore. Additional discussion included the addition of a symposium based on a topic of interest to the host institution, committee, and/or South Dakota. Guidelines for the symposia would include Friday afternoon symposium and poster session and contributed paper session on Saturday. Bob Tatina reported that the membership committee created and sent out an electronic poster for membership and information about the 2004 meeting. The Council decided to follow up on membership again after the meeting. Consen-sus on membership was that membership will be the responsibility of all mem-bers and to include one-on-one invitations to new faculty to join. Membership in the Academy will be a requirement for publishing a paper in the Proceedings. In addition, membership will also be a requirement for attendance at meetings and presentation of papers or posters. Donna Hazelwood and Kristel Bakker sent out eight award checks for $25.00 each to each of the five South Dakota Regional Science Fairs. The regional science fairs and representatives are: 1) Aberdeen, Jodie Ramsay; 2) Brookings, Neil Reese; 3) Mitchell, Bob Tatina; 4) Rapid City, Andy Detwiler; and 5) Timberlake, Neil Reese. Each of the SDAS representatives will provide for publication in the Proceedings student name, school, teacher name, year in school, title of project. The abstract will also be provided, but will remain in the archives of the Academy and will not be published. Ken Higgins reported that due to a change in software at the print shop, the 2003 Proceedings will be available and mailed out May 2004. The 2005 annual meeting will be held at the University of Sioux Falls, Sioux Falls, SD, and will include a symposium based on a topic of interest to USF. The Executive Committee would like to follow a format similar to that for the 2004 meeting with a half-day symposium Friday 1 April and contributed paper

Proceedings of the South Dakota Academy of Science, Vol. 83 (2004) 1

session Saturday 2 April. The 2006 meeting is scheduled to be held at Dakota Wesleyan University. The Audit Committee will consist of Mark Gabel and Audrey Gabel. The Nominating committee, Dave Swanson from USD, Mike Hildreth from SDSU, and Dave Siemens from BHSU, will have the following positions to fill. Second Vice President, and two members-at-large. Gary Earl and Neil Reese will be going off as members-at-large. On behalf of the By-laws committee, Miles reported that the membership will need to vote on changes in the by-laws concerning the wording on selection of fellows, splitting the Secretary-Treasurer position into the positions of Secre-tary and Treasurer. Registration for the Annual Meeting began 7:00 p.m. Thursday 1 April, continued 7:00 a.m. Friday 2 April and again 7:00 a.m. Saturday 3 April. Ken Higgins brought Terri Symens and Di Drake to assist with registration. President Elect Andrew Detwiler gave an interesting presidential address entitled “In Search of a Golden Age for Science in South Dakota”. The business meeting was opened by President Miles Koppang. He noted that five former presidents were in attendance: Sharon Clay, Chuck Estee, Mark Gabel, Nels Granholm, and Jim Martin. The Treasurer report was provided by Donna Hazelwood for Kristel Bakker. Mark Gabel and Audrey Gabel served as Auditing Committee. Neil Reese moved and Ken Higgins seconded that the treasurer’s report be accepted as written. The motion passed by voice vote. The CD at Dakotah Bank will be allowed to roll-over for another term. Elections were held for officers for 2004-2005. Bob Stoner moved and Char-lie Lamb seconded nominations cease and members cast a unanimous ballot in favor for Second Vice-President Mike Wanous. The motion carried by voice vote. Bob Stoner moved and Neil Reese seconded that nominations for 2004-2006 member-at-large cease and members cast a unanimous ballot for John Naugard and Dave Swanson. The motion carried by voice vote. Resolutions were proposed by the Resolutions Committee, Perry Rahn, Howard Coker, and Gary Johnson. Donna moved and Charles seconded the following resolutions: 1) thank Cedar Shore Resort and the local planning com-mittee as listed in the program for the 2004 Academy meeting; 2) thank you for Andy Detwiler for his presidential address on “In Search of a Golden Age for Science in South Dakota” 3) thank you to Jim Martin for hosting the paleon-tology field trip, and Jeff Palmer for hosting the ornithology field trip; 4) a big thank to Gary Moulton for the keynote address, 4) commending President Miles Koppang for his direction and leadership, 6) thank you also to Jim Martin for hosting a symposium, 7) thank you to moderators Steve Chipps and Bob Tatina for moderating the Friday Symposium, 8) thanks to the secretaries Di Drake and Terri Symens for assisting with registration at the meeting; and 9) a special thanks goes to Editors Ken Higgins and Steve Chipps for their oversight of timely publication of the Proceedings. Charlie Lamb moved and Dave Swanson seconded acceptance of the resolutions. The motion carried by voice vote. The by-laws committee, Miles Koppang, Bob Stoner, and Chuck Estee, introduced changes that need to be updated that have been voted upon. The

2 Proceedings of the South Dakota Academy of Science, Vol. 83 (2004)

language of section 7 article 1 addressing for designation and approval of Fel-lows was modified. In addition the committee proposed that the by-laws include wording on the split of the Secretary-Treasurer position to include that the Secre-tary and Treasurer each have an equal vote. Bob Stoner moved and Jim Sorenson seconded approval. The motion passed by voice vote. Ken Higgins, notified the membership that the 2003 Proceedings will be the last year in which he will be Proceedings Editor. The current Associate Proceed-ings Editor, Steve Chipps will assume duties as Editor for the 2004 Proceedings, and Ken will become Associate Editor, Ken requested electronic copies of the Presidential Address, By-laws changes and Resolutions Committee. Neil Reese reported that he and Tom Holmlund have downloaded in PDF format all issues from 1996-2002 and posted them to the website. He will wait one year to down-load the 2003 Proceedings. Ken moved and Jim Lefferts seconded acceptance of the Proceedings report. The motion passed by voice vote. Ken requested that the Academy provide honoraria for assistance in the fol-lowing amounts to Terri Symens $300.00, Di Drake $100.00, and Nancy Pre-shum $100.00. Neil moved and Donna seconded a motion to give the amounts requested. The motion carried by a voice vote. Discussion followed about the format and location of future meetings. Howard Coker moved and Bob Stoner seconded that the Executive Committee be given the authority to consider the alternative sites for the annual meeting in addition to meetings hosted by institutions. The motion carried by voice vote. Discussion about having a symposium in concurrence with the general meeting resulted in agreement that the choice of symposium not be restricted to the host institution but could be expanded to include ideas from membership. Consensus was that tying to the community was a good idea. Outgoing President Miles Koppang turned the meeting over to President Andy Detwiler, and First Past-President Charlie Lamb presented Miles with a certificate of appreciation. The 2005 meeting of the Academy will be at The University of Sioux Falls, and will follow the format of Friday symposium and papers and posters on Sat-urday. Ken noted that normally the host institution for the annual meeting prints the program. For the 2004 meeting at Chamberlain, Ken will be in charge of printing the program.

Committee positions for 2004-2005 includeMembership Bob Tatina to be filledBy-laws Miles Koppang Bob Stoner Chuck EsteeFellows to be filledResolutions to be filledNominations to be filledPublicity for 2005 meeting to be filled


Several items for consideration at the fall meeting of the Executive Committee were discussed. 1) the 2005 meeting at USF; 2) nomination of individuals for Fellow; 3) recruitment of new members; 4) a new operations manual to replace the one that apparently has been lost; 5) assisting the five regional science fairs, providing prizes at the middle school level; 6) the 2006 meeting scheduled for Dakota Wesleyan University; and 7) the format and locations of future meet-ings.

Respectfully submitted,Donna Hazelwood, DSU

SDAS Secretary

SOUTH DAKOTA ACADEMY OF SCIENCE2003-2004 EXECUTIVE COMMITTEE

President Miles Koppang, USD Chemistry, 677-5693 [email protected]

President-Elect Andrew Detwiler, SDSM&T IAS, 394-1995 [email protected]; FAX394-6061

First Vice-President Robert Tatina. DWU Biology, 995-2712 [email protected]

Second Vice-President James Sorenson, MMC Biology, [email protected]

Secretary Donna Hazelwood, DSU Natural Sciences 256-5187 [email protected]; FAX 256-5643

Treasurer Kristel Bakker, DSU Natural Sciences 256-5182 [email protected]; FAX 256-5643

Proceedings Editor Kenneth F. Higgins, SDSU Wildlife, 688-4779 [email protected]; FAX 688-4515

Assistant Proceedings Editor Steve Chipps, SDSU Wildlife, 688-5467 [email protected]; FAX 688-4515

First Past President Charles Lamb, BHSU Biology, 642-6026 [email protected]


Second Past President Members-at-Large2002-2004 Gary W. Earl, Augustana, Chemistry [email protected]; FAX 357-9772

2002-2004 R. Neil Reese, SDSU Biology/Microbiology, 688-4568 [email protected]

2003-2005 Krisma DeWitt, Mount Marty College Chemistry 668-1530 [email protected]

2003-2005 Bill Soeffing, University of Sioux Falls, Biology 331-6759 [email protected] of the South Dakota

2004 REGIONAL SCIENCE FAIRS

The SDAS supports student interest and participation in science. Toward that end, representatives from the Academy presented eight $25.00 awards at each of the five SD regional science fairs for a total of 40 awards. The awards were distributed as follows:

2004 Regional Science Fair Location: DWU, Mitchell, SDSDAS representative: Bob Tatina

Student School Title of Project

Clarence Vanderlei Avon Middle School Can You Fix Your Dirt?Emily Bartunek Avon Middle School Natural Or CommercialSarah Luebke Corsica Middle School Things Are Heating UpKaycee Kopfmann & Ashley Losing

Wessington Springs Middle School

Cordless Crystal Radio

Samantha Hohn & Amber Whing

Ethan Middle School Which Type Of Water Is Best For Plants?

Tanner Duba & Cory Wegehaupt

Ethan Middle School Golf Balls & Distance

Amanda Hintz & Alicia Schrank

Stickney Middle School Corn Germination

Elizabeth Bosworth White Lake Middle School Flour Power



2004 Regional Science Fair Location: Northern State University Campus—Barnett Center—Aberdeen, SDSDAS representative: Jodie Ramsay

Student School Year Teacher Title of Project

Ben Volk &Chad Mitzel

Herreid School 7th Gary Weismantel Rise, Rise, Rise

Reid Turner AberdeenSimmons School

7th Gus Erickson Which Brown County Soil Grows The Best?

Cole Hogg Faulkton School 7th Gloria Bode Do Different Traits Af-fect Multiple Births?

Hannah Wickard Ipswich School 8th Arlene Schopp Does It Dehydrate?Britany Schanzenbach & Erica Bender

Selby Area School

7th Gary Fahrni Storage “Do’s & Don’ts” For Common Liquids

Bob Brown, Teacher AberdeenSimmons School

2004 Regional Science Fair Location: Timber Lake, SDSDAS representative: Neil Reese

Student School Year Teacher Title of Proejct

Cody Aberle, Brett Mayer, & Scott Biegler

Timber Lake School

9th LuAnn Lindskov How Does The Diameter, Type And Length Of Wire Affect Conductivity?

Megan Johnson Timber Lake School

9th LuAnn Lindskov Do Antacids Affect The Way Antibiotics Work In The Hu-man Stomach?

Jillaine Pfeifle & Emily Reinbold

Timber Lake School

9th LuAnn Lindskov Pain Tolerance

Jessica Stradinger Isabel School 9th Eric Mertens How Does Ethylene Affect Crickets?

Travis Davis & Tal Wammen

Harding County School

8th Matthew DeBow Toybox Construction


2004 Regional Science Fair Location: Brookings, SDSDAS representative: Neil Reese

Student School Year Teacher Title of Project

Nathan Iverson & Darin Anderson

Deubrook Area School

8th Mike Knust Which Pop Is The Most Harmful?

Ty Fuller Henry School 10th Janelle Stahl Do Taller People Have A Greater Lung Capacity Than Shorter People?

Julianne Goltz Hamlin Middle School

8th Michelle Bartels Smart Snacks

Cody Kreins West Central Middle School

6th David Heck Power Of Soda Pop

Brian Sullivan De Smet School

6th Mark Birkel Paper Towel Strength

Aldo Magana Henry School 12th Janelle Stahl Propel Electro Magnetic Cannon

Mark Birkel Teacher

De Smet School

2004 Regional Science Fair Location: Rapid City, SDSDAS representative: Andrew Detwiler

Student School Year Title of Project

Katherine O’Donahue Spearfish Middle School 6th Brain Games: Optical IllusionsJames Sorensen Wall Middle School 6th Feel the PowerTara Lorraine Correll Edgemont Middle

School7th How Does Your Brain React?

Chanz Timothy John Paddack

Black Hills Christian Academy (Rapid City)

8th Can You Remember?

Preston Engesser & Justin Frasier

Spearfish Middle School 7th Brain Power 2

Zachary Ryan Smith West Middle School (Rapid City)

8th Tidal Turbine

Kirstie Bradley Spearfish Middle School 6th What Type Of Material Is Best When Made Into A Hat?

Skye O’ Brien Belle Fourche Middle School

8th Homemade or Commerical Carpet Cleaners?

SOUTH DAKOTA ACADEMY OF SCIENCEMARCH 31, 2004—PREPARED BY DI DRAKE

Proceedings Disbursements/Receivables

2003 2002 2001 1996-2000 Total

Lay Out Formatting 2,490.35 2,251.25 2,836.5 4,185.00 11,763.10Publication 4,000.00 * 3,978.76 4,997.00 14,127.47 27,103.23Reprints 1028.42 * 1037.39 1178.98 3859.46 7104.25Conversion to PDF File 1,600.00 0 0 0 1,600.00Miscellaneous Printing 16.39 108.47 33.26 174.71 332.83Supplies,Phone,Postage 101.60 328.61 450.29 429.69 1,310.19TOTAL EXPENSES 9,236.76 7,704.48 9,496.03 22,776.33 49,213.6

TOTAL INVOICED 7,625.00 7,055.00 11,060.00 24,888.00 50,628.00

Profit / Loss to date (1,611.76) (649.48) 1,563.97 2,111.67 1,414.4

* Estimate

Cash Balance in SDSU-SD Academy of Science Proceedings Account

2003 2002 2001 1996-2000 Total

Total Paid ExpensesPaid by SDAS-Treasurer

(9,236.76) (7,704.48) (9,496.03) (22.776.33)4,747.13

(49,213.60)4,747.13

Payments ReceivedPaid to SDAS-Treasurer

6,750.00 6,780.00 10,770.00 23,624.00(240.00)

47,924.00(240.00)

Balance of SDSU Account (2,486.76) (924..48) 1,273.97 5,354.80 3,217.53

Percent of Paid Invoices 0.89 0.96 0.97 0.95 0.95Unpaid Invoices 875.00 275.00 290.00 1,264.00 2,704.00

Estimated cash on hand after 2003 expenses: $3,217.53

REPORT OF THE RESOLUTIONS COMMITTEE

The membership of the South Dakota Academy of Science thanks the Cedar Shores Resort for making the facilities available for the 2004 Academy annual meeting. The Academy extends its thanks to the local planning committee for making arrangements for the meeting. The members are: Bob Tatina, Steve Chipps, Ken Higgins, Terri Symens, Donna Hazelwood, Kritel Bakker, and Jim Martin. We appreciate the efforts to include two symposia at this year’s meeting. Jim Martin and David Parris were the moderators for “Following in the footsteps of Lewis and Clark: The geology and paleontology of the Missouri River”. Steve Chipps and Robert Tatina were moderators for “Recovering Lewis and Clark’s


River: today’s perspectives on the history, management, and ecology of the Mis-souri River”. We thank Gary Moulton for his lecture “The living legacies of Lewis and Clark”. President Miles Koppang is to be commended for his direction and leader-ship provided the Academy during this past year. Thanks for secretarial help from Terri Symens, Di Drake, and Nancy Pre-shum. Thanks to Donna Hazelwood for being our Treasurer this past year. Thanks to President-elect Andy Detwiler for his presidential address “Search-ing for the Golden Age of Science in South Dakota.” The Academy is pleased with the contributions of the paleontologists at this year’s meeting. This includes the field trip “The geology and paleontology of the Missouri River.” We also thank Kristel Bakker and Jeff Palmer for leading a field trip “Birds of the Missouri River.” The Academy expresses their thanks to Ken Higgins and Steve Chipps, edi-tors of the Proceedings, for their dedication in producing a quality publication.

Respectfully submitted,Perry Rahn, Chairman

Gary JohnsonHoward Coker

EXECUTIVE SUMMARY—PROCEEDINGS EDITORS

Volume 82 for 2003 totaled 379 pages and the production cost for 230 copies was $5,730.35. Our overall Proceedings account balance to date is $3,217.53. Copies have been mailed to all current members, all life members, all the State libraries and to abstract indexing providers.

Respectfully submitted by:Steven R. Chipps and Kenneth F. Higgins, Co-EditorsFor the South Dakota Academy of Science Proceedings

August 31, 2003


SDAS BUDGET SUMMARY FOR 2003

Beginning Balance: April 2003 $9416.60

SDAS 2003 Meeting Registration copying DSU -27.50 Miscellaneous (copying, etc) SDSMT -121.30 Faculty Lounge SDSMT -50.00 Banquet Radisson -1,487.35 Terri Symens -200.00 Di Drake -100.00 Linda Embrock -100.00

Memberships, Registration, Banquet +3,412.00

2003 Executive Board Meeting in Chamberlain -116.49

Science Fairs -1,025.00

Miscellaneous costs (stamps) -37.00

Dallas Jewelry, Vermillion (Engraving) -360.00

Cash for 2004 SDAS -300.00

Ending Balance April 2004 $9276.99

Respectfully submitted,Kristel K. Bakker, Ph.D.

Treasurer, SDAS

THE BY-LAWSof the

SOUTH DAKOTA ACADEMY OF SCIENCE

ARTICLE IMembers

SECTION 1. The membership of the Academy shall consist of members, associate members, life members, honorary members, and corporate members.

SECTION 2. Members shall be those persons interested in scientific work and research with the exception of those persons specified in Article I, Sections


3, 5, and 6. The right to vote and hold office shall be reserved to members and life members.

SECTION 3. Associate members shall be undergraduate/graduate students interested in scientific work: students may become full members upon recom-mendation by a majority of the membership committee and upon election by three-fourths of membership voting at an annual meeting.

SECTION 4. Life members. (a) Any member or applicant for member-ship who at one time shall contribute to the funds of the Academy ten times the annual dues may be elected a life member of the Academy and shall be exempt from payment of all dues and fees thereafter. (b) An individual who has been a member of the South Dakota Academy of Science for twenty-five years and who has reached the age of sixty-five years, shall be elected to life membership. He shall be exempt from payment of all dues and fees thereafter.

SECTION 5. Honorary members. Any person who has rendered con-spicuous service in the advancement of science may be elected to honorary membership of the Academy. Honorary members shall be exempt from all dues and fees.

SECTION 6. For election to membership, the candidate’s name must be proposed by two members, be approved by a majority of the Committee on Membership, and receive the assent of three-fourths of the members voting. Honorary and life members will be elected by a vote of members at an annual meeting upon recommendation of the Membership Committee. A three-fourths majority of those voting will be required for election.

SECTION 7. Academy Fellows. A member of the South Dakota Academy of Science may be designated as an Academy Fellow; an honor afforded to those members whose years of service to the Academy and the promotion of science in the state and nation is meritorious. Nominations for fellows will be solicited annually from the membership. Nominees must be active members of the Acad-emy for a total of at least 10 years. The nominee will have made outstanding contributions in the area of specialization whether in research, teaching, exten-sion, service, or private service activities. The nomination will include a letter of recommendation about the nominee including degrees received, professional positions held, membership in honorary academic societies, honors and awards received since baccalaureate degree, service to SDAS in appointed and elected positions, and professional achievements in science (teaching, extension, inves-tigative, service, leadership). Review of nominations and designation of the title Honorary Fellow will be initially conducted by the Executive Committee. When the membership of active Honorary Fellows reaches a critical mass (more than 10 members), the review and designation of Honorary Fellow will be conducted by the active group of Honorary Fellows. If the membership of active Fellows falls below the critical mass, then the Executive Committee will assist with the selection of Honorary Fellows.


SECTION 8. The Junior Academy of Science shall consist of high school students interested in science. The organization of the Junior Academy, mem-bership requirements, dues, and time of meetings, shall be regulated by the Committee of the Academy for Junior Academy Activities. Ordinarily, the Junior Academy will hold its annual meeting at the time and place designated for the annual meeting of the Senior Academy. The annual report of the Junior Academy shall be published in the Proceedings of the South Dakota Academy of Science for the current year.

SECTION 9. Corporate members shall consist of such organizations and business concerns which shall wish to affiliate with the South Dakota Academy of Science. Such members shall pay annual dues as determined and published by the Academy.

ARTICLE IIOfficers

SECTION 1. The officers of the Academy shall consist of a President; a President-elect who will automatically succeed to the President at the time of the next annual election of officers or as specified in Article II, Section 5; a First Vice-President; a Second Vice-President; a Secretary, a Treasurer; and an Editor of the Proceedings of the South Dakota Academy of Science. These officers, with the exception of the President, the Secretary and the Treasurer, shall be elected annually by a majority vote of those members voting at the annual meeting or as specified in Article II, Section 5. The Secretary and Treasurer shall serve for a term of three years. At the annual meeting prior to the third year of the term of the Secretary and Treasurer, a Secretary-elect and Treasurer-elect shall be elected by a majority vote of those members voting at the annual meeting or as specified by Article II, Section 5. The Secretary-elect and the Treasurer-elect will auto-matically succeed to the Secretary and Treasurer, respectively, at the time of the next annual election of officers or as specified by Article II, Section 5. All officers shall perform the duties usually pertaining to their offices.

SECTION 2. The President-elect at each annual meeting shall appoint a nominating committee of five active members, no two of whom shall be from the same county. It shall be their duty to recommend candidates for the several offices and for elected committees. This slate of officers shall be presented to the members and be voted on at the next annual meeting. This list shall pres-ent the names of two nominees for the office of Second Vice-President, one of whom shall be elected to the office. Nominations for candidates other than those presented by the nominating committee may be made by any member in accordance with Robert’s Rules of Order.

SECTION 3. It shall be one of the duties of the President-elect to prepare an address which shall be delivered before the Academy at the opening of the annual meeting.


SECTION 4. The President shall present to the members at the annual meeting such actions of the Council as are contained in Article III, Section 2 and those subject to ratification by the membership.

SECTION 5. Upon death or resignation of the President, the President-elect shall immediately assume the office of the President; upon the death or resignation of the President-elect or First Vice-President, the vacant office shall be filled by the officer immediately below the vacancy. The remaining vacancy or any other vacancy in any position on the Executive Council or any other elec-tive position in the Academy, with the exception of the First and Second Past Presidents, shall be filled by the Executive Council voting upon two persons for each position nominated by the President.

ARTICLE IIICouncil

SECTION 1. The Executive Council of the South Dakota Academy of Science shall consist of the President, First Past-President, the President-elect, First and Second Vice-Presidents, and Secretary-Treasurer of the Academy. In addition, the Second Past-President, the Editor of the Proceedings of the South Dakota Academy of Science, and four members at large, to be elected at the time of the election of other officers and in the manner of their election, are to be included. These four members at large are to be elected from the membership to serve for two-year terms except that two shall be elected each year. At the time of the initial election of these members, two will be elected for a one-year term.

SECTION 2. To the Council shall be entrusted the management of the affairs of the Academy during the intervals between regular meetings. No ex-traordinary act of the Council shall remain in force beyond the next following annual meeting without specific ratification by the members. The President may call special meetings of the Council and shall call such a meeting upon written request by three members. The members of the Executive Council shall consti-tute the membership of the Board of Directors of the South Dakota Academy of Science.

ARTICLE IVCommittees

SECTION 1. The standing committees of the Academy shall be a Com-mittee on Publications, a Committee on Membership, a Committee on Junior Academy Activities, a Resolutions Committee, and a Nominating Committee.

SECTION 2. The Committee on Publications shall consist of the President, the Secretary, and the Editor. In addition, this committee may call upon one or more members to serve in any specific case. This committee shall be empowered to specify the conditions for the publication of papers in the Proceedings of the South Dakota Academy of Science. The Editor shall serve as chairperson of this committee.}


SECTION 3. The Committee on Membership shall consist of eight mem-bers who reside in different regions of the state and who represent different levels of education, areas of research, and fields of employment. They shall be elected annually by ballot. The Nominating Committee shall present a list of candidates at the time of the annual meeting. The chairperson shall be appointed from the committee by the President. The Membership Committee shall conduct an an-nual membership campaign among all educational, service and research institu-tions of the state. This campaign will include the distribution of informational brochures and membership applications.}

SECTION 4. The Committee on Junior Academy Activities shall consist of five members to be elected annually by ballot. The nominating committee shall present a list of candidates at the time of the annual meeting. The chairperson shall be appointed from the committee by the President. The Junior Academy Committee shall conduct all business and meetings pertaining to the operation of the South Dakota Junior Academy of Science.

SECTION 5. The President shall appoint the Academy Representative to the Council of the American Association for the Advancement of Science and to the Academy Conference.

SECTION 6. The President shall appoint annually an Auditing Committee of three members which shall audit the books of the Secretary-Treasurer prior to the annual meeting and report the results of this audit at the business meeting.

SECTION 7. The Resolutions Committee shall consist of three members to be appointed annually by the Academy President. The chairperson shall be appointed from the committee by the President. The Resolutions Committee is charged with the responsibility of presenting any special policies, recommenda-tions, or recognitions to the membership of the Academy at the annual meeting for a vote of adoption.

ARTICLE VMeetings

SECTION 1. The regular meetings of the Academy shall be held at such time and place as the Council Members may designate. Special meetings may be called by the Council and shall be called upon the written request of fifteen members.

SECTION 2. The annual meeting shall be in the spring of each year. At that time the Annual Banquet and Academy Lecture will be held. This annual meeting will ordinarily be held at one of the colleges or universities of the state of South Dakota. The host institution shall be determined by the Executive Council. The host institution will select the speaker who will deliver the Annual Academy Lecture.


ARTICLE VIPublications

SECTION 1. The regular publications of the Academy shall include the Proceedings of the South Dakota Academy of Science, which will contain such papers, presented by members and associate members of the Academy, as are deemed suitable by the Committee on Publications.

SECTION 2. The publication of papers by persons not members of the South Dakota Academy of Science and of papers by members of the Junior Academy shall be determined by the Editorial Committee as stated in Article IV, Section 2.

SECTION 3. All members in good standing shall receive gratis the current issues of the publications of the Academy.

ARTICLE VIIOrder of Business

SECTION 1. The following shall be the regular order of business:1. Call to order and report by the President2. Reports of the Secretary-Treasurer3. Reports of Officers4. Reports of Standing Committees5. Election of New Members6. Reports of Special Committees7. Unfinished Business8. New Business9. Election of Officers10. Appointment of Special Committees11. Adjournment The program may be interpolated in the order of business according to con-venience.

ARTICLE VIIIAnnual Meetings and Fiscal Year

SECTION 1. No meeting of the Academy shall be held without notice sent by the Secretary-Treasurer to all members at least ten days before the date of the proposed meeting.

SECTION 2. The members present at the annual meeting shall constitute a quorum of the Academy. The majority of the Council Members shall constitute a quorum of the Council.

SECTION 3. The fiscal year of the Academy shall extend from January 1 through December 31.


ARTICLE IXAdministration of Funds for the Junior Academy

SECTION 1. Funds received annually from the American Association for the Advancement of Science for research activities shall be awarded by the Ju-nior Academy Activities Committee of the Academy to selected Junior Academy members in accordance with the intent of the policy of the AAAS. The two annual AAAS honorary memberships shall be awarded by this committee. The committee will inform the Secretary-Treasurer in writing of the names of recipi-ents of awards and their home addresses.

SECTION 2. The dues of the members of the Junior Academy and other funds acquired shall be deposited with the Secretary-Treasurer and expended by the Treasurer of the Junior Academy through such procedure as the Secretary-Treasurer may indicate.

ARTICLE XDues and Assessments

SECTION 1. Each member shall pay an initiation fee and annual dues as determined by the Academy.

SECTION 2. Undergraduate students shall become associate members through recommendation by a member and by payment of an annual fee as determined by the Academy.

SECTION 3. Members who shall allow their dues to remain unpaid for two consecutive years (computed at the time of the closing session of the Annual Meeting), after having been annually notified by the Secretary, shall have their names stricken from the roll. No member who is in arrears shall be entitled to vote.

ARTICLE XIDuties of the Secretary and the Treasurer

SECTION 1. The Secretary shall keep a record of all meetings of the Acad-emy and of the Executive Council, and shall prepare these records for publica-tion. The Secretary shall conduct all correspondence relating to the Academy. The Secretary shall have charge of the preparation, printing, and mailing of circulars, blanks, announcements of meetings, and other materials of a similar nature. At the annual meeting the Secretary shall present to the incoming Lo-cal Arrangements Chairperson a document including customary procedures for arranging the upcoming meeting. After each annual meeting, the Secretary shall send to each incoming member of the Executive Council a copy of the By-Laws.

SECTION 2. The Treasurer shall collect dues, mail notices of arrears in dues, receive and disburse all funds of the Academy, and have custody of all


funds. The books are subject to annual audit. The Secretary-Treasurer is autho-rized to expend such funds as are necessary to conduct the business of the office. All other disbursements are to be authorized by the Executive Council or by the members at the time of the annual meeting.

SECTION 3. The Treasurer shall be empowered to accept such grants, gifts, and bequests as may be offered to the Academy. Stipulations as to the use of such sums as may be made by the donors shall be subject to the approval of the Executive Council.

SECTION 4. The Treasurer shall have charge of the distribution, sale, and exchange of the published transactions of the Academy, under such restrictions as may be imposed by the Council. The Treasurer shall have charge of all books, collections, and materials belonging to the Academy.

ARTICLE XIIPresentation of Papers

SECTION 1. The Secretary shall notify all members in ample time before any meeting at which papers may be presented, and papers to be read must be filed with the Secretary by title and abstract at a time to be designated by the Sec-retary. Papers presented by associate members must, in addition, be co-authored, sponsored, or approved by a member. It shall be the duty of the Secretary to ar-range the program for the annual meeting and to provide copies of the program for the use of the members.

ARTICLE XIIIAdditions, Amendments, or Changes

in theArticles of Incorporation and the By-Laws

SECTION 1. The Articles of Incorporation may be amended by a three-fourths vote of members voting at an annual meeting, provided that notice of the desired change has been sent by the Secretary to all members at least twenty days before such meeting. Members may vote by mail and have their vote recorded in the minutes.

SECTION 2. By-Laws may be adopted, suspended, or amended by a three-fourths vote of the members present at any annual meeting.

SECTION 3. Any changes in the Constitution or the By-Laws of the Acad-emy, or additions or amendments thereto, shall be made in compliance with such laws of the State of South Dakota as govern or regulate educational and scientific corporations.

SECTION 4. The rules contained in Robert’s Rules of Order shall govern the Academy in all cases in which they are not inconsistent with the By-Laws of the South Dakota Academy of Science.



PRESIDENTIAL ADDRESS

In Search of a “Golden Age” for Science in South Dakota

Address to the South Dakota Academy of ScienceCedar Shores Resort, Oacoma, SD

April 3, 2004

Presented by Andrew DetwilerSouth Dakota School of Mines & Technology

Rapid City, SD 57701

As we develop (stumble?) through our careers as scientists, teachers, research-ers, academic administrators, and so forth, here in South Dakota, we often indulge our human tendency to complain about the general state of things. When my colleagues at the School of Mines and Technology sit together at a lunch table in the cafeteria or at coffee break in the faculty lounge we often find ourselves comparing the conditions under which we perform our teaching, research, and service, to those of colleagues at institutions elsewhere and even to conditions years earlier at our own institution. As might be expected, the grass usually appears greener outside South Dakota’s borders, and within our borders it appears greener for times sufficiently far removed from the present that only one or two of our faculty colleagues have even dim recollections of those times. A review of the annual presidential addresses printed in the Proceedings of the Academy demonstrates that Academy presidents often use their presidential address to assess the state of science in the state. Perhaps by the time presidents arrive in this position we are so well practiced at such assessment, as a result of the hundreds of hours of lunches and coffee breaks that have been interspersed between the classes and research sessions that filled our days and evenings over the years, that it is a natural presentation to make. For this reason if for no other, it seems appropriate to me in 2004 to continue this tradition. This year the fo-cus will not be so much on comparison to what is or has transpired beyond our borders, but on comparison to what has happened in South Dakota before the present. Although not formally trained in philosophy or psychology, it seems to me that no one here would challenge the idea that we scientists, like humans in almost every line of endeavor, often yearn for a “golden age”, one in which condi-tions are nearly perfect for us to practice our profession. What would constitute a “golden age” for science? Most of us would, I think, say that such conditions in such an age must include the flexibility to independently pursue new knowledge with at least modest but automatic backing from our home institution. Why must we always be preparing proposals and budgets and threading the intricate and uncertain maze of the federal bureaucracy seeking the funds we need to pursue our research ideas, even for very small and inexpensive projects? Why must we suffer multiple rejections for each success in such searches for funding? Most of us are self-confident enough to expect that reasonable resources should be available for the asking, given our competence in our fields and our own per-


ceptions of our value to society. (Unfortunately, polls of the public are mixed in terms of their confidence in us to make valuable contributions on their behalf.) Most of us might also support the notion that to a large degree the best science is accomplished when scientists pursue their own notions of what is important – “curiosity-driven” science. Why must we restrict ourselves to the topics and approaches mentioned in the broad agency announcements, requests for proposals, etc., very often of a distinctly applied nature? What do the research bureaucrats who write these solicitations know about what is important in our fields? Obviously, given the way the world works today, more than scientific cri-teria often go in to choices of topics for funding opportunities, but perhaps in a “golden age”, extra-scientific issues would have minimal influence? Beyond this, where is the flexibility to allow us to pursue some of the deeper questions that no funding agency will support, the “meta-science” of combining knowledge from different disciplines in order to update our general worldview? This is stuff with no direct application to any societal need. For instance, an ex-ample of such a question is how are we to understand our “common” worldview? Is it based on objective observations of the physical universe, and is there only one possible worldview? Or is our worldview a social construct, one of multiple possible worldviews constructed by different social groupings, each consistent with the observed evidence? How does one find the time and support to consider these meta issues and develop an understanding of them? Would not scientists with a better grasp of these issues be better at successfully conducting their research and communicating its implications to their colleagues, particularly those in other fields? This obviously is a hard concept to sell to research program managers who need concrete results on an annual basis to justify their program budgets. Few of us would disagree with the idea that our institutions should provide adequate support for us as we train the next generation of scientists to follow us. Where are the institutional teaching and research assistantships that will make this happen? In South Dakota today, these resources are very scarce. In a “golden age” would there not be selfless cooperation between institu-tions, true sharing of resources in order to achieve the common good? Without pointing a finger in any direction in particular, based on my own experiences over 17 years in South Dakota, I feel safe in observing that cooperation between researchers around the state could be greatly improved, resulting in benefits both to science and to the researchers involved. In my opinion, such improvements require changes in attitudes of mid- and upper-level academic administrators as well as lower down in the hierarchy at the individual researcher level. Taking these factors as a group, there are a few at the School of Mines who speculate that some decades ago conditions might have more closely approached some definition of “golden”, however tarnished, compared to the present. My retired colleague Cyrus Cox, who taught electrical engineering for 40 years at the School of Mines beginning in 1952, recently reminisced that when he ar-rived on campus there were only 5 people on staff who were not faculty. Among these were the president and the groundskeeper. Faculty constituted a substantial majority of all employees. Today at this same institution, less than 1/3 of em-


ployees are faculty. If our institutions were not so top heavy with administration, constructing administrative mazes to control activities at our institutions, would not our lives as faculty and as researchers be more ideal? Professor Cox hints that this might be so. Or does increased administrative staff in some way work out to be a net benefit to research activities? Does our image of a “golden age” need revision? The Academy proceedings yield cross-sectional views of science in the state beginning with the founding of the Academy in 1915. I have been in South Dakota only since 1987, and attended my first Academy meeting when it met in Rapid City in 1993, just over one decade ago. In preparing these thoughts, after suitable procrastination, the meeting was fast approaching before I even got to the library. I judiciously limited my search through the proceedings for evidence of progression toward or away from a “golden age” to the last 20 years of Academy presidential addresses. Most addresses in this era from 1983-2002 were thoughtful and interesting. They included a mixture of disciplinary talks, state-of-science talks, and an occasional talk on pedagogy, so there was not a new assessment of the state of science in the state every year. Among those addresses on the state of science, a few themes and trends emerged, but no evidence for progression toward or away from a “golden age”. One strong theme appeared in the 1990’s when there were several discus-sions dealing with the “small state syndrome”. This syndrome is characterized by South Dakotans making apologies for their lack of resources to conduct sci-ence, and the limitations on the types of science that can be conducted that are imposed by this deficiency. I believe sensitivity to this issue was heightened with the beginnings of the various EPSCoR program competitions funded by several federal science agencies beginning in the 1980’s, which in some sense made us more sensitive to the disparities between science in many other states and in South Dakota. However, we were a small state in 1984 and still are small today. Another theme from this same period is featured most strongly in Sharon Clay’s 1997 address when she depicted the depressed morale within the state sys-tem of higher education at that time. This was aggravated by enhanced controls imposed from above (Pierre) and by generally increased teaching loads caused by loss of faculty lines during the “reinvestment for efficiency” conducted within the state system of higher education at the time. However, the same morale problems, some times stronger, some times weaker, have existed at least since I arrived in South Dakota in 1987 If there ever was a golden age for science in South Dakota, it must have hap-pened before 1984! With the meeting only days away, I found time for a quick survey of presi-dential addresses from the first 20 years of academy meetings, 1915-1934. If a golden age was to be found, might it not be as far removed as possible from the present? During the 1915-1934 era there was virtually no federal support for sci-entific research anywhere and Academy members had to find support at local levels for their activities. Science then in South Dakota could be characterized as “small”, in contrast to the “large” science that became much more common in the more populous states following World War II. It is clear that Academy

members had somewhat more autonomy in conducting their research during this earlier era, but that resources were even more of a limiting factor then than they are now. The Academy was more heterogeneous then than now, including secondary school teachers, a significant contingent of government workers, and even church ministers, so that topics discussed at meetings covered a spectrum somewhat broader than at our more recent meetings. The first two presidential addresses were delivered by Hilton Ira Jones in 1915 and 1916. In them he focuses much of his attention on the state of science in South Dakota at the time. He emphasizes the need for the state government to start funding research applied to state problems (a need still much in the public eye today!), including such items as surveys of state geological and agricultural resources, to support economic development. He complains strongly about the attitude of legislators and college administrators who regard teaching as the only important duty of faculty members, and mentions that faculty members at the time were typically spending 15-25 hrs/wk in the classroom. Do these concerns sound familiar? Do conditions then resemble even re-motely the criteria for a “golden age” discussed above? If not in that era, might there be a “golden age” somewhere between the teens and the present? There were some very lean years in the late ‘20’s and early ‘30’s. The “Uni-versity” (now USD) published the Academy proceedings from 1915 onward, but due to financial constraints the collected proceedings from 1928-1934 were published in 1934 as one very slim volume. The Depression years certainly were not a “golden age” for science in South Dakota, in any sense of the term. If there has been a “golden age” during our period of record, then it must have occurred after the Great Depression and before the second term of the Reagan presidency, somewhere between 1934 and1984. Before our 2005 meet-ing, I pledge to finish my survey of Academy proceedings and report on this middle period, although I truly think it is unlikely a “golden age” will be found in there. Perhaps there hasn’t been a “golden age” yet, but there will be one in our fu-ture? A little optimism makes present impediments easier to bear. My own hunch is that the search for a golden age will yield a result other than what we seek, perhaps as the Spanish found the High Plains prairies instead of cities of gold in their search for the 7 cities of Cibola. Or perhaps, like Dorothy in Frank Baum’s Land of Oz, we have visited the place we truly seek, but not realized it? Taking this latter line of thought, I would like to close with a quote from the 1922 presi-dential address to the Academy by A. N. Hume, a soil scientist at “State” (now SDSU). This was delivered following the end of the tragic first great world war of the century, and subsequent world-wide flu epidemic, at a time when South Dakota was growing in population and economic activity.

…may I submit very seriously that the scientific spirit is indeed the hope of the world. Instead of finding just rare moments, perhaps once a year, when we are not seeking students or religious converts or adherence to some politi-cal faith or favor with men - in fact when we are not seeking anything at all except that some problem may be made to appear in the light of truth


– such moments ought to become not exceedingly rare, but they ought to be the rule of life.

We need not flatter ourselves that our meeting here will have much visible effect upon society, but we may get some peace of mind out of the fact that the spirit in which we meet is indeed the spirit which can make for human progress. If the South Dakota Academy of Science shall diminish by a little the rarity with which groups of people assemble themselves together to con-template truth for truth’s sake, it will do more to avert conflict on earth than an ordinary peace conference.

Merely by participating in Academy meetings and functions, we are at least visiting, if not taking up permanent residence, in a kind of “golden age”, with a diversity of topics being discussed thoughtfully and respectfully, and with new ideas and collaborations for research being fostered. It is doubtful that our Acad-emy will be able to do much directly concerning significant increases in research funding or the conditions under which its members conduct their activities at their home institutions, but perhaps through our discussions we can contribute indirectly towards improvements in these areas. It is a privilege to be a member of our Academy and as an officer to help facilitate its activities. I look forward to an interesting and challenging year as we continue our efforts to encourage the growth of scientific research, and the dissemination and applications of scientific results, for the benefit of the people of our state and our planet.


Complete Senior Research Papers

presented at

The 89th Annual Meeting

of the



USE OF LOW VACUUM ELECTRON MICROSCOPY TO QUICKLY ESTIMATE BACTERIAL

POPULATIONS ON INCUBATING SALMONID EGGS

Michael E. BarnesSouth Dakota Department of Game, Fish and Parks

McNenny State Fish HatcherySpearfish, SD 57783

Hans Stephenson and Mark Gabel Department of Biology

Black Hills State UniversitySpearfish, SD 57799

ABSTRACT

Low vacuum scanning electron microscopy (SEM) was used to quickly estimate microbial populations attached to the external egg membrane of land-locked fall chinook salmon (Oncorhynchus tshawytscha) eyed eggs. The eggs required no preservation or treatment prior to placement inside the SEM in low vacuum mode, and could be viewed for approximately 10-15 min before severe desiccation occurred. Bacterial numbers were estimated for 7 d post-eyed stage eggs reared in vertical-flow tray incubators and treated with either 1,667 mg/L formalin for 15 min daily in the AM or 1,667 mg/L formalin for 15 min twice daily (AM and PM). Estimated bacterial levels were twice as high on the eggs, and percent hatch was significantly lower, in the AM-treated trays compared to the eggs in trays receiving treatments in both the AM and PM. As a result of estimated bacterial loads, three trays in the AM group were shifted to both AM and PM treatments at eight days post-eyed stage. Survival to hatch in these was significantly greater compared to trays that were not switched. This study documents the feasibility of using low vacuum SEM to quickly estimate bacterial populations on eyed eggs and subsequently customize anti-microbial chemical treatments during hatchery rearing.

Keywords

Chinook salmon, Oncorhynchus tshawytscha, electron microscope, bacteria, eyed eggs, formalin

INTRODUCTION

Enumeration of microbial populations associated with salmonid egg incu-bation has been performed using culture media and plating techniques (Trust


1972; Barker et al. 1989, 1991; Barnes et al. 1999, 2000a). More recently, Ste-phenson et al. (2003) used high vacuum scanning electron microscopy (SEM) to count individual bacteria on landlocked fall chinook salmon (Oncorhynchus tshawytscha) eyed eggs. The results from these techniques are valid. However, all of them are labor intensive and have a potential lag time of several days between the collection of samples and reporting of the results. Because bacteria may play some role in egg mortality (Sauter et al. 1987; Barker et al. 1989, 1991; Stephenson et al. 2003), estimating the size of bacterial populations during egg incubation could provide the basis for a prescribed anti-microbial chemical treatment regime. Such treatments are usually conducted at the same concentration and duration throughout incubation, despite fluctua-tions in microbial levels (Trust 1972; Barnes et al. 1997, 2000b, 2001). The objective of this study was to evaluate the use of low vacuum SEM to quickly estimate the number of bacteria attached to the external membrane of in-cubating landlocked fall chinook salmon eggs, so that anti-microbial treatments could be quickly adjusted in a production setting.

METHODS

A JEOL 5600LV SEM (JEOL USA, Inc., Peabody, Massachusetts) was used for this study. It was operated in low vacuum mode at a magnification of 3000x, acceleration voltage of 20 kV, and a pressure of 30 to 33 Pa. Eggs were not pre-served or treated in any way prior to viewing under low vacuum conditions. The landlocked fall chinook salmon eggs examined under the microscope were spawned on 10/15/2002 and incubated under the conditions described by Barnes et al. (1999). After autopicking to remove any dead eggs on 11/11/2002, nine Heath vertically-stacked incubator trays (Flex-a-lite Consolidated, Tacoma, Washington) were loaded at 6,422 (1500 ml) eyed eggs per tray. Six trays re-ceived formalin (37% formaldehyde, 6 to 14% methanol; Parasite-S, Western Chemical Inc., Ferndale, Washington) treatments at a concentration of 1,667 mg/L for 15 min daily at 08:00. Three trays received the same formalin treat-ment at 08:00, but also received an additional, identical formalin treatment at 15:30. Three of the trays in the once-daily treatment stack were switched on 11/20/2002 (eight days post-eyed stage) to twice-daily formalin treatments until complete hatch on 12/2/2002. Chemical treatments were started on 11/12/2002 and discontinued on 12/2/2002, at which time egg mortalities were recorded from all nine trays. All formalin treatments were administered using a Masterflex model 7524-00 microprocessor peristaltic pump (Cole-Parmer Instrument Company, Chicago, Illinois). Because of the well-established benefits of post-eyed stage formalin treatments on Lake Oahe chinook salmon eggs (Barnes et al. 1997; 2003), and because the primary thrust of this experiment was to evaluate the utility of view-ing live eggs in a low vacuum SEM, no true control (i.e. trays that received no chemical treatments) was included in this experiment. Ten to twelve eggs were removed from the bottom tray in each treatment group on 11/19/2002 and transported in plastic bags containing incubation


water 18 km (20 min) from McNenny State Fish Hatchery to the electron mi-croscope at Black Hills State University, Spearfish, South Dakota. The procedure employed during the electron microscopy session was as follows, with the ap-proximate time for each step in parenthesis:

1. The SEM was focused under high vacuum conditions on a lead ball roughly the same size as the salmon eggs to be viewed, and subsequently the microscope was switched to low vacuum mode (5 min).

2. A single salmon egg was placed on a delicate task wiper, blotted to remove excess water, and put into the microscope chamber (2 min).

3. The egg was viewed and its image was captured. The number of bacteria attached to the external egg membrane was estimated from two randomly selected sites (15 min).

4. Steps 2 and 3 were repeated until three eggs from each treatment had been viewed. Bacterial estimates from the eggs were averaged and converted to number of bacteria/mm2.

Data were analyzed using analysis of variance with the SPSS (9.0) statistical analysis program (SPSS 1999). Pairwise mean comparisons were performed us-ing Fisher’s Protected Least Significance Difference, with significance predeter-mined at P < 0.05 (Ott 1984). All embryo survival percentage data were arcsine transformed prior to analysis to stabilize the variances (Ott 1984).

RESULTS AND DISCUSSION

The external egg membranes of the salmon eggs were clearly visible for approximately 15 min until egg desiccation occurred. No fungal growth was observed on any of the eggs, but bacteria could be readily observed. There were an estimated 1,500 to 3,000 bacteria/mm2 from the eggs receiving once daily formalin treatments and between 0 and 1,000 bacteria/mm2 on the eggs treated with formalin twice daily. Survival to hatch was significantly different between the trays of eyed-eggs receiving once-daily formalin treatments through hatch, compared to those trays receiving twice-daily formalin treatments and those trays that were switched from once-daily to twice-daily treatments midway during incubation (Table 1). However, the increase in survival with increasing formalin treatment may not be biologically significant, because the eggs receiving once-daily formalin experi-enced only 0.75% greater mortality than eggs in the other two groups.


Table 1. Mean (± SE) percent survival to hatch for eyed landlocked fall chinook salmon eggs incubated in vertically-stacked incubation trays subjected to one of three formalin treatment regimes. Means followed by different letters are significantly different at P ≤ 0.05 (n = 3).

Treatment Hatch

Daily 15 min, 1,667 mg/L formalin 94.79 + 0.10zDaily 15 min, 1,667 mg/L formalin for first 9 d followed by twice for the remaining 13 d

95.65 + 0.24y

Twice daily 15 min, 1,667 mg/L formalin 95.54 + 0.19y

Our results show that using a low vacuum SEM is a novel technique that can be used to very quickly estimate microbial populations on salmonid eggs. It is also possible that using a multiple specimen holder for the SEM could reduce the time required for estimation even further. Although this would save loading time, it would also increase the risk of specimen dehydration. We recognize that scanning electron microscopes with low vacuum capa-bilities might not be in close proximity to many hatcheries incubating salmonid eggs. However, in those situations where it is available, it is possible that chemi-cal usage could quickly be adjusted in relation to estimated microbial popula-tion size. Treatments could be increased as microbial populations increase and decreased as populations decrease. Although in our experiment chemical treat-ments were increased in response to elevated microbial populations, it is highly probable that chemical treatments could also be adjusted downward, particularly during the lower bacterial numbers associated with initial egg incubation (Barnes et al. 1999, 2000a). Given the public pressure to decrease chemical outflows in hatchery effluents (Winton 2001), methods to decrease hatchery chemical dis-charges or even more quantitatively justify the use of chemicals during hatchery rearing are extremely important. Costs associated with chemical use during egg incubation might also decrease with increased monitoring of microbial popula-tions attached to incubating eggs.

ACKNOWLEDGEMENTS

We thank the BRIN Undergraduate Fellows Program, Black Hills State University Nelson Grant, the spawning crews at Whitlocks Spawning Station, the hatchery staff at McNenny State Fish Hatchery, NSF Grant # BCS 9871165, and the South Dakota State Library reference librarians.

LITERATURE CITED

Barker, G. A., S.N. Smith and N.R. Bromage. 1989. The bacterial flora of rainbow trout, Salmo gairdneri Richardson, and brown trout, Salmo trutta L., eggs and its relationship to development success. Journal of Fish Diseases 12:281-293.


Barker G.A., S.N. Smith and N.R. Bromage. 1991. Commensal bacteria and their possible relationships to mortality of incubating salmonid eggs. Jour-nal of Fish Diseases 14:199-201.

Barnes, M.E., R.J. Cordes and W.A. Sayler. 1997. Use of formalin during in-cubation of eyed eggs of inland fall chinook salmon. The Progressive Fish-Culturist 59:303-306.

Barnes, M.E., A.C. Gabel and R.J. Cordes. 1999. Bacterial populations dur-ing inland fall chinook salmon egg culture in vertical-flow tray incubators. North American Journal of Aquaculture 61:252-257.

Barnes, M.E., A.C. Gabel and R.J. Cordes. 2000a. Bacterial populations during rainbow trout egg culture in vertical-flow tray incubators. North American Journal of Aquaculture 62:48-53.

Barnes, M. E., K. Wintersteen, W. A. Sayler, and R. J. Cordes. 2000b. Use of formalin during incubation of eyed rainbow trout eggs. North American Journal of Aquaculture 62:54-59.

Barnes, M. E., W. A. Sayler, and R. J. Cordes. 2001. Use of formalin during incubation of eyed brown trout eggs. North American Journal of Aquacul-ture 63:333-337.

Barnes, M. E., H. Stephenson, and M. Gabel. 2003. Use of hydrogen perox-ide and formalin treatments during incubation of landlocked fall chinook salmon eggs. North American Journal of Aquaculture 65:152-154.

Ott, L. 1984. An introduction to statistical methods and data analysis. PWS Publishers, Boston.

SPSS. 1999. SPSS Base 9.0. SPSS, Inc. Chicago, Illinois.Sauter, R. W., and five coauthors. 1987. A study of bacteria present within

unfertilized salmon eggs at the time of spawning and their possible relation to early lifestage death. Journal of Fish Diseases 10:193-203.

Stephenson, H, M. Gabel, and M. E. Barnes. 2003. Microbial inhibition in response to treatments of hydrogen peroxide and formalin on landlocked fall chinook salmon eyed eggs as determined by scanning electron microscopy. North American Journal of Aquaculture 65:324-329.

Trust, T.J. 1972. The bacterial population in vertical flow tray hatcheries during incubation of salmonid eggs. Journal of the Fish Research Board of Canada 29:567-571.

Winton, J.R. 2001. Fish health management. Pages 559-640 in G. Wedemeyer editor. Fish Hatchery Management, second edition. American Fisheries So-ciety, Bethesda, Maryland.


RELATIVE ABUNDANCE OF SMALL MAMMALS IN NATIVE AND RESTORED TALLGRASS PRAIRIE

Samuel J. Kezar and Jonathan A. JenksDepartment of Wildlife and Fisheries Sciences

South Dakota State UniversityBrookings, SD 57007

ABSTRACT

Relative abundance was assessed for small mammals captured on native and restored tallgrass prairie habitats. Prairie voles (Microtus ochrogaster) (n = 30), meadow voles (Microtus pennsylvanicus) (n = 4), deer mice (Peromyscus spp.) (n = 5), northern short-tailed shrews (Blarina brevicauda) (n = 3), pygmy shrews (Sorex hoyi) (n = 2), and thirteen-lined ground squirrels (Citellus tridecemlineatus) (n = 3) were captured using snap traps. Relative abundance of voles (Microtus spp.) was higher in native prairie (163.3) than in restored prairie (10.2). Rela-tive abundance of deer mice (20.4), thirteen-lined ground squirrels (13.3), and shrews (20.4) was higher in restored than native prairie (< 5.1) habitat. Results indicated that restored prairie habitats provide adequate components (forage, cover) to support viable small mammal communities.

Keywords

Native prairie, prairie voles, Microtus ochrogaster, relative abundance, re-stored prairie, small mammals, South Dakota

About 2.6 million ha (6.5 million acres) of land consists of pasture or hay lands in South Dakota (Johnson and Larson 1999). In eastern South Dakota, agricultural lands represent the most abundant habitat for wildlife (Smith et al. 2002). For example, in 2002, 111,500 ha (275,500 acres) of row crops were harvested in Brookings County, South Dakota (SDASS 2003). Small mammals use these habitats and represent an integral part of wildlife populations (Sullivan et al. 2003). Pinkert et al. (2002) documented the presence of deer mice (Pero-myscus maniculatus) and white-footed mice (Peromyscus leucopus) in cropland habitats in eastern South Dakota. Limited information is available on small mammal abundance within the tallgrass prairie (both native and restored) tracts that characterize eastern South Dakota. The purpose of this study was to determine if small mammal abundance differed between native and restored tallgrass prairie habitats. We hypothesized that relative abundance of small mammals would be higher in native (undis-turbed) prairie than in restored (seeded) prairie habitat.


STUDY AREA

Habitat for this study was limited because most native tallgrass prairie is privately owned in eastern South Dakota (Higgins et al. 2000). Brookings Prairie is 16 ha in size and is owned by the City of Brookings, South Dakota. Geographical location of the prairie is North 44º 15’ 09.0”,West 96º 48’ 39.9”. Native tallgrass prairie at this site was characterized by grasses such as big blue-stem (Andropogon gerardii), switchgrass (Panicum virgatum), little bluestem (Schizachyrium scoparium), various sunflowers (Helianthus spp.), sageworts (Artemisia spp.), goldenrods (Solidago spp.), and purple coneflower (Echinacea angustifolia) (Johnson and Larson 1999). In addition, non-native species (e.g., smooth brome, Bromus inermis) also characterize the vegetation. The restored prairie site was a 64 ha section of privately owned land located at North 44º26’ 37.1”,West 96º 48’ 17.7”. Restored prairie grasses on this site included big bluestem, Indiangrass (Sorghastrum nutans), and switchgrass, which was the most abundant grass at the site.

MATERIALS AND METHODS

Sampling was conducted from 11 to 15 August 2003. Mouse-sized snap traps (Victor®, Woodstream Corporation, Lititz, PA) (Stickel 1946) were baited with a 50:50 mixture of oatmeal and peanut butter (Schemnitz 1996). Using a 7 x 7 trap grid, traps were placed 9.09 m apart as described by Pinkert et al. (2002). Assuming that each trap covered a radius of 4.55 m, the total sample area covered was 0.405 ha (1 acre) (Pinkert et al. 2002). Traps were numbered 1 to 49 and placed numerically left to right, top to bottom, starting in the north-east corner of the grid. Each site was checked twice daily; at sunrise and no earlier than 2 hours before sunset. Traps with a captured animal were replaced with new traps to reduce pheromones that could affect capture rates. Animals captured were placed on ice for later identification. Abundance was calculated as the total number of individuals caught per 1000 trap nights. Relative abundance was determined by individual species of small mammal and for total number of small mammals captured. Alpha level for comparisons was set at 0.05 and Chi-Square tests were used to compare frequencies of small mammals captured from the two habitats. Analysis was conducted at the genus level (i.e., Microtus, for both prairie [Microtus ochrogaster] and meadow voles [Microtus pennsylvanicus]). An index of species richness was calculated for the two habitats where;

Species Richness = number of individuals of species captured total number of small mammals captured

Kruskal-Wallis One-Way Analysis of Variance was used to test for significant dif-ferences in relative abundance. Coefficient of Variance also was calculated as an index to variability in small mammal populations within the two habitats. All statistical analyses were conducted using Program Systat (SPSS 2000).


RESULTS

A total of 47 individual small mammals was captured over 392 trap nights (Table 1). Prairie voles (n = 30) were the most abundant species captured fol-lowed by deer mice (n = 5), meadow voles (n = 4), thirteen-lined ground squirrels (n = 3), northern short-tailed shrews (n = 3), and pygmy shrews (n = 2) (Table 1). A total of 34 small mammals was captured on the native prairie site wherease 13 small mammals were captured on the restored prairie site. Prairie voles (n = 28) were the most abundant small mammals captured in native prairie, whereas deer mice (n = 4) were most abundant small mammal captured in restored prairie.

Table 1. Small mammals captured 11-15 August 2003 in Brookings County, South Da-kota.

Pero-myscus

Prairie Vole

Meadow Vole

Thirteen-Lined

Ground Squirrel

Northern Short-Tailed Shrew

Pygmy Shrew

Total

Native Prairie 1 28 4 0 0 1 34Restored Prairie 4 2 0 3 3 1 13Total 5 30 4 3 3 2 47

Small mammal relative abundance differed (χ23 = 150.24, P <0.001) for na-

tive and restored prairie habitats. Relative abundance of voles (Microtus spp.) was higher in native prairie than in restored prairie (Table 2). Deer mice, thir-teen-lined ground squirrels, and shrew relative abundance was higher in restored than in the native prairie habitat. Pygmy shrew relative abundance was similar across prairie sites. The index of species richness for restored prairie (0.83) was higher than native prairie (0.67). Small mammal abundance was 6.25 times more variable in native than the restored prairie.

Table 2. Relative abundance of small mammals captured 11-15 August 2003 in Brookings County, South Dakota.

Peromyscus Microtus Thirteen-Lined Ground Squirrel

Pygmy/Northern Short-Tailed Shrew

Total

Native Prairie

5.10 163.27 0.00 5.10 173.46

Restored Prairie

20.41 10.20 15.31 20.40 66.32

DISCUSSION

Pearson and Ruggiero (2003) found that transect arrangements for small mammal trapping surveys were more efficient than trap grids. However, trap grids were used in this experiment to allow comparisons to previously published information on small mammal populations in eastern South Dakota. To mini-


mize affects of adjacent habitat, trap grids were placed at a maximum distance from edge habitat. Native prairie was an important habitat for prairie voles. Dense vegetation and a well-developed litter layer in open prairie are common habitat components used by prairie voles (Walker 1976). Brady and Slade (2001) found that prairie vole abundance had little ecological effect on other small mammal communities, which suggests that abundance of voles captured was not an indication of domi-nance. Moreover, vole abundance, in itself, may not indicate habitat quality (Van Horne 1984). There was a considerable difference in the relative abundance of the species captured (Table 2). Microtus species captured in the native prairie site had the highest relative abundance (163.27). However, Peromyscus species and shrew species had the highest relative abundance in the restored prairie site. Both Pinkert et al (2002) and Terrall et al. (2002) documented higher relative abun-dance of Peromyscus species and shrews than microtus species in grassland habitats in eastern South Dakota. Their results were similar to those for the restored prairie habitat in our study and may indicate that many grassland tracts of land in eastern South Dakota have small mammal populations indicative of restored prairie. Based on small mammals captured at the restored prairie site, seeded prairie did provide suitable habitat for a variety of small mammals, albeit of lesser rela-tive abundance. Diversity of forbs at the restored site was limited when com-pared to the native prairie site. Yet, despite the higher variety of forbs on the native prairie site, our index of species richness was higher for the restored prairie site; an indication that the site did provide a variety of niches for small mammals. Alder (1988) and Sieg (1988) stated that habitat variability can have a profound affect on the distribution of mammals. Results indicate that restoration efforts are capable in reproducing habitats suitable for small mammal communities.

ACKNOWLEDGMENTS

Special thank you to South Dakota State University for support, City of Brookings and private land owner L. Carson for access to his property, Dr. G. E. Larson for peer review, and C. L. Kezar for help with field work.

LITERATURE CITED

Alder, H. A. 1988. The role of habitat structure in organizing small mammal populations and communities. Proceedings of a symposium on the manage-ment of amphibians, reptiles, and small mammals in North America. Pages 289-299. in U.S.D.A. Forest Service General Technical Report. RM-166.

Brady, M. J., and A. Slade. 2001. Diversity of a grassland rodent community at varying temporal scales: the role of the ecologically dominant species. Jour-nal of Mammalogy 82: 974-984.


Higgins, K. F., V. J. Smith, J. A. Jenks, J. J. Higgins and G. A. Wolbrink. 2000. A provisional inventory of tall grass prairie tracts remaining in eastern South Dakota. South Dakota Agricultural Experiment Station Report EC912, South Dakota State University, Brookings, South Dakota. 123pp.

Higgins, K. F., E. D. Stukel, J. M. Goulet, and D. C. Backlund. 2000. Wild Mammals of South Dakota. South Dakota Department of Game, Fish and Parks, Pierre, South Dakota.

Johnson, J. R., and G. E. Larson. 1999. Grassland Plants of South Dakota and the Northern Great Plains. South Dakota State University, College of Ag-ricultural and Biological Sciences, Ag. Communications, Brookings, South Dakota.

Pearson, D. E. and L.F. Ruggiero. 2003. Transect versus grid trapping arrange-ments for sampling small-mammal communities. Wildlife Society Bulletin 31(2): 454-459.

Pinkert, M. K., J. R. Meerbeek, G. D. Scholten, and J. A. Jenks. 2002. Impact of crop harvest on small mammal populations in Brookings County, South Dakota. Proceedings of the South Dakota Academy of Science 81:39-45.

Schemnitz, S. D. 1996. Capturing and Handling Wild Animals. Research and Management Techniques for Wildlife and Habitats. The Wildlife Society. Edited by T. A. Bookhout. Fifth Edition: 106.

Sieg, C. H. 1988. The value of Rocky Mountain juniper (Juniperus scopulorum) woodlands in South Dakota as small mammals in North America. Pages 328-332. in U.S. Forest Service General Technical Report. RM-166.

Smith, V. J., J. A. Jenks, C. R. Berry, Jr., C. J. Kopplin, and D. M. Fecske. 2002. The South Dakota Gap Analysis Project. Final Report. Research Work Or-der No. 65. Department of Wildlife and Fisheries Sciences, South Dakota State University, Brookings, S.D.

South Dakota Agricultural Statistics Service. 2003. Acres planted and harvested, SD., 1998-2002. South Dakota Agriculture Statistics Bulletin. U. S. Depart-ment of Agriculture.

SPSS. 2000. SYSTAT Version 10. SPSS Science and Marketing Department, Chicago, Illinois.

Stickle, L. F. 1946. Experimental analysis of methods for measuring small mam-mal populations. Journal of Wildlife Management 10: #150-159.

Sullivan, T. P., D. S. Sullivan, D. B. Ransome, P. M. F. Lindgren. 2003. Impact of removal-trapping on abundance and diversity attributes in small-mammal communities. Wildlife Society Bulletin 31:464-474.

Terrall, D. F., N. G. Cochran, and J. A. Jenks. 2002. Variation in small mam-mal richness among ecotypes in eastern South Dakota. Proceedings of the South Dakota Academy of Science 81:147-152.

Van Horne, B. 1983. Density as a misleading indicator of habitat quality. Jour-nal of Wildlife Management 47:893-901.

Walker, J. A. 1976. Relative abundance and habitat preference of some small mammals on southeastern Nebraska. Transactions Nebraska Academy of Science 3:74-77.


ASSESSING POTENTIAL TRANSPORTOF TYLOSIN IN THE LANDSCAPE

Jared K. Oswald and Todd P. TrooienAgricultural and Biosystems Engineering Department

Sharon A. Clay and Zhuojing LiuPlant Science Department

Robert ThalerAnimal and Range Science Department


ABSTRACT

Tylosin is an antibacterial chemical used in livestock production. Tylosin is highly water soluble, only partially sorbed to manure, and potentially excreted by animals that receive it in their feed. Thus, tylosin can be contained in manure and is potentially mobile, but the fate of tylosin after application to agricultural fields is unknown. The objectives of this project were: (1) determine the concen-tration and mass transport in runoff of tylosin, applied either alone or within manure, (2) determine the effect of manure on infiltration rate, (3) determine the effect of tylosin in the manure on infiltration rate, and (4) determine the ef-fect of landscape position on infiltration rate. A small sprinkler infiltrometer was used to measure infiltration rate at three landscape positions in a field. Runoff was collected and tylosin concentrations were measured. When 176 mg of dry tylosin was applied to the soil surface, up to 6% (10.4 mg) of the applied tylosin was recovered in the runoff. When manure containing 0.344 mg of tylosin was applied, up to 23% (0.08 mg) of the tylosin was recovered in the runoff. Infiltra-tion rates were reduced by manure and by implement traffic. These results show that tylosin can be transported in runoff. In areas where manure that contains tylosin is applied to fields, runoff reduction or prevention measures may be re-quired to prevent tylosin from reaching surface water resources.

Keywords

Runoff, infiltration, antibacterial chemical, manure, transport

INTRODUCTION

Antibacterial chemicals are often fed to animals at subtherapeutic levels for growth promotion. Approximately 7.8 to 11 million kg of antibacterial chemi-cals are used annually in animal production; approximately 1.3 million kg are


used annually in humans (Hileman, 2001). Antibacterial chemicals that have been fed to animals can be excreted from the animal unchanged (Halling-Sorenson et al., 1998). Once in the environ-ment, these chemicals can disrupt biogechemical cycles or become pollutants. Many public health officials are concerned that resistance can result from the wide-scale use of antibacterials in animals. One common antibacterial chemical fed to swine is tylosin. The sorption of tylosin to soil (unpublished data) is similar to atrazine (Liu et al., 1997), making tylosin potentially mobile in the landscape. The environmental fate of antibacterial chemicals such as tylosin, when ap-plied in either solid or liquid animal wastes, is unknown. One potential pathway for transport of tylosin is in runoff water. If tylosin leaves agricultural fields in runoff, it may be transported to surface waters and adversely affect the aquatic ecosystem. Given the potential mobility of tylosin, the goal of this study was to determine if this commonly used antibacterial chemical can be transported in runoff water. Because runoff and infiltration rate are so closely related, the change of infiltration rate (if any) due to tylosin and manure (in which the ty-losin is excreted) must also be understood. The specific objectives of our study were: (1) determine the concentration and mass transport of tylosin, applied either alone or within manure, in runoff, (2) determine the effect of manure on infiltration rate, (3) determine the effect of tylosin in the manure on infiltration rate, and (4) determine the effect of landscape position on infiltration rate.

PROCEDURE

The research site was located in the SE 1/4 of section 18, R48W, T107N, Moody County, SD. The soil is classified as Houdek clay loam. The slope on which this research took place was 100 m long with 3 m of elevation difference, resulting in a consistent slope of approximately 3%. The site has been well man-aged and relatively little soil erosion has taken place in recent years. The site had corn (Zea mays L) growing during the test and crop residue from the previous year’s soybean (Glycine Max [L] Merr.) was on the soil surface. Tylosin had never been applied to this field. For the analysis of tylosin in runoff, four different treatments were analyzed at three different landscape positions and in trafficked and nontrafficked inter-rows. The treatments were an untreated control, broadcast application of dry tylosin as Tylan 40, manure from swine that had ingested feed containing tylosin as Tylan 40 at the label rate, and manure from swine that did not receive tylosin in their feed. The landscape positions were toe slope, back slope, and shoulder slope. Each treatment was replicated twice, resulting in a total of 48 plots. Infiltration rates were measured with a Cornell Sprinkler Infiltrometer (Og-den et al., 1997). The water-filled infiltrometer was placed on top of the 240-mm steel cylinder pounded 75 mm into the ground and leveled. The initial height of the water column inside the infiltrometer was recorded. Runoff samples were col-lected every two minutes for tests run on trafficked sites and every three minutes on nontrafficked sites. The application rate was calculated as the original water


column height minus the final water column height, divided by the time of ap-plication. The infiltrometer used a marriotte tube to maintain constant applica-tion rate. The application rate was approximately 230 mm/hr, which is similar to a 50-year, 10-minute storm for eastern SD (Hershfield, 1961). Infiltration rate was calculated by subtracting measured runoff rate from application rate. Steady state infiltration rate was calculated as the average of the infiltration rates at the end of the test. The actual number of observations used to calculate the steady state infiltration rate was determined by inspection for each plot. The Cornell infiltrometer applies water by dripping from about 100 mm above the soil surface. This low elevation of water application means that water drops have little chance to gain kinetic energy due to gravity. Thus, drops strike the soil surface with relatively little energy. This reduces the formation of a crust that might seal the soil surface and reduce the infiltration rate (Mohammed and Kohl, 1987), resulting in greater measured infiltration rates. However, ample runoff is still generated because of the high application rate. Treatments were applied 24 hours prior to the infiltration tests. Major debris was cleared from the treatment area before pounding in the steel cylinder. For the control plots, 75 mm of water was applied inside the cylinder to saturate the soil. For the dry tylosin treatment, 75 mm of water was applied then 2 g Tylan 40 (containing 176 mg tylosin) was broadcast evenly within the cylinder after the water had infiltrated. For the manure and manure/tylosin treatments, 82 grams of solid fresh manure (either with or without tylosin) was mixed with approxi-mately 1 liter of liquid manure and applied inside the cylinder. This application was equivalent to 17500 kg of solid manure per ha. The average tylosin concen-tration in the manure was 4.2 μg of tylosin per g of solid manure, resulting in an average mass 0.344 mg of tylosin per plot. No statistical tests were performed on any of the data in this study because of the small number of replications. Except where noted, results are presented as the averages of two plots. Water was analyzed directly, without dilution or concentration, for tylosin (as tylosin phosphate) in 10 μl samples with LC/MS (Rabolle and Spliid, 2000). Tylosin in a manure matrix was extracted with a methanol/water mix, the metha-nol was evaporated off, and the aqueous phase was analyzed for tylosin with LC/MS. The limit of quantitation (LOQ) was 0.01 mg/l. The limit of detection was less than the LOQ; any tylosin concentration measured as less than 0.005 mg/l was considered to be zero.


Tylosin was found only in the solid swine manure and was not detected in any of the liquid manure samples analyzed. Of the three samples of solid manure analyzed, the average tylosin concentration was 4.2 μg of tylosin/g of solid ma-nure (sample values of 3.8, 3.8, and 4.9 mg/g). Each plot reached a steady state infiltration rate during its test (Fig. 1). The pretreatment of each plot with water allowed each plot to reach steady state with-in the 30 to 45 minute test period. The concentrations of tylosin in runoff were


relatively constant during the test, resulting in a constantly increasing cumulative mass of tylosin transported in the runoff (Fig. 1). It is not known for how long the tylosin transport would continue or what the total transported mass would be.

Application of dry tylosin resulted in transport of 0.1% to 6% of the applied tylosin in the runoff (Fig. 2). The mass of tylosin lost varied from 0.07 to 10.4 mg. In the absence of manure, tylosin transport was consistently greater at lesser infiltration rates. Infiltration rates were reduced by compaction due to imple-ment traffic (Fig. 2).

Figure 1. Runoff and infiltration rates and cumulative tylosin in runoff from a plot in a toe slope location, tylosin treatment, and non-trafficked. The steady state infiltration rate was 139 mm/hr, using the last four observations.

Figure 2. Percent of the applied tylosin leaving the plot in the runoff for the plots re-ceiving 176 mg of dry tylosin.


Application of tylosin in manure resulted in less mass transport but a greater percentage of the applied tylosin was transported (Fig. 3) when compared to application of dry tylosin. The mass of tylosin transported ranged from zero in one plot to a maximum of 0.08 mg, which is 23% of the tylosin applied in the manure. This is equivalent to a transport rate of 18 g/ha. Total mass transport could be greater, if precipitation continued long enough, because tylosin was still being transported at the end of our tests (Fig. 1). The manure reduced the infiltration rates so the effect of increased tylosin transport at lesser infiltration rates was not manifest (Fig. 3).

The addition of manure caused the formation of a green film on the soil surface that formed a seal and reduced infiltration rates. In nontrafficked areas, the control plots had infiltration rates ranging from 162 mm/hr at the toe slope to 234 mm/hr at the shoulder slope position. The manure plot infiltration rates ranged from 29 mm/hr at the toe slope to 33 mm/hr at the shoulder slope (Fig. 4), reductions of 82% and 86%. Manure is generally understood to have many beneficial long-term effects on soils and agriculture. Our data show that manure application can reduce infiltration rates in the short term (24 hours after a heavy application). In trafficked areas, the effect of manure on infiltration rate was inconsistent (Fig. 5). The effect of the traffic and resulting soil compaction reduced the in-filtration rates by the same amount as the addition of manure. The infiltration rates of the manure plots were similar to those in the nontrafficked areas, ranging from 19 mm/hr at the back slope to 38 mm/hr at the shoulder slope position. The infiltration rates of the control plots were much less than those in nontraf-ficked areas, ranging from 12 mm/hr to 45 mm/hr. The greatest infiltration rate decrease occurred at the shoulder slope position, where the infiltration rate aver-age decreased from 210 mm/hr to 12 mm/hr, a decrease of 94%.

Figure 3. Percent of the applied tylosin leaving the plot in the runoff for the plots re-ceiving 0.344 mg of tylosin in manure.


Tylosin in the manure had no effect on infiltration rates. In nontrafficked areas, tylosin/manure plots had an average infiltration rate of 29 mm/hr and the aver-age infiltration rate of the manure plots was 30 mm/hr. The infiltration rate varied by landscape position, with the greatest infiltra-tion rate (average of 216 mm/hr) at the back slope position (Fig. 6). The least infiltration rate among the three landscape positions was at the toe slope, 168 mm/hr. This reduction of up to 22% of the maximum infiltration rate (going from the back slope position to the toe slope position) is much less than the infiltration rate reductions of up to 86% due to manure application (Fig. 4) or of up to 94% due to traffic (Figs. 4 and 5). Tylosin measured in this study was transported in solution. There was very little sediment in the runoff in our tests and no attempt was made to extract tylosin from the small amount of sediment collected in the runoff. Thus, if any tylosin was transported while sorbed to sediment, we did not measure it. These results show that tylosin can be transported in runoff. In areas where manure that contains tylosin is applied to fields, runoff reduction or prevention

Figure 4. Steady state infiltration rate averages for manure and control plots location in nontrafficked areas.

Figure 5. Steady state infiltration rate averages of manure and control plots location in trafficked areas.


measures may be required to prevent tylosin from reaching surface water re-sources. Because the tylosin is transported in solution, erosion or sedimentation transport reduction may not reduce the transport of tylosin.

CONCLUSIONS

1. Tylosin can be transported in runoff. When 176 mg dry tylosin was broad-cast on the soil surface, up to 10.4 mg was recovered in the runoff. When manure containing 0.344 mg of tylosin was applied to the soil surface, up to 0.08 mg was recovered in the runoff.

2. Manure added to the soil decreased infiltration rates in non-trafficked areas by as much as 86% when tested 24 hours after application.

3. The presence of tylosin in the manure had no effect on infiltration rates.4. Infiltration rates varied by landscape position, varying from 168 mm/hr to

216 mm/hr, but the variation was smaller than reductions caused by traffic or manure application.

REFERENCES

Halling-Sorenson, B., S. N. Nielsen, P. F. Lanzky, F. Ingerslev, and H. C. H. Lutzhoft. 1998. Occurrence, fate, and effects of pharmaceutical substances in the environment- a review. Chemosphere 36:357-393.

Hershfield, D. M. 1961. Rainfall frequency atlas of the united states for dura-tions from 30 minutes to 24 hours and return periods from 1 to 100 years. Weather Bureau Technical Paper No. 40, US Weather Bureau, Washington, DC.

Hileman, B. 2001. Furor over animal antibiotic use. Chemical Engineering News. Feb 19, 2001. pg 47-52.

Liu, Z., S. A. Clay, J. Gaffney, and D. Malo. 1997. Atrazine and alachlor adsorp-tion characteristics to benchmark soil series in eastern South Dakota. Proc SD Acad Sci 76:147-158.

Figure 6. Steady state infiltration rate averages for control and tylosin plots at three landscape positions in nontrafficked areas. Each bar is the average of four plots.

Mohammed, D. and R. A. Kohl. 1987. Infiltration response to kinetic energy. Trans ASAE 30(1):108-111.

Ogden, C. B., H. M. van Es, and R. R. Schindelbeck. 1997. Miniature rain simulator for field measurement of soil infiltration. Soil Sci Soc Am J 61:1041-1043.

Rabolle, M. and N. H. Spliid. 2000. Sorption and mobility of metronadizole, olaquindox, oxytetracycline, and tylosin in soil. Chemosphere 40:715-722.

ACKNOWLEDGEMENTS

This research was supported by funds from the South Dakota Agricultural Experiment Station. The LC/MS was purchased with funds from the South Da-kota NSF EPSCoR Grant no. 0091948. We thank Dale Tjarks for allowing us to perform this research on his land. SD AES article number 3431.



FOX SQUIRREL WEIGHT AND AGE STRUCTUREIN MAST AND NON-MAST FORESTS

Steven L. Reighard, Aaron D. Bucholz, Justin A. Haahr and Jonathan A. JenksDepartment of Wildlife and Fisheries Sciences


ABSTRACT

We conducted a 58-day study from 6 September to 2 November 2003 to evaluate size and age structure of fox squirrels (Sciurus niger) in mast and non-mast forests in the upper Midwest. Hourly observation and harvest ratios indicated squirrel density was greater in the mast than non-mast forest. No sig-nificant differences were found in adult male (p = 0.264) or subadult male (p = 0.114) squirrel weights for mast and non-mast forests. Adult female (p = 0.003) and subadult female (p = 0.022) weights were significantly heavier in mast than non-mast forests. Subadult to adult ratios were similar in the mast and non-mast forests indicating high fecundity rates, which may be due to subadult male emigration into the non-mast forest. Results indicated a need for mast crops in agriculture areas for squirrel production.

Keywords

Age structure, agriculture, emigration, forage, fox squirrel, habitat, harvest, immigration, Iowa, mast, Midwest, populations, Sciurus niger, sink.

This study evaluated size and age structure of fox squirrels (Sciurus niger) between a predominately mast (> 80% mast) and non-mast (> 80% non-mast) forest in the upper Midwest. Numerous research projects have examined fox squirrels relative to mast crops (Allen 1948, Nixon et al. 1975, Harty and Stains 1976, Korschken 1981, Nixon et al 1991, Schroeder and Vangilder 1997) and suggested that mast is important as a food source for this species. These studies also indicate that mast availability is related to survival rates of young, fecundity, and immigration/emigration rates. However, others have indicated that habitat fragmentation and mast crop availability have a small impact on fox squirrels in agricultural areas (Fouch 1962, Kline 1964, McClellan 1985, Schmidt and Brown 1996, Nupp and Swihart 2000). Studies also have compared size and age structure of fox squirrels (Thoma and Marshall 1960, McCloskey 1968, Short and Duke 1971, Goodrum 1972, Harnishfeger 1977, Reynolds and Silvy 1983) across the U.S. with variable results. No research is available comparing mast and non-mast forests associated with fox squirrel size and age structure. This study analyzed these parameters to assess the importance of mast forests to squirrel populations. We hypothesized fox squirrels from the mast forest would be larger in size and the population would be younger in age structure compared to the non-mast forest.


STUDY AREA The mast forest, Study Site #1, was 117 ha, on private property, and owned by Anderson Farms Inc., Oliver Shonkwiler Farms, and John Paton Farms. The legal description was E 1/2 Sec 7 and W 1/2 Sec 8, Emmet Township, T-100-N and R-34-W located northwest of Estherville, Iowa (Figure 1). The area was comprised of two large ravines perpendicular to the west fork of the Des Moines River, with many adjoining smaller ravines (Figure 2). The main soil types lo-cated within the site are Lester loam, Millington silt loam, and Vinje silty clay loam (Jones 1997). The forest is adjoined to the north, west, and south by row crop agricultural, and pasture to the east.

Figure 1. Location of study sites in Emmet County, Iowa (Natural Resources Conservation Service 2001).


Black walnut (Juglans nigra), bur oak (Quercus macrocarpa), and northern red oak (Quercus rubra) comprise over 80% of the woody vegetation in this study area. Other species include American basswood (Tilia americana), American elm (Ulmus americana), black maple (Acer nigrum), black willow (Salix nigra), eastern hophornbeam (Ostrya virginiana), green ash (Fraxinus pennsylvanica), sandbar willow (Salix interior), silver maple (Acer saccharinum), slippery elm (Ulmus ru-bra), and sugar maple (Acer saccharum). The non-mast forest, Study Site #2-Ryan Lake, was 128 ha owned by the state of Iowa, and managed by the Iowa Department of Natural Resources. Legal description of the site was the SW 1/4 Sec 27, SE 1/4 SE1/4 Sec 28, NE 1/4 NE

Figure 2. Topographic maps of study sites (Geographical Survey Bureau 2000).


1/4 Sec 33, and the NW 1/4 Sec 34, Center Township, T-99-N and R-33-W lo-cated southeast of Estherville, Iowa (Figure 1). The area is a prairie pothole lake that has been drained into Drainage Ditch #60 (Figure 2). The old lakebed is surrounded by forest and is divided into five sections by four shelterbelts. Open areas and surrounding lands are planted to row crop agriculture. Soil types with-in this site are Blue Earth mucky silt loam, Calcousta mucky silt loam, and Palms muck (Jones 1997). Woody vegetation on this site was over 80% American elm, boxelder (Acer negundo), eastern cottonwood (Populus deltoides), and green ash. Remaining forest coverage was American basswood, black walnut, black willow, burr oak, hackberry (Celtis occidentalis), red mulberry (Morus rubra), Russian olive (Elaeagnus angustifolia), silver maple, and sugar maple.

METHODS

Fox squirrel observation and harvest data were collected in this study. Ob-servation data were collected during hunting outings. The hunting route was circular in pattern to eliminate double counting bias. Squirrels were harvested during the regular open hunting season using non-toxic shot (Iowa 2003). They were collected by still-hunting using 20 gauge shotguns, with 21.3 gram, 6 shot steel loads. A Faulks model LS-85 squirrel call was utilized for location purposes by imitating chatter barks (Zelley 1971). All squirrels were weighed in the lab with an A & D Silver Scale, model SK5000WP, (A & D Company, Tokyo Japan) to the nearest 2 g. Sexing and aging was done by mastology of the females and scrotum assess-ment of the males (Petrides 1951, Kirkpatrick 1955, Thompson 1958, Godin 1960, Dubock 1979, Dimmick and Pelton 1996). This method is the most accurate method to sex and age live or freshly harvested squirrels (McCloskey 1977). Age classes utilized were adult, subadult (spring born), and juvenile (summer born). Hypotheses were evaluated using Lillifor’s tests, paired t-tests, and least squares means statistical analysis in ANOVA (SYSTAT 2002).

RESULTS

Between 6 September and 2 November 2003, 102 fox squirrels were har-vested from mast (n = 61) and non-mast (n = 41) forests. Observation data was normally distributed for the mast (p=0.023) and non-mast (p=1.000) forest. Number of squirrels observed was significantly higher (p = 0.021) in mast (n = 113, 2.58/hr) than non-mast (n = 75, 1.32/hr) forest. Harvest rate was signifi-cantly higher (p = 0.013) in the mast (1.39/hr) than in the non-mast (0.72/hr) forest. Juveniles were excluded from further analysis due to limited data (Table 1).


Table 1. Fox squirrels harvested by sex and age class in mast and non-mast forests in the upper Midwest, 6 September – 2 November 2003.

Forest Type

Adult Male

Adult Female

Subadult Male

Subadult Female

Juvenile Male

Juvenile Female

Total

Mast 10 7 22 14 1 7 61Non-Mast 6 6 19 8 2 - 41

All weight data were normally distributed. Weights did not differ between subadult males (p = 0.144) and adult males (p = 0.264) collected from the two study sites. Adult female weights were significantly larger (p = 0.003) in mast (mean=875g) than the non-mast (mean=778g) forest (Table 2). Subadult fe-males also were significantly heavier (p = 0.022) in the mast (mean=705g) than the non-mast (mean=650g) forest (Table 2). Adult females weighed more than adult males in the mast forest. This was biologically significant (p=0.086) but not statistically significant (≤ 0.05). There was no significant weight difference (p = 0.516) between adult females and adult males in the non-mast forest.

Table 2. Fox squirrel mass (g) by sex and age class in mast and non-mast forests in the upper Midwest, 6 September – 2 November 2003.

Forest Type Adult Male Mean SE

Adult Femal Mean SE

Subadult Male Mean SE

Subadult Female Mean SE

Mast 825 20 875 16 738 18 705 15Non-Mast 793 13 778 20 704 12 650 14

The subadult to adult age structure was similar in the mast and non-mast forests; 2.12:1 and 2.25:1, respectively. Subadult male to adult female ratios were similar in the mast and non-mast forests; 3.14:1 and 3.17:1, respectively. Adult male to adult female ratios were higher in mast (1.42:1) than non-mast (1:1) forests. Subadult female to adult female ratios were higher in mast (2:1) than non-mast (1.33:1) forests. Subadult male to subadult female ratios were lower in the mast (1.57:1) than the non-mast (2.38:1) forest.

DISCUSSION

Results from our study were similar to an Ohio study (Nixon et al. 1974) in that mast crops were shown to be important to fox squirrel populations. Obser-vation and harvest data indicated the fox squirrel population density was greater in the mast than non-mast forest. Many studies have shown that immigration decreases with abundant food availability (Allen 1943, Baumgartner 1943, Nixon et al. 1974, Nixon et al. 1975, Harnishfeger 1978). Fall dispersal of fox squirrels is short in duration (Baumgartner 1943, Nixon et al. 1974) and peaked during our study period (6 September to 2 November) (Allen 1943, Delong 1986). Comparisons of weight between sites indicated food availability and nu-tritional value were higher in the mast forest. Korschgen (1981) indicated that


75% of the fox squirrels fall diet in his Missouri study consisted of mast. How-ever, studies in Michigan (Fouch 1962), Iowa (Kline 1964), Illinois (McClellan 1985), and Indiana (Nupp and Swihart 2000) indicate no difference in harvest, survival, or fecundity in relation to mast crops in agricultural areas. Fecundity is reduced in many species during times of insufficient food sup-ply. Nixon et al. (1975) indicated summer breeding was drastically reduced during periods of low mast crops. In our study, only two juvenile squirrels were harvested from the non-mast site compared to 8 harvested from the mast site. Juvenile to adult ratios in this study were much higher than those in Iowa (Kline 1964, McCloskey 1968), Ohio (Nixon et al. 1974), and Illinois (Har-nishfeger 1978). Ratios of juvenile to adults at both study sites were similar. We believe this result was due to emigration of juvenile males into the non-mast forest. Studies have indicated that dispersal primarily involves juvenile males (Allen 1943, Baumgartner 1943, Nixon et al. 1974, Delong 1986). The ratio of juvenile males to juvenile females in the non-mast forest was higher than documented in other studies (Fouch 1962, Kline 1964, Nixon 1965, McClos-key 1968, Nixon et al. 1975). Because of this result and the lower nutritional requirement for males compared to females, we believe that non-mast forests act as a sink for dispersing juvenile male squirrels. Vulnerability to harvest may have led to bias in this study. Nixon et al. (1974) indicated juvenile males were more vulnerable to harvest than juvenile females. A study by Reynolds and Silvy (1983) was inconclusive on vulnerability rates. Another possible bias was hunting pressure. The mast forest was privately owned and we were the only individuals harvesting squirrels on the property. The non-mast forest was a public hunting area. We did not encounter other hunters while at the site but were only there for 57 hours during the study. We also did not determine the amount of the mast crop during the study.

MANAGEMENT IMPLICATIONS

Mast crops are a valuable nutritional resource to many species. This study demonstrated the importance of mast crops to fox squirrel populations. Mast crops should be an important factor when managing forest habitats for wildlife. Various mast species should be managed to ensure mast availability to maintain fox squirrel populations. This study indicated a need for mast crops in agricul-tural areas.

ACKNOWLEDGMENTS

We thank S. L. Reighard, Jr. and J. Knutson Jr. for their assistance with harvest and data collection. We would also like to thank K. Monteith and Q. Phelps for helpful comments on manuscript drafts.


LITERATURE CITED

Allen, D. L. 1943. Michigan fox squirrel management. Michigan Department of Conservation, Game Division. Publication 100. Lansing, Michigan, USA.

Allen, J. M. 1948. Mast production and squirrel crops. 10th Midwest Wildlife Conference, Indiana Division of Fish and Wildlife, Indianapolis, Indiana.

Baumgartner, L. L. 1943. Fox squirrels in Ohio. Journal of Wildlife Manage-ment 7:193-202.

DeLong, K. T. 1986. The effects of sex and age on fall dispersal in fox squirrels (Sciurus niger) in Iowa woodlots. Iowa Academy of Science Proceedings 93:1.

Dimmick, R. D. and M. R. Pelton. 1996. Criteria of sex and age. Pages 205-208 in T. A. Bookhout, editor. Research and management techniques for wild-life and habitats. Fifth edition, The Wildlife Society. Bethesda, Maryland, USA.

Dubock, A. C. 1979. Methods of age determination in gray squirrels (Sciurus carolinensis) in Britain. Journal of Zoology 188:27-40.

Fouch, W. R. 1962. Mast crops and fox squirrel populations at the Rose Lake Experimental Station. Michigan Academy Science 47:211-217.

Goodrum, P. D. 1972. Adult fox squirrel weights in eastern Texas. Journal of Wildlife Management 36:159-161.

Godin, A. J. 1960. A compilation of diagnostic characteristics used in aging and sexing gamebirds and mammals. Thesis, University of Massachusetts, Amherst, Massachusetts, USA.

Harnishfeger, R. L. 1977. Reproductive biology of the fox squirrel in Thompson Woods. Thesis, Southern Illinois University, Carbondale, Illinois, USA.

Harnishfeger, R. L., J. L. Roseberry, and W. D. Klimstra. 1978. Reproductive levels in unexploited woodlot fox squirrels. Illinois State Academy of Science Transactions 71:342-355.

Harty, F. M. and H. J. Stains. 1976. Fox squirrel utilization of catkins and acorns of the black walnut group. Transactions Illinois State Academy of Science 69:188-191.

Iowa Department of Natural Resources. 2003. Iowa hunting, fishing, and trap-ping regulations. Des Moines, Iowa, USA.

Jones, R. 1997. Soil survey of Emmet County. Iowa. Natural Resources Conservation Service. U.S. Department of Agriculture, Estherville, Iowa, USA.

Kirkpatrick, C. M. 1955. The testis of fox squirrel in relation to age and seasons. American Journal of Anatomy 97:229-255.

Kline, P. D. 1964. Iowa squirrels: hunting statistics, sex and age ratios, and the influence of mast and agriculture. Iowa Academy of Science 71:216-227.

Korschgen, L. J. 1981. Foods of fox and gray squirrels in Missouri. Journal of Wildlife Management 45:260-266.

McClellan, S. 1985. Survival rates in squirrels. Illinois Natural History Survey Reports 243:3-4.

McCloskey, R. J. 1968. Chronology of reproduction of the fox squirrel in Iowa. Thesis, Iowa State University, Ames, Iowa, USA.


McCloskey, R. J. 1977. Accuracy of criteria used to determine age of fox squir-rels. Iowa Academy of Science Proceedings 84:32-34.

Natural Resources Conservation Service. 2001. Farm and home plat and direc-tory. U.S. Department of Agriculture, Belmond, Iowa, USA.

Nixon, C. M., R. W. Donohoe, T. Nash. 1974. Overharvest of fox squirrels from two woodlots in western Ohio. Journal of Wildlife Management 38:67-80.

Nixon, L. P. Hanson, and S. Havera. 1991. Growth patterns of fox squirrels in east-central Illinois. American Midland Naturalist 125:168-172.

Nixon, M. W. McClain, and R. W. Donohoe. 1975. Effects of hunting and mast crops on a squirrel population. Journal of Wildlife Management 39:1-25.

Nupp, T. E. and R. K. Swihart. 2000. Landscape level correlates of small-mam-mal assemblages in forest fragments of farmland. Journal of Mammalogy 81:512-526.

Petrides, G. A. 1951. Notes on age determination in squirrels. Journal of Mam-malogy 32:111-112.

Reynolds, J. R. and N. J. Silvy. 1983. Differential vulnerability of fox and gray squirrels age classes to hunting. Southeastern Association of Fish and Wild-life Agencies Proceedings 37:23-27.

Schmidt, K. A. and J. S. Brown. 1996. Patch assessment in fox squirrels: the role of resource density, patch size, and patch boundaries. American Naturalist 147:360-380.

Schroeder, R. L. and L. D. Vangilder. 1997. Tests of wildlife models to evaluate oakmast production. Wildlife Society Bulletin 25:639-646.

Short, H. L. and W. B. Duke. 1971. Seasonal food consumption and body weights of captive tree squirrels. Journal of Wildlife Management 35:435-439.

SYSTAT. 2002. Systat Software, Inc. Richmond, California, USA.Thoma, B. L. and W. H. Marshall. 1960. Squirrel weights and populations in a

Minnesota woodlot. Journal of Mammalogy 41:272-273.Thompson, D. R. 1958. Field techniques for sexing and aging game animals:

Special Report No. 1. Wisconsin Department of Natural Resources. 45p.United States Geological Survey. 2002. Topography Maps. U.S. Department of

the Interior, Washington D.C., USA.Zelley, A.R. 1971. The sounds of the fox squirrel, Sciurus niger rufiventer. Journal

of Mammalogy 52:597-603.


ERGODIC THEORY AND ITS APPLICATION TOTHE ANALYSIS OF NUCLEIC ACID SEQUENCES

O. Michael MelkoDepartment of Mathematics

Northern State UniversityAberdeen, SD 57401

ABSTRACT

Basic facts about measure-preserving transformations and the ergodic theo-rem are reviewed with an emphasis on discrete spaces. These ideas are then used to derive an estimate for the number of approximate matches between a fixed word and a long sequence of randomly generated characters over an alphabet. This is followed by a discussion of generalizations, and potential applications to molecular probe design.

Keywords

Ergodic theory, statistical mechanics, information theory, biosequences, mo-lecular probes, nucleic acid sequences, approximate match counts

INTRODUCTION

The goal of this paper is to outline an approach for the statistical analysis of words in biosequences based on ideas from statistical mechanics and informa-tion theory. Although the ideas discussed here may be applicable to other biose-quences, such as proteins, we restrict are attention nucleic acids. The structure of the paper is as follows: the next two sections dispense with the mathematical preliminaries, after that, we develop the application that is the main focus of this paper. Finally, we discuss some unanswered questions, and other avenues of pos-sible research.

THE INDIVIDUAL (BIRKHOFF) ERGODIC THEOREM

In this section, we introduce the mathematical formalism used in this paper. In the interest of brevity, we merely summarize key concepts and results. Details, including proofs of the stated theorems, may be found in (Billingsley 1965), or (Reed et al. 1980). Note also that the latter reference contains an introduction to the concepts of measure theory used here. Let Ω denote a probability space with σ-field F and probability measure ρ : F → [0,1] set X in Ω is measurable if X ∈ F. This is a mathematical technical-ity that insures the measure, or probability of occurrence, ρ(X) of X makes sense. A function f between two measure spaces is measurable if the pre-image f -1(Y)


is measurable whenever Y is measurable. (Another technicality insuring the exis-tence of the integral of a function.) We denote the space of real-valued Lebesgue integrable functions on Ω by L1(Ω,ρ). The expectation of a measurable function f ∈ L1(Ω,ρ) will be denoted by E[f ]. Recall that, by definition, E[f ] is the aver-age value of f over Ω with respect to ρ, and is given by the integral

E[f ] := ⌠Ω

f d ρ

⌡

Note also that under certain circumstances (such as when Ω is a finite set) the integral reduces to a sum. We say that a map τ : Ω → Ω (which we always assume to be invertible) is ρ-measure preserving if ρ(τ(X)) = ρ(X) for every X ∈ C. A ρ-measure preserving transformation is said to be ergodic if the only invariant measurable sets in F are trivial in the sense that they have measure 0 or 1. That is, if X ∈ F, and if τ(X) = X, then ρ(X) = 0 or 1. We think of Ω as the set of possible states of some system, and a measurable set X ∈ F as an ensemble of allowable states, or an event. Further, we regard an integrable function f on Ω as an observable. The set X ⊆ Ω is proper if it represents an ensemble of states with probability of occur-rence 0 < ρ(X) < 1. That is, neither the occurrence nor the nonoccurrence of X is certain. Thus, a ρ-measure preserving transformation τ is ergodic if there are no proper measurable subsets of Ω which are invariant under the action of τ. Note that two measurable sets, X and Y, may have the property that X ≠ Y and yet ρ((X – Y) ∪ (Y – X)) = 0. We then say that X and Y differ on a set of mea-sure zero. Two functions are said to agree almost everywhere (a.e.) if the subset in Ω on which they differ has measure zero. With these definitions, the individual ergodic theorem can be stated as follows:

Individual Ergodic Theorem: Let τ : Ω → Ω be a ρ-measure preserving transformation on Ω. Then for any f ∈ L1 (Ω, ρ), there exists some function f # ∈ L1 (Ω, ρ) such that

lim — ∑ f (τr ω) = f #(ω) a.e.

Furthermore, if τ is ergodic, we have

f #(ω) = E[f ] a.e.

Let’s consider briefly what this theorem means. We may think of the set of iterates {τr}r=0 of τ on Ω as a discrete version of the time evolution of a dynamical system. The above theorem then states that the “time average” of the observable f (initially evaluated at some “generic” state ω) is tends to its average value over the state space Ω. This is essentially the 0th law of thermodynamics, which asserts that any system approaches an equilibrium state. There is a stronger property than ergodicity called mixing. Its usefulness lies in the fact that it’s often easier to show a transformation is mixing than to show that it’s ergodic.

1nn→∞ r=0

n−1

∞


Definition: An invertible ρ-measure preserving transformation τ : Ω → Ω is said to be mixing if, for any pair of measurable sets X, Y ∈ F , we have

lim ρ(X ∩ τn(Y)) = ρ(X) ρ(Y)

A word about the meaning of this definition is perhaps also in order. It essen-tially says that, as n grows larger and larger, the set τn (Y)spreads out uniformly over Ω. Hence, as our system evolves (n → ∞), any proper ensemble of states Y will visit a given proper ensemble X infinitely often. (For large enough n, the intersection of sets on the left-hand side of the above equation must always have positive measure, and hence is nonempty). Furthermore, in the limit, X and τn(Y) are not correlated as events.

Theorem: If τ : Ω → Ω is mixing, then it is also ergodic.

Note that there is another form of the ergodic theorem, called the mean (von Neumann) ergodic theorem. Von Neumann’s version describes the behavior, under averaging, of the iterates a unitary operator U on the Hilbert space of square-integrable functions L2 (Ω, ρ). These two ergodic theorems are related through a program known as “Koopmanism”, which establishes a connection by defining U via U(f )(ω) := f (τ(ω)), where τ is assumed to be ρ-measure preserv-ing. Koopman’s program is to describe interesting properties of τ in terms of the spectral properties of U (Reed et al. 1980). The latter approach is beyond the scope of this paper, but see the discussion below.

BERNOULLI SEQUENCES AND THE SHIFT OPERATOR

Now we consider a particular example. Our approach is adapted from (Bill-ingsley 1965). Let A = {a1, ⋅ ⋅ ⋅ , ad} denote a finite set with cardinality d = | A |, and let the corresponding σ-field F = 2A be the collection of all subsets of A. We interpret A as the set of possible states of a random variable ω, and we think of ω as an “experiment”, similar to a coin toss, whose random outcome is one of the pos-sible states specified by A. The case of particular interest to us is when A is the alphabet {A, C, G, T} of nucleotides appearing in a DNA sequence. Suppose we are given a probability measure ρ:A → [0,1] on A. Since A has finite cardinality, we may write ρj := ρ(aj) for the probability of occurrence of the state aj ∈ A. The stipulation that ρ is a probability measure implies, of course, that

ρ1 + ⋅ ⋅ ⋅ + ρd = 1

We want to calculate the expectation that the outcome of ω falls within a subset X ⊂ A. To that end, we define the indicator function IX : A → {0,1} of a set X by the rule that IX (a) = 1 if a ∈ X, and IX (a) = 0 otherwise. Then expectation of IX

is given byE[IX] = 1 ⋅ ρ(X) + 0 ⋅ ρ(X c) = ρ(X)

n→∞


Where X c denotes the complement of X in A. We have just given a formal description of a single Bernoulli trial. Next, we wish to discuss doubly infinite sequences of such trials. A Bernoulli sequence with values in A is of the form ω = (⋅⋅⋅, ω-1, ω0, ω1, ⋅⋅⋅) where ωj : Z → A is an experiment of the type just described for each integer j. The state space Ω is taken to be the set of all such sequences. Our next task is to define an appropriate σ-field F on Ω, and a probability measure ρ on F. Let E ⊆ Am, where Am denotes the m-fold Cartesian product of A. For integers m, n with m ≥ 1, define a cylinder of length m starting at n in Ω to be a set of the form

Cn,m (E) := {ω ∈ Ω | (ωn, ⋅⋅⋅, ωn+m-1) ∈ E}

The desired σ-field F is then generated by means of (possibly countably infinite) unions and complementation of cylinders in Ω. Note that ∅ and Ω are members of F. A probability measure ρ:A → [0,1] extends naturally to the Cartesian product Am, which we also denote by ρ. This is accomplished by setting ρ(X1 × ⋅⋅⋅ × Xm) := ρ(X1) ⋅⋅⋅ ρ(Xm) on “cuboids” of the form X1 × ⋅⋅⋅ × Xm (where X j ⊆ A for 1 ≤ j ≤ m), and extending to Am via set-theoretic operations. The measure of a cylinder Cn,m(E) is then defined to be ρ(Cn,m(E)) := ρ(E). Standard measure-theoretic arguments (Billingsley 1965) allow us to extend ρ to all of F. We have defined our underlying space of interest, and now we wish to de-scribe the transformation that gives us our “dynamics”. The Bernoulli shift opera-tor is given by

(τ ω)n = ωn+1

This operator takes a given Bernoulli sequence ω and creates a new one by shift-ing each entry one position to the left. Note that the sequences in Ω must be doubly infinite in order for τ and its inverse to be well defined. Observe, also, that τ is ρ-measure preserving. This fact is obvious on cylinders.

Theorem: The Bernoulli shift operator τ is ρ-measure preserving on the space of doubly infinite Bernoulli sequences . Furthermore, it is mixing, and hence ergodic.

APPROXIMATE MATCH COUNTS IN NUCLEIC ACID SEQUENCES

There is a large degree of randomness to the nucleotide sequence of a ge-nome: patterns exist, but the signal to noise ratio is generally weak. The statistics of such sequences has been the subject of study ever since data first became avail-able, and various approaches to modeling the data have been considered. The best results are obtained by means of hidden Markov models, which postulate that the outcome of a given experiment (i.e., the nucleotide appearing at a given locus of the genome) depends on the outcome of one or more of its predeces-sors. One application of this approach is the detection of statistically significant regional variations in nucleotide content (e.g., so-called “CpG islands”) within a genome. Details of this approach may be found in (Durbin et al. 1998).


The other main approach is to model DNA with Bernoulli trials, in which adjacent nucleotides are assumed to occur randomly and independently of each other. This model remains useful for many purposes. In (Melko et al. 2004), it was used to derive a formula for the expectation of approximate match counts between a molecular probe q and a long nucleotide sequence γ. The sequence could be a fragment of a genome, or an entire genome, for example. A molecular probe is a short nucleotide sequence (sometimes referred to as a “word”), which is chosen for its specificity to a particular sequence. Suppose that q has length m (i.e., consists of m nucleotides). We are interested in finding an estimate of the number of locations along γ where q and γ have k or fewer mismatches, where 0 ≤ k ≤ m. This number is an example of what we mean by an approximate match count; we refer to it as the k-mismatch count between q and γ, and denote it by θq(γ). We would like to estimate the value of θq(γ) from structural information about q and γ (such as GC-content). In order for q to be specific to γ, θq(γ) should be relatively large, while at the same time θq(γ∗) should be small for any sequence γ∗ chosen from some specified collection of sequences excluding γ. Implicit in this discussion is the assumption that the number of mismatches between q and a specific oligomer of length m (or m-mer) of γ correlates well with the stability of the duplex that they form. Assessing the validity of this assumption would require a careful study of the chemistry of DNA hybridization. We now take A to be the alphabet of nucleotides {A, C, G, T}, and view a sequence γ = (γ0, ⋅⋅⋅, γn-1) of length n, where γj ∈ A for 0 ≤ j ≤ n-1, as the partial outcome of an experiment ω = (⋅⋅⋅, ω-1, ω0, ω1, ⋅⋅⋅). Thus, each γj, for 0 ≤ j ≤ n-1, is the outcome of the random variable ωj. If we let ω(i, ⋅⋅⋅, j) denote the finite subsequence of ω starting at i and ending at j, then γ is a possible outcome of ω(0, ⋅⋅⋅, n-1). We estimate the probability of occurrence of a nucleotide in each ωj by calculating the relative frequency of occurrence of each nucleotide in γ. Here, it is convenient to index these probabilities by the elements of A, so we write {ρA, ρC, ρG, ρT} for the probabilities, which satisfy ρA + ρC + ρG + ρT = 1. Note that these probabilities depend on the choice of genome γ, and that θq(ω(0, ⋅⋅⋅, n-1)) is a random variable with possible outcome θq(γ). A k-mismatch between q and an m-mer in ω(0, ⋅⋅⋅, n-1) is an in-stance in which τr(ω) falls inside a specified cylinder in F, for some r ≥ 0. To specify the cylinder, first define the ball of radius k about q in Am to be Bk(q) := {p ∈ Am | d(q, p) ≤ k}, where d(q, p) denotes the number of mismatches between q and p. (The quantity d(q, p) is essentially the Hamming distance of information theory.) The desired cylinder is simply C0,m (Bk(q)). For notational convenience, we will assume that q and k are fixed, and set C = C0,m (Bk(q)). Our main result is then as follows:

Theorem: For values of n that are large in comparison to the length m of q, we have the relation

θq(ω(0, ⋅⋅⋅, n-1)) ≈ (n-m+1)ρ(Bk(q))

k k

k

k

k k

k


Proof: First, observe that

θq(ω(0, ⋅⋅⋅, n-1)) = ∑ IC(τrω)

The term n-m+1 occurs here because it is the number of contiguous m-mers contained in a sequence of length n. The individual ergodic theorem states that

lim ——— ∑ IC(τrω) = E[IC] a.e.

where τ is the Bernoulli shift. Hence, for large n, we have

——— ∑ IC(τrω) ≈ E[IC] a.e.

This, together with the fact that E[IC] = ρ(C) = ρ(Bk(q)), implies

∑ IC(τrω) ≈ (n-m+1) ρ(Bk(q)) a.e.

from which the result follows.

DISCUSSION

We begin by deriving an exact formula for the expectation of k-mismatch counts.

Proposition: The expectation of the random variable θq(ω(0, ⋅⋅⋅, n-1)) satisfies

E[θq(ω(0, ⋅⋅⋅, n-1))] = (n-m+1) ρ(Bk(q))

Proof: Take the expectation of both sides of the first equation in the proof of the previous theorem, and use the linearity of expectation to get

E[θq(ω(0, ⋅⋅⋅, n-1))] = ∑ E[IC oτr]

The result now follows from the fact that E[IC oτr] = ρ(Bk(q)) for all integers r (since τ is measuring preserving).

r=0

n−m+1

1n→∞ r=0n-m+1

n−m+1

1r=0n-m+1

n−m+1

r=0

n−m+1

r=0

n−m+1

k

k

k

k


It is interesting to consider the meaning of this proposition in light of the previous theorem. It gives an exact formula for the expectation of θq(ω(0, ⋅⋅⋅, n-1)). Hence, if we do N experiments SN :={ωa ∈ Ω|1≤a≤N}, the mean value of the set of corresponding k-mismatch counts ΓN := {θk

q(ωa(0, Λ, n-1)) | 1 ≤ a ≤ N} will tend to (n-m+1) ρ(Bk(q)) as N → ∞. This tells us nothing about the variation in the set of numbers ΓN. Thus, in order to estimate the possible deviation of θq(ω(0, ⋅⋅⋅, n-1)) from its expectation, we must calculate its variance. A formula for the variance of θq(ω(0, ⋅⋅⋅, n-1)) is derived in (Régnier et al. 1998), but it is computationally intractable for large values of n. On the other hand, the theorem of the last section gives us an estimate for the k-mismatch count θq(ω(0, ⋅⋅⋅, n-1)), where ω ∈ Ω is a single experiment. Its accuracy depends only on the length of the subsequence ω(0, ⋅⋅⋅, n-1) in ω. The problem remains, however, to calculate a bound for the error in this estimate. A careful examination of one or more of the known proofs of the ergodic theorem may provide insight on how to proceed. Note, also, that the calculation of ρ(Bk(q)) is not entirely trivial. This was done in (Melko et al. 2004) for strand-symmetric Bernoulli sequences, in which the underlying probability measure on A is assumed to satisfy ρA = ρT and ρC = ρG. The resulting closed-form formula derived there was referred to as the “perturbed Binomial distribution”. Strand symmetry is observed to be approxi-mately true in the nucleotide sequences of many genomes. This symmetry is also known to biochemists as Chargaff’s second parity rule after its discoverer, Erwin Chargaff. Deviation from strand-symmetry in DNA sequences is given by two structural parameters, referred to as AT-skew and GC-skew. Either a closed-form expression or an algorithm should be given for calculating ρ(Bk(q)) when the probability measure on A does not satisfy strand-symmetry: i.e., AT-skew and GC-skew should be taken into account. The distribution of nucleotides in actual DNA sequences is not homoge-neous (with respect to location in the genome). Examples of this include the presence of CpG-islands already mentioned, and the presence of “isochores”. Thus it would be worthwhile, if possible, to extend the main result to the case where the probability distribution for each entry ωj in a Bernoulli sequence Ω varies with the index j. One problem here is that the Bernoulli shift is no longer measure preserving. Thus, one must produce an alternative measure preserving transformation. This simplest possibility along these lines is to consider a mea-sure on Ω which is periodic with respect the index j. Then a power of the Ber-noulli shift becomes measure preserving. A related problem would be to extend our result to Markov models. A “Markov shift” on Ω is discussed in (Billingsley 1965). It may be also prove interesting to reformulate the main result in terms of von Neumann’s ergodic theorem. Indeed, one version of the proof of the indi-vidual ergodic theorem in (Billingsley 1965) makes use of the unitary operator alluded to previously. Finally, we close with a phenomenological discussion, and some speculation. The random processes considered here are rather artificial in the sense that they don’t reflect the process by which DNA actually evolves. This observation also applies to the hidden Markov models discussed in (Durbin et al. 1998). The abil-

k

k

k

k


ity of such models to detect CpG-islands reflects the fact that they capture some information about biochemical processes in DNA. However, the Markov process is assumed to progress along the length of the DNA sequence (from left to right, say) as if the location of a nucleotide in the sequence were a time parameter. This is clearly not how DNA evolves. It would be interesting to devise a more realistic model reflecting the nature of molecular evolution. The genome of an organism is, effectively, an information repository for proteins that carry out the functions of life. Mutations occur at the level of DNA in the form of insertion, deletion, and substitution of individual nucleotides (as well as recombination of potentially large segments). However, the process of selection is carried out at the level of proteins. A mutation in the DNA sequence of an organism can be ineffectual, or cause a change in a protein that either gives the mutant some advantage, or causes a fatal breakdown in some aspect of protein function. Figure 1 summarizes this relationship between DNA and the proteins it encodes. Note that the idea of insertion, deletion, and sub-stitution of nucleotides has been used as a basis for scoring alignments of DNA sequences. An account of this can be found in (Waterman 1995). The recent paper (Karev et al. 2002) introduces a birth-death-innovation (BDI) model for the evolution of protein domains, which is based on the clas-sical theory of branching processes (Feller 1968). The core of the BDI model is a “master equation”, which is a nonlinear system of ordinary differential equa-tions. Asymptotically, this model is shown to produce a power-law distribution of protein domains, which is in agreement with experimental observation. This suggests developing a toy 1+1 lattice-type model of molecular evolution (similar to an Ising model of ferromagnetism). Here, our state space would be the space Ω of doubly infinite sequences over a finite alphabet, and we assume that the time variable is discrete. One way this model would differ from other lattice models is that it would allow the insertion/deletion (or creation/annihilation) of elements in a sequence. The BDI model of (Karev et al. 2002) might provide some insight on how to accomplish this. In particular, we would like to know if there are exactly solvable models of this type.

Figure 1: Relation of DNA mutation to birth-death-innovation models of protein evolution.


LITERATURE CITED

Billingsley, P. 1965. Ergodic theory and information. John Wiley & Sons, New York. 193 pp.

Durbin, R., S. Eddy, A. Krogh, and G. Mitchison. Biological sequence analysis. Cambridge University Press. 1998. 368 pp.

Feller, W. 1968. An introduction to probability theory and its application, Vol. I (3rd edition). Wiley Text Books, New York. 528 pp.

Karev, G.P., Y.I. Wolf, A.Y. Rzhetzky, F. S. Berezovskaya, and E.V. Koonin. 2002. Birth and death of protein domains: a simple model of evolution explains power law behavior. BMC Evolutionary Biology 2:18.

Melko, O. M., and A. R. Mushegian. 2004. Distribution of words with a pre-defined range of mismatches to a DNA probe in bacterial genomes. Bioin-formatics. 20: 67-74.

Reed, M., and B. Simon. 1980. Methods of modern mathematical physics I: functional analysis. Academic Press, San Diego. 400 pp.

Régnier, M. and W. Szpankowski. On the approximate pattern occurrences in a text. Compression and complexity of sequences 1997 (proceedings). IEEE Computer Society 1998: 253-264.

Waterman, M.S. 1995. Introduction to computational biology. Chapman Hall. 448 pp.


ACCURACY OF HOME SOIL TESTKITS ON SOUTH DAKOTA SOILS

Rhoda BurrowsDept. of Horticulture, Forestry, Landscape & Parks


ABSTRACT

A variety of soil test kits are available for home use, but the accuracy of those “quick” tests has not been independently verified. We tested four kit types on five South Dakota garden and field soils, and compared the kit results to soil analyses by the South Dakota State University (SDSU) Soil Testing Laboratory. To emulate use of the kits by homeowners rather than trained scientists, Master Gardener and student volunteers were given the kits and soils and were instruct-ed only to follow each kit’s manual. In general, pH results were similar to SDSU soil test lab results, although one kit overestimated that of the lower pH soils by 1.0 to 2.0 units. The kits tended to underestimate nitrogen, phosphorus, and to a lesser extent, potassium levels. Each kit contained fertilizer recommendations based upon the test results and sometimes upon the crop grown, and kit rec-ommendations were generally higher than SDSU recommendations. Fertilizer applications based on the kit results were mostly lower than those based solely on the crop grown without any consideration of soil test results. Thus, use of the kits by homeowners could result in decreased incidence of excessive fertilizer applications. Although the kits may be useful for home gardeners who wish to determine if their soils’ N, P, and K levels are low or high, they should not be relied on for commercial production where soil nutrient levels should be more precisely managed.

Keywords

Soil tests, nutrient determination

INTRODUCTION Home gardeners are strongly encouraged to have their soil tested to de-termine appropriate level of supplemental fertilization. In many cases, South Dakota soils already have an excess of potassium and phosphorous. If the hom-eowner has applied manure or other fertilizers in the recent past, even nitrogen may be adequate without further application. In the absence of a soil test, gar-deners are often encouraged to apply an amount of fertilizer based on the total average consumption of the crop in question. Application of these materials when adequate or surplus levels are already present can lead to contaminated surface or groundwater, excess soil salts, and other problems.


A variety of soil fertility test kits are available for home use, offering conve-nience, nearly instant results, and reduced cost compared to testing by university or independent soil testing laboratories. However, the results of these “quick” tests are more qualitative (i.e., low to high) than quantitative, and their accuracy has not been independently verified. Our objective was to determine the useful-ness of home soil test kits in determining garden soil pH, nitrogen, phosphorus and potassium content, and resulting fertilizer recommendation by testing a variety of South Dakota soils and comparing the results with that of the South Dakota State University Soil Test Laboratory.

METHODS

Soil Test Kits

• AccuGrow Soil Test Strips, Hach Company, P.O. Box 4659, Elkhart, IN 46514-0659 U.S.A.

• No-Wait Soil Test Kit, Farnam Companies Inc., P.O. Box 34820, Phoenix, AZ, 85067 U.S.A

• LaMotte Garden Soil Test Kit, LaMotte Company, PO Box 329, Chester-town, MD 21620 - USA

• Environmental Concepts Soil pH-N-P-K Test Kit, LusterLeaf Products, Ft. Lauderdale, FL USA

The soil tests of all kits are based on colorimetric methods, using either the color reaction of a soil:reagent solution or that of small strips dipped into the soil:reagent solution. Users then compare their results against a printed color chart to determine results. Kits differed in the amount of soil used for each test—as little as 1 mm3 and up to approximately 10 mm3. In all cases, the soil amount was determined by volume rather than weight, for the convenience of the user. Use of distilled water was strongly recommended.

Test Soils

1. Medium textured; home garden with 10 years of annual leaf compost addi-tions, 7% organic matter

2. Medium textured; home garden with 15 years of chemical fertilization (no compost), 6.5% organic matter

3. Medium textured; Crop field with manure history, 5.5% organic matter4. Coarse textured (very sandy) pasture soil, no fertilization, 0.6% organic mat-

ter5. Medium textured; research farm (periodic cropping, some chemical fertiliza-

tion), 3.5% organic matter

To emulate use of the kits by homeowners rather than trained scientists, Master Gardener and student volunteers were provided the kits, deionized water, and the test soils, with no instruction except to follow the directions provided by


each kit. Each soil x kit combination was tested by a minimum of four testers. The Environmental concepts kit was dropped from testing when the first three volunteers found the kit nearly impossible to use. A sample of each soil was also submitted to the South Dakota State University (SDSU) Soil Testing Laboratory for comparison. For comparison purposes, N, P, and K results were expressed on a relative scale of 0.5 to 5, with 0.5=very low and 5=very high. According to kit recommendations, “low” to “medium” levels (i.e., a score of 1 to 3) generated a recommendation for fertilizer addition according to the product literature. For each soil, analysis of variance was used to compare differences in test kit results for pH, N, P, and K.

RESULTS In general, pH results of the test kits were similar to SDSU soil test lab results (Table 1). The exception was Accugrow, which tended to give higher pH values, especially in the lower pH sandy soil (Soil 4). No one test kit showed consistently greater or less variability for pH. Variability of N,P, and K results within a given test kit was greatest for Soils 2 and 4 (Table 1). As with pH, no one test kit showed consistently greater or less variability for these three nutrients across soils (Table 1). The soils tested all had relatively high amounts of potassium, and potassium results were generally the least variable. However, variability did not always coincide with high or low values. For example, there was no variability within test kit results for Soil 5, but AccuGrow and LaMotte found high levels of nitrogen while No-Wait found low levels resulting in a fertilizer application recommendation. Each kit contained fertilizer recommendations based upon the test results and sometimes upon the crop grown, and kit recommendations generally were higher than SDSU recommendations. NoWait underestimated nitrogen com-pared to the SDSU lab and other kits, resulting in recommendations for nitrogen fertilization (Table 2) for all five soils. Few differences between kits were noted for phosphorus values of the higher organic matter soils, but both NoWait and LaMotte resulted in phosphorous fertilizer recommendations beyond what was recommended by SDSU (Table 2). Accugrow results did not lead to recom-mendations of either phosphorous or potassium fertilization in any of the five soils, even in Soil 5, the only soil of the five for which the SDSU laboratory recommended supplemental phosphorous and potassium.

DISCUSSION

Our results indicate that home soil test kits are limited in precision, although they do reflect general trends of soil pH and nutrient levels. Our results are similar to results from somewhat more sophisticated colorimetric on-farm tests discussed by Allan and Killorn (1996) and Liebig et al (1996). Liebig et al. also found higher nitrate levels by the standard laboratory analysis compared to field tests, and attributed the difference to sample handling differences. However, all


Table 1. Mean soil pH and relative nutrient levels of soil test kit results: 0.5=Very Low, 1=Low, 2=Medium Low, 3= Medium, 4=High, 5=Very High

pH N P K

Soil 1

No-Wait 7.5 (6)* 0.5 (0)* 3.6 (19)* 4.3 (12)* AccuGrow 8.0 (0) 1.0 (0) 4.5 (0) 4.5 (0) LaMotte 7.6 (6) 1.6 (68) 4.0 (0) 4.8 (11) SDSU 7.8 3.0 5.0 5.0 Soil 2

No-Wait 7.3 (6) 0.6 (35) 3.4 (14) 3.7 (14) AccuGrow 7.7 (4) 1.4 (64) 3.5 (14) 3.3 (14) LaMotte 7.7 (4) 2.0 (87) 4.0 (0) 4.0 (25) SDSU 6.9 3.0 5.0 5.0Soil 3



No-Wait 7.5 (7) 1.0 (0) 1.0 (0) 3.9 (5) AccuGrow 8.0 (0) 4.0 (0) 3.0 (0) 4.0 (0) LaMotte 7.0 (1) 4.0 (0) 1.6 (56) 3.0 (54) SDSU 7.5 5.0 3.5 3.5

*Percent variability (coefficient of variation) within test kit and soil.

our subsamples were taken from the same air-dried and sieved soil samples, so the lower N determination by the “quick” colorimetric methods may indicate a need for recalibration of the “quick” test charts. Variability in individual nutrient test results can be due to several factors. The soils themselves, even though well-mixed and drawn from the same bag, may not be completely uniform. Subsamples used for the test kits are small, from approximately 1- to 10 mm3, depending on the kit, so it is quite possible the soil itself varied from sample to sample. Secondly, there may be differences in how


Table 2. Fertilizer recommendations based on test kit results (lb. actual element per 1000 ft.2) by soil.

N P K

Soil 1

No-Wait 3-5 0-0.25 0 AccuGrow 1 trt* 0 0 LaMotte 4 3 1.5 SDSU 1 0 0Soil 2

No-Wait 1-5 0-.25 0-1 AccuGrow 1-2 trts* 0 0 LaMotte 2 3 3 SDSU 1.2 0 0Soil 3

No-Wait 3-5 0 0 AccuGrow 1 trt* 0 0 LaMotte 2 3 1.5 SDSU 0 0 0Soil 4

No-Wait 1-5 0.25-1 0-1 AccuGrow 1-2 trts* 0 0 LaMotte 5 5 2.5 SDSU 3 0 0Soil 5

No-Wait 1-3 0.25-.75 0-0.5 AccuGrow 0 0 0 LaMotte 2 5 3-5 SDSU 0 0-.85 1Crop (no soil test):

Vegetables 1-3 1-3 1-3 *trt = "application per (fertilizer) package instructions"

various persons interpret the match between the color of the sample with reagent and the printed color swatches. Thirdly, there may be simple errors in measuring and/or timing mixing and settling of reagents, especially by lay people unused to laboratory procedures. Organic matter (OM) may affect the variability of nitrate availability in the soil over time, but did not appear to influence the variability


of the test kit results, as the low-OM soil (Soil 4) showed no less variability than those with relatively high OM (Soils 1 and 2). Potassium within-kit results were generally the least variable, perhaps be-cause the five soils tested all had medium-high to high amounts of potassium. However, comparison of P results within Table 1 reveal that similarly low or medium test values (eg., Soils 4 and 5) can be associated with high or low vari-ability, even by the same test kit. Thus, the variability observed within any one test kit or nutrient had no apparent connection to the level of nutrient in any particular soil. This rules out the possibility of recommending one test kit over another for a given soil type (eg., high or low OM, medium or coarse texture, high or low nutrient-containing). Fertilizer recommendations for any given soil can vary considerably, depend-ing on not only crop, but also on expected growing conditions and target yield, as well as less tangible considerations. The LaMotte kit recommends at least minimal application of N, P, and K for every soil, regardless of soil test values, and its recommended P and K applications were always quite high. In most South Dakota soils, this addition would be unnecessary and, in cases where the soils have high salts, even detrimental.

CONCLUSIONS

Fertilizer applications based on the No-Wait and AccuGrow kit results are lower than recommendations based solely on the crop grown without consideration of soil test results (Table 2). Thus, even though these kits generally recommended higher application levels than indicated by the SDSU laboratory, use of these kits by homeowners would still result in decreased incidence of excessive fertilizer applications. However, although the kits may be useful for home gardeners who wish to determine if their soil’s N, P, and K level is low or high, they should not be relied upon for commercial production where soil nutrient levels should be more precisely managed. The kits also lack the ability to determine soil texture, OM content, or soluble salt levels—all included in the standard SDSU “Garden Soil” analysis and useful in determining soil management by both homeowners and commercial growers.

ACKNOWLEDGEMENTS

Thanks to the many Master Gardener volunteers who donated their time to test soils with these kits, and to the South Dakota State University Soil Testing Laboratory for providing calibrated analysis of the soils.


LITERATURE CITED

Allen, D.L. and R. Killorn. 1996. Assessing soil nitrogen, phosphorus, and potassium for crop nutrition and environmental risk, pp. 187-201. In: Methods for assessing soil quality, J.W.Doran and A.J. Jones, eds. SSSA Spe-cial Publication 49.

Liebig, M.A., J.W. Doran, and J.C. Gardner. 1996. Evaluation of a field test kit for measuring selected soil quality indicators. Agron. J. 88:683-686.


A COMPARATIVE STUDY OF SEEDCHARACTERISTICS IN THE

CHENOPODIACEAE AND AMARANTHACEAE

Elke Kuegle and Mark GabelBiology Department

Black Hills State UniversitySpearfish, SD 57799

ABSTRACT

The Amaranthaceae and Chenopodiaceae are herbs and shrubs primarily known for their weedy nature. Both families belong to the order Caryophyllales which has characteristic seed shape and embryo position. Numerous character-istics have previously been used to compare the families with differing results. Until this time, no studies have conducted a comprehensive examination of the seed characteristics of the two families. Twenty-five characters were examined in 26 species traditionally placed in the Chenopodiaceae and in 34 species of the traditional Amaranthaceae. Our results indicate that seed characters do not separate the traditional taxonomic division, and thus support recent studies combining the two families.

Keywords

Amaranthaceae, Chenopodiaceae, seed, cluster analysis, principal compo-nent analysis

INTRODUCTION

The Caryophyllales as described by Cronquist (1981) and Cronquist and Thorne (1994) are nearly synonymous with the Centrospermae of previous authors (Lawrence, 1951; Rodman, 1990). Members of the core Caryophyllales are the Amaranthaceae and Chenopodiaceae which are thought to be closely related, but are traditionally separated. In a study using both phenetic and cladistic analyses of morphological, ana-tomical, palynological, chemical, and chromosomal data, Rodman et al. (1984) included the Chenopodiaceae and Amaranthaceae in the “cohort Amaranthares” in the suborder Chenopodiineae. Later (1990), Rodman presented revised and recoded data, but no visual analyses and indicated that the Amaranthaceae are nested within the Chenopodiaceae. A study of trichomes by Carolin (1983) indicated that the Amaranthaceae were contained within the Chenopodiaceae. Rodman (1994) refuted this arrangement indicating that the Amaranthaceae would have had to reduce ovule number to a single basal ovule and then reverse


this change to account for the multiovulate gynoecia of Celosia. A study of the Caryophyllales using chloroplast DNA, by Downie and Palmer (1994a) found one major clade of the Caryophyllales (based upon only five taxa) to be composed of the Chenopodiaceae and Amaranthaceae, and the other clade composed of all remaining families. They (1994b) stated that “both the Chenopodiaceae and Amaranthaceae appear to be monophyletic.” In 1997, Downie et al. conducted a study of chloroplast DNA (OFR2280), and suggested that the Amaranthaceae are polyphyletic with Celosia and Fro-elichia forming one clade and the Chenopodiaceae and Amaranthus another. Sarcobatus (traditionally Chenopodiaceae) grouped with a clade that included the Phytolacaceae and Nyctaginaceae. Behnke (1997), describing sieve elements and Cuénoud et al. (2002) using nuclear and plastid DNA have shown that Sar-cobataceae should be elevated to a family level. The latter authors also indicated a clade of the Chenopodiaceae and a separate “well-supported” clade composed of genera from the Amaranthaceae. The Chenopodiaceae have traditionally been described as a family of about 100 genera and about 1500 species most commonly found in drier and alkaline or saline habitats (Kuhn et al., 1993; Flora of North America Editorial Commit-tee, 2003). The Amaranthaceae have been delimited as containing about 65 gen-era and 900 species and are most common in warmer climates (Townsend, 1993; Flora of North America Editorial Committee, 2003). Both families contain numerous herbaceous species and a few shrubs or sub-shrubs and are primarily known for their weedy nature. Historically, the two groups have been considered as separate but related families (Cronquist 1981; Flora of North America Edito-rial Committee 2003). The Angiosperm Phylogeny Group (1998) included the Chenopodiaceae in the Amaranthaceae. Judd and Ferguson (1999) maintained the separation of the two families, but wrote “…the separation of the Chenopodiaceae and Ama-ranthaceae is more or less arbitrary and very probably results in a paraphyletic Chenopodiaceae …” As exemplified above, numerous characteristics have been used to compare the families with differing results. No studies have ever done a comprehensive examination of the seed characteristics of the two families. It is the purpose of this study to compare the seed characteristics of the traditionally defined Amaranthaceae and Chenopodiaceae to determine if they can help elucidate the relationship among the families, and if they support or refute combining the two families.


Seeds from 34 species of Amaranthaceae and 26 species of Chenopodiaceae were freed from the fruit wall and cleaned. A cursory examination of seed coat morphology was conducted using a dissecting microscope. The waxy layer ob-scuring seed coat cell details of some species, especially Amaranthus, was removed by submerging the seed in a solution of either 50% EtOH or 10% papain for a few minutes. Cross sections were prepared to view the internal anatomy of the


seeds by cryofracture or by microtome sectioning after embedding the seeds in paraffin or L.R. White resin. Seeds were mounted on a carbon stub for detailed examination with the scanning electron microscope (SEM). Images obtained by the SEM and dissecting microscope observations were then used to establish a matrix, with 25 distinct characters. To measure the phenetic relationships of the two families, a cluster analysis using a similarity matrix of the seed characters was used to construct a dendro-gram or phenogram. The same data was used to perform principal component analysis (PCA) by calculating correlation coefficients among the seed characters. Both analyses used NTSYS 2.11S (Rohlf, 2003).

RESULTS

Results of the measurement of 25 seed characters of 60 species (65 speci-mens) are shown in Table 1. These data were then used in cluster analyses and principal component analyses. The cluster analysis (Fig. 1) indicates that seed characters do not provide a clear distinction between the Amaranthaceae and Chenopodicaceae. Three major groups were delimited, one with primarily Ama-ranthus (traditional Amarathaceae), Iresine (Amaranthaceae), and two species of Suadea (Chenopodiaceae).

Figure 1. Cluster analysis (resulting in a dendrogram or phenogram) of 25 characters from 65 seed samples of the traditional Amaranthaceae and Chenopodiaceae. The figure was derived from a standardized similarity matrix by computing and cluster-ing distance coefficients. The first letter of each label refers to traditional placement (A = Amaranthaceae, C = Chenopodiaceae), while the second and third letters refer to the genus. Numbers refer to sample (Table 1). Axis aspect has not been preserved to allow easier viewing.


The second group was composed mostly of Atriplex (Chenopodiaceae), with Gomphrena (Amaranthaceae), one Celosia (Amaranthaceae), two Amaran-thus species and another Suadea (Chenopodiaceae). The third group included most of the Chenopodium (Chenopodicaceae), as well as Kochia (Chenopodi-aceae), Celosia (Amaranthaceae) and several species of less well-represented gen-era. The PCA very broadly separated the two families (Fig. 2), but there were numerous genera which were traditionally members of one family which were placed within a group of the other family. Examples of the “misplaced” gen-era include Suaeda (100) (Chenopodiaceae) Chenopodium 83 and Gomphrena (Amaranthaceae). The PCA supported the placement of the genera in the cluster analysis (Fig. 1). While the details of the PCA presented in Fig. 2 appear rather obscured by overlapping labels, it should be noted that the NTSYS software (Rohlf, 2003) allows the user to freely move the model in apparent three di-mensional space, thus allowing a visualization of each datum point. While the analyses of the seed characteristics broadly separate the two families, there are numerous outlying species that do not group as expected.

Figure 2. Principal Component Analysis of 25 characters from 67 seed samples of the traditional Amaranthaceae and Chenopodiaceae. The figure was derived from a stan-dardized correlation matrix and presents orthogonal coordinate axes such that points have variance in as few dimensions as possible. The first letter of each label refers to traditional placement (A = Amaranthaceae, C = Chenopodiaceae), while the second and third letters refer to the genus. Numbers refer to sample (Table 1).


DISCUSSION

It is obvious from Figs. 1 and 2 that seed characters do not clearly separate the Amaranthaceae and Chenopodiaceae, thus supporting the hypotheses of Downie et al. (1997) and the Angiosperm Phylogeny Group (1998). Frequently the taxa which were placed unexpectedly or “misplaced” are problematic genera discussed by other authors (Suaeda, Judd and Ferguson, 1999). Of the 76 characters employed by Rodman (1994) in a study of the Caryo-phyllales, only seven differentiated the Amaranthaceae from the Chenopodiaceae. The seven characters, presence of: embryo chlorophyll, arillate seeds, myricetin, 6-hydroxyflavenol, flavonol sulfate, 6-hydroxyflavones, and type of leaf wax were combined with our dataset. The two families unsurprisingly were separated. The validity of selecting only variant characters in such an analysis is questionable. It should also be noted that Rodman (1994) was analyzing family positions in an order, and did not include data for species or genera. It is notable that some species are exceptions to the generalizations about families. One example is the presence of an aril on the seed of Chamissoa altissima which is a member of the Amaranthaceae (Townsend, 1993). We did not include Sarcobatus (traditionally in the Chenopodiaceae) in our study. The genus has been elevated to family status (Sarcobataceae) by Behnke (1997) or may be closely related to the Nyctanginaceae (Cuénoud et al., 2002). It is important to remember that these are phenetic data, and we make no conclusions about the primitive or derived nature of the character states. We can conclude, however, that seed characters alone do not support the maintenance of two separate families. The seed characters studied, showed no clear distinction between the Chenopodiaceae and Amaranthaceae. This could be due to variabil-ity within these two groups or to the characteristics selected.

ACKNOWLEDGEMENTS

The authors wish to thank the curators of ISC, MO, and BHSC. The BHSU Scanning Electron Microscope Facility was used in this study. The Na-tional Geographic Society is thanked for a grant to MG. The Nelson Scholarship Committee (BHSU) and BRIN (NIH) are acknowledged for a grant to EK.

LITERATURE CITED

Angiosperm Phylogeny Group. 1998. An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85:531-553.

Behnke, H. D. 1997. Sarcobataceae – a new family of Caryophyllales. Taxon 46:495-507.

Carolin, R.C. 1983. The trichomes of the Chenopodiaceae and Amaranthaceae. Bot. Jahrb. Syst. 103:451-466.


Cronquist, A. 1981. An integrated system of classification of flowering plants. Columbia University Press, New York.

Cronquist, A. and R.F. Thorne. 1994. Nomenclatural and Taxonomic History. pp 5-25 in Behnke, N.D. and T.J. Mabry (eds.) Caryophyllales: Evolution and Systematics. Springer-Verlag, Berlin.

Cuénoud, P., V. Savolainen, L.W. Chatrou, M. Powell, R. Grayer and M.W. Chase. 2002. Molecular phylogenetics of Caryophyllales based on nuclear 18S rDNA and plastid rbcL, atpB and matK DNA sequences. American Journal of Botany 89:132-144.

Downie, S.R. and J.D. Palmer. 1994a. Phylogenetic relationships using restric-tion site variation of the chloroplast DNA inverted repeat. In Caryophyllales: Evolution and systematics N.D. Behnke and T.J Mabry (eds.) p. 223-233. Springer-Verlag, Berlin.

Downie, S.R. and J.D. Palmer. 1994b. A chloroplast DNA phylogeny of the Caryophyllales based on structural and inverted repeat restriction site varia-tion. Systematic Botany 19: 236-252.

Downie S.R., D.S. Katz-Downie and K.J. Cho. 1997. Relationships in the Caryophyllales as suggested by phylogenetic analysis of partial chloroplast DNA ORF2280 homolog sequences. American Journal of Botany 84: 253-273.

Flora of North America Editorial Committee. 2003. Flora of North America. Volume 4: Caryophyllales. Oxford, New York.

Judd, W.S. and T.K. Ferguson. 1999. The genera of Chenopodiaceae in the southeastern United States. Harvard Papers in Botany 4:365-416.

Kuhn, U., V. Bittrich, R. Carolin, H. Freitag, I.C. Hedge, P. Uotila and P.G. Wilson. 1993. Chenopodiaceae. In The families and genera of vascular plants, vol 2, Magnoliid, hamamelid and caryophyllid families, K. Ku-bitzki, J.G. Rohwer and V. Bitterich (eds.) p. 253-281. Springer-Verlag, Berlin.

Lawrence, G.H.M. 1951. Taxonomy of Vascular Plants. MacMillan, New York.

Rodman, J.E. 1990. Centrospermae revisited. Part I. Taxon 39:383-393. Rodman, J.E. 1994. Cladistic and phenetic studies. In Caryophyllales: Evolu-

tion and systematics N.D. Behnke and T.J Mabry (eds.) 279-301. Springer-Verlag, Berlin.

Rodman, J.E., M.K. Oliver, R.R. Nakamura, J.U. McClammer, Jr. and A.H. Bledsoe. 1984. A taxonomic analysis and revised classification of Centro-spermae. Syst. Bot. 9:297-323.

Rohlf, J. 2003. NTSYS 2.11S. Exeter Software, Setauket, NY.Townsend, C.C. 1993. Amaranthaceae. In The families and genera of vascular

plants, vol 2, Magnoliid, hamamelid and caryophyllid families. K. Kubitz-ki, J.G. Rohwer and V. Bitterich (eds.) 70-91. Springer-Verlag, Berlin.


Tabl

e 1.

Ta

ble

of s

peci

es o

f se

eds

stud

ied

and

char

acte

rs d

eter

min

ed.

The

lett

ers

prec

edin

g th

e sp

ecie

s na

mes

ref

er t

o tr

adit

iona

l pla

cem

ent

of t

he s

peci

es.

A =

Am

aran

thac

eae,

C =

Che

nopo

diac

eae.

Cha

ract

er n

umbe

r de

scri

ptio

ns a

re f

ound

at

the

end

of t

he t

able

. Num

ber

follo

win

g sp

ecie

s na

me

refe

rs t

o sa

mpl

e nu

mbe

r.

80 Proceedings of the South Dakota Academy of Science, Vol. 83 (2004)Ta

ble

1 co

ntin

ued.

Ta

ble

of s

peci

es o

f se

eds

stud

ied

and

char

acte

rs d

eter

min

ed.

The

let

ters

pre

cedi

ng t

he s

peci

es n

ames

ref

er t

o tr

adit

iona

l pl

acem

ent

of t

he s

peci

es.

A =

Am

aran

thac

eae,

C =

Che

nopo

diac

eae.

Cha

ract

er n

umbe

r de

scri

ptio

ns a

re f

ound

at

the

end

of t

he t

able

. Num

ber

follo

win

g sp

ecie

s na

me

refe

rs t

o sa

mpl

e nu

mbe

r.


Tabl

e 1

cont

inue

d.

Tabl

e of

spe

cies

of

seed

s st

udie

d an

d ch

arac

ters

det

erm

ined

. T

he l

ette

rs p

rece

ding

the

spe

cies

nam

es r

efer

to

trad

itio

nal

plac

emen

t of

the

spe

cies

. A

= A

mar

anth

acea

e, C

= C

heno

podi

acea

e. C

hara

cter

num

ber

desc

ript

ions

are

fou

nd a

t th

e en

d of

the

tab

le. N

umbe

r fo

llow

ing

spec

ies

nam

e re

fers

to

sam

ple

num

ber.

Tabl

e 1

char

acte

r de

scri

ptio

ns: 1

. See

d co

at g

lauc

ous,

2. M

argi

n ri

dged

, 3. S

urfa

ce t

uber

cula

te o

r pa

pilla

te, 4

. Sur

face

gra

nula

r, 5

. Sur

face

pit

-te

d, 6

. Sur

face

ver

y sm

ooth

wit

h a

glas

sy lu

ster

, 7. C

alyx

and

/or

funi

culu

s pe

rsis

tent

, 8.

Rad

ial s

tria

tion

vis

ible

at

mag

nifi

cati

on o

f up

to

65x,

9.

Fine

ret

icul

ate

patt

ern

visi

ble

at m

agni

fica

tion

of

up t

o 65

x, 1

0. C

ell s

hape

of

seed

coa

t at

mar

gin

elon

gate

d (1

) or

squa

mou

s (0

), 11

. Cel

l sha

pe

of s

eed

coat

cen

ter

elon

gate

d (1

) or

squa

mou

s (0

), 12

. Cel

l sha

pe o

f se

ed c

oat

at h

ilum

elo

ngat

ed (1

) or

squa

mou

s (0

), 13

. Sm

all s

ulcu

s at

hilu

m

indi

cati

ng f

olde

d em

bryo

, 14.

Not

ch a

t on

e po

int

on m

argi

n ev

iden

t, 1

5. If

not

ch p

rese

nt in

14,

mea

sure

men

t w

as o

btai

ned

for

leng

th o

f no

tch:

1/

10 m

m a

nd l

arge

r (1

), 1/

10 m

m a

nd s

mal

ler

(0),

16.

Tip

pro

trud

ing,

17.

H

ilum

in

a no

tch

form

ed b

y an

ext

ensi

on (

1),

wit

hout

ext

ensi

on (

0),

18. S

eed

scar

dep

ress

ed a

nd r

ound

(0)

, ova

l (1)

, or

tria

ngul

ar (

2) in

sha

pe, 1

9. S

urfa

ce in

dica

ting

the

inne

r em

bryo

, 20.

Sev

eral

pro

min

ent

vein

s an

d/or

rid

ges

runn

ing

leng

thw

ise

on s

eed

coat

, 21.

Sca

r at

cen

ter

of s

eed,

22.

Out

line

of s

eed

ellip

tica

l (1)

or

circ

ular

(0),

23. E

nd o

f se

ed o

bvi-

ousl

y tr

unca

ted,

24.

See

d co

at c

olor

bla

ck (

1), o

ther

(0)

, 25.

Len

gth/

wid

th r

atio

of

seed

.


NATURALLY OCCURRING ACID ROCKDRAINAGE AND IMPACTS TO THE UPPER RAPID

CREEK WATERSHED NEAR ROCHFORD, SD

Scott L. Miller and Arden D. DavisDepartment of Geology and Geological Engineering

Scott J. Kenner and A.J. SilvaDepartment of Civil and Environmental Engineering

South Dakota School of Mines and TechnologyRapid City, SD 57701

ABSTRACT

Naturally occurring acid rock drainage has historically and is presently nega-tively affecting reaches of North Fork Rapid Creek, North Fork Castle Creek, and Castle Creek in the Upper Rapid Creek Watershed of the Black Hills near Rochford, South Dakota. Samples were collected and assessments were made for these creeks during 2002 and 2003. The acid rock drainage has a pH of ap-proximately 2.5 to 3.5 and contains high concentrations of iron, aluminum, and sulfate. Uncontaminated surface water has a pH of approximately 7 to 8.5 and contains high concentrations of calcium, magnesium, and bicarbonate. When the acid drainage mixes with uncontaminated water of these creeks, natural buffering reactions occur, causing iron and aluminum hydroxides to precipitate. Surface water chemistry from upstream to downstream was little affected by the introduction of the acid rock drainage. However, the metal hydroxide precipi-tates significantly affect the streambed and the fisheries habitat. The metal hy-droxide precipitate coats the stream bottom and cements the sediments together, negatively affecting vegetation and macroinvertebrate habitat. Some precipitate is transported downstream, degrading the stream habitat for several hundred meters. Much of the plant and animal life in these areas is stressed, leaving some stream reaches devoid of all life and destroying the fisheries habitat. Natural chemical reactions cause the negative effects to attenuate, restoring the stream’s water quality and habitat within approximately one to two km downstream.

Keywords

Acid rock drainage, iron bog, fisheries habitat


INTRODUCTION

Naturally occurring acid rock drainage is actively discharging from iron bogs into North Fork Rapid Creek, North Fork Castle Creek, Castle Creek, and some unnamed tributaries. Iron bogs can develop where ground water discharges at the land surface in the form of springs. The discharge is acidic, contains high concentrations of iron, aluminum, and sulfate, and negatively affects the envi-ronment. The Upper Rapid Creek Watershed is in the north-central Black Hills of South Dakota, near the small community of Rochford, approximately 60 km northwest of Rapid City (Figure 1). This research was part of a larger project to

Figure 1. Location of North Fork Rapid Creek and generalized geologic map of the Black Hills, South Dakota (modified from Dewitt and others, 1989).


assess the Upper Rapid Creek Watershed for the South Dakota Department of Environment and Natural Resources, the South Dakota Department of Game, Fish, and Parks, and the Black Hills Flyfishers (in press).

HYDROGEOLOGIC SETTING

The geology of the Upper Rapid Creek Watershed consists of Precambrian metamorphic rocks overlain with younger Paleozoic sedimentary rocks to the north and west (Figure 1). Precambrian rock types include phyllite, schist, banded iron formation, chert, and quartzite (DeWitt and others, 1989). In the western Black Hills, the Paleozoic rocks are nearly flat lying and consist of, from older to younger, the Deadwood (sandstone), Winnipeg (shale), Whitewood (dolostone and limestone) formations, the Englewood and Pahasapa (Madison) limestone, and dolostone, and the Minnelusa Formation. North Fork Rapid, North Fork Castle, and Castle creeks are gaining streams, where flow increases downstream due to ground water discharging from bedrock and alluvium. The bulk of surface water in all three streams originates as spring flow from the Pahasapa limestone (Carter and others, 2002). Not surprisingly, uncontaminated water quality in these streams is similar to ground water quality of the Pahasapa (Madison) Limestone in the western Black Hills. Surface water in these three creeks and ground water from the Pahasapa limestone typically contain calcium, magnesium, and bicarbonate with trace levels of inorganics (e.g., chloride, fluoride, potassium, sodium and sulfate) and metals (e.g., alumi-num). The pH is typically 7.5 to 8.5. Water quality from Precambrian rocks varies greatly but generally contains iron, aluminum, and sometimes sulfate. Acid rock drainage may develop where pyrite is abundant in the Precambrian bedrock. Precambrian rock does not always contain abundant pyrite, therefore, not all ground water and spring dis-charge from the Precambrian rock is high in acid and metals. Acid rock drainage occurs both naturally and due to anthropogenic distur-bances, primarily mining. Most acid rock drainage is found at mine sites where it is commonly referred to as acid mine drainage. Acid rock drainage results from the interaction of sulfide minerals with oxygen and water. Iron-oxidizing bacteria often catalyze these reactions. Pyrite (FeS2) is the most abundant and widespread sulfide mineral and is the primary source for acid rock drainage. Pyrrhotite (FeS) is less common than pyrite and is a minor source material for acid rock drainage (Deer and others, 1980). In this region, the Precambrian metamorphosed sedimentary rocks most likely originated as marine shale under reducing conditions and contain abundant finely disseminated pyrite. This microscopic pyrite is more reactive than larger crystals as the finer crystals have a higher surface area to volume ratio (PDEP, 1998). Figure 2 shows a reach of Castle Creek strongly affected by acid drainage and Figure 3 shows both alumi-num and iron hydroxide precipitates at the confluence of an unnamed tributary with Rapid Creek. Iron bogs form where iron-rich, acidic ground water discharges to the land surface as springs or directly into the creek along the banks or within the chan-


Figure 2. Iron hydroxide precipitate coating stream substrate on reach of North Fork Rapid Creek. Note reach is devoid of most life in channel.

Figure 3. Confluence of acid rock drainage and uncontaminated stream. Note red-or-ange iron hydroxide and white aluminum hydroxide precipitates.


nel. This spring water comes from a subsurface reducing environment, is acidic (pH approximately 2.5 to 4), and contains high concentrations of iron, alumi-num, and sulfate. At the surface, the discharged water is exposed to oxygen from the atmosphere and mixes downstream with well-oxygenated, unpolluted surface water with pH approximately 7 to 8. The resulting mixture has a pH in the range of approximately 6.5 to 7.5 and is nearly saturated in oxygen. These con-ditions cause the reduced iron (i.e., Fe2+ or Fe (II)) and aluminum in solution to oxidize. The neutral water is over-saturated in iron and aluminum causing iron and aluminum hydroxides to precipitate, flocculate and accumulate at the spring and along the bottom of the channel. These chemical reactions occur quickly and precipitates can be observed immediately at springs and where acidic water mixes with surface water. The oxidized iron precipitates discolor the water light red-orange. Over longer periods of time (i.e., decades to millennia), the metal hydroxides accumulate, dewater and harden through diagenesis, cementing the stream sediments together. Iron-cemented conglomerate was encountered along streams, flood plains, and stream terraces indicating historic discharges of acid rock drainage in this area. Iron bog deposits contain as much as 50-percent iron. Because of their high iron content, many of the larger bogs in this area were mined for iron from the early 1900s up through the 1950s (USGS, 1975).

STUDY METHODS AND RESULTS

Field and analytical water quality data collected quarterly during 2002 and 2003 were used to characterize the surface and ground water quality in North Fork Rapid Creek, North Fork Castle Creek, and Castle Creek watersheds. Field data included ferrous (Fe2+) iron concentration, redox potential (Eh), pH, dis-solved oxygen, temperature, and specific conductance. Analytical data included general water chemistry consisting of major and minor ions and several metals. Analyses were performed by an EPA-certified laboratory. Selected water quality data collected from Upper Rapid Creek watersheds during 2002 and 2003 are summarized in Table 1. Concentrations of water quality parameters for surface water and spring and ground water vary by up to two orders of magnitude. Piper trilinear diagrams (Piper, 1944) and Stiff diagrams (Stiff, 1951) were used to classify water quality type and visually inspect and compare water quality based on major ions. Ground water and bog water were calcium-mag-nesium-sulfate rich water. Surface water and Pahasapa limestone water were calcium-magnesium-bicarbonate rich water. Using the U.S. Geological Survey geochemical model, pHREEQC (version 2.7) and visible and near infrared reflectance spectroscopy, precipitates appear to be iron hydroxide or yellow boy (Fe(OH)3), geothite (FeO•OH), limonite (FeO•OH•nH2O), hematite (Fe2O3), jarosite (KFe3(SO4)2(OH)6), and an aluminum hydroxide, gibbsite (Al(OH)3) or alunite (KAl3(SO4)2(OH)6).


Table 1: Summary of Selected Water Quality Data from Upper Rapid Creek

Stream Water Quality Spring Discharge and Ground Water Quality

Total Iron (mg/L) ND – 0.68 160 – 235Ferrous Iron, Fe2+ (mg/L) ND – ~2 ~3 – 7Ferric Iron, Fe3+ (mg/L) 0.11 – 2. 3 10 – 130Aluminum (mg/L) 0.02 – 0.28 11 – 23Sulfate, SO4

2+ (mg/L) 9 – 197 2,600 – 8,100Specific Conductance (μS/cm) 0.221 – 0.467 0.610 – 2.811DO (mg/L) ~8 – 10 ~1 – 35pH 6.5 – 8.6 2.5 – 4.8Redox Potential (mvolts) ~100 – 250 ~250 – 510

ND = Not detected above detection limit

SUMMARY

Naturally occurring acid rock drainage has historically and is presently dis-charging directly to North Fork Rapid Creek, North Fork Castle Creek, Castle Creek, and some unnamed tributaries. Based on surface and ground water data collected for this investigation in 2002 and 2003, the discharge is acidic, contains high concentrations of iron, aluminum, and sulfate, is naturally occur-ring, and originates from pyrite-rich Precambrian metamorphic rocks. Where the acidic drainage mixes with uncontaminated surface water, red-orange iron hydroxide and white aluminum hydroxide deposits quickly precipitate out of solution and coat the stream substrate. Some of the precipitate remains sus-pended and is transported downstream, negatively affecting the stream habitat for several hundred meters. With time, diagenetic processes cause the precipitate to dewater and lithify, cementing the stream substrate and ultimately forming iron-cemented conglomerate. While the overall upstream to downstream surface water chemistry is little affected, the precipitates have a significant negative effect on reaches of the stream’s habitat. Damage or destruction of the biotic habitat is the primary effect and results from mineral precipitation on and within stream substrate. Plant and animal life are stressed in reaches of the streams. Where effects are great, reaches of stream may be devoid of most life. The absence of plants and macroinvertebrates results in a habitat unfavorable to fish. Natural chemical reactions attenuate the effects, ultimately restoring the stream’s water quality and allowing the stream’s habitat to become favorable to plants, macroinvertebrates, and fish. This natural restoration appears to occur within approximately one to two km downstream from acid rock drainage dis-charge locations. The neutralizing capacity of North Fork Castle Creek surface water is exceeded by acid rock discharge and stream quality does not improve be-fore its confluence with Castle Creek. Approximately three to four km of North Fork Castle Creek upstream from its confluence with Castle Creek are negatively


affected by acid rock drainage. Based on visual observation, the damage to the stream may be exacerbated by cattle grazing. Iron bogs are depositional features and many were mined from the early 1900s up through the 1960s because of their high iron content (approximately 50-percent iron). While iron bog mines do impact the environment, in this study area, the mines are relatively small. When compared to the total naturally occurring acid discharge in the two watersheds, the mines probably have a neg-ligible effect on overall water chemistry. Even with the absence of mining, iron bogs and acid rock drainage would be present and affecting the environment in this study area.

ACKNOWLEDGEMENTS

We would like to acknowledge the South Dakota Department of Environ-ment and Natural Resources for their technical support and use of their auto-mated sampling equipment, and the South Dakota Department of Game, Fish and Parks, for their technical support. Dr. Edward Duke of the South Dakota School of Mines and Technology performed the visible and near infrared reflec-tance spectroscopy. This study was financed through research grants received from the South Dakota Department of Environment and Natural Resources, the South Dakota Department of Game, Fish and Parks, the Black Hills Flyfishers, and the Geological Society of America.

REFERENCE CITED

Carter, J.M, D.G. Driscoll, J.E. Williamson, and V.A. Lindquist, 2002, Atlas of Water Resources in the Black Hills Area, South Dakota, U.S. Geological Survey, Hydrologic Investigations Atlas HA-747, 120 p.

Deer, W.A., Howie R.A., and Zussman, J., 1980, An Introduction to the Rock-Forming Minerals, Longman, London, pp. 445-461.

DeWitt, E., J.A. Redden, D. Buscher, A.B. Wilson, 1989, Geologic Map of the Black Hills Area, South Dakota and Wyoming, U.S. Geological Survey, Mis-cellaneous Investigations Series MAP I-1910.

Pennsylvania Department of Environmental Protection, 1998, Coal Mine Drain-age Prediction and Pollution, Brady, K.B.C., Smith, M.W., Schueck, J. edi-tors, 1998, pp. 1-1 to 1-22.

Piper, Arthur M., 1944, A Graphic Procedure in the Geochemical Interpretation of Water-Analysis, Hydrology Papers, Transactions, American Geophysical Union, pp. 914-23.

Stiff, Henry A., Jr., 1951, The Interpretation of Chemical Water Analysis by Means of Patterns, Journal of Petroleum Technology, vol. 3, no.10, pp.15-17.

U.S. Geological Survey, March 1 2004, pHREEQC (Version 2)—A Computer Program for Speciation, Batch-Reaction, One-Dimensional Transport, and In-verse Geochemical Calculations, Version 2.8.01.

U.S. Geological Survey, 1975, Mineral and Water Resources of South Dakota, U.S. Government Printing Office, Washington, pp. 96-98.


HYDROGEOLOGY OF LOWER SPEARFISH CANYON

Perry H. Rahn Department of Geology & Geological Engineering


ABSTRACT Presently Barrick Gold Corporation diverts Spearfish Creek into a tunnel in the lower portion of Spearfish Canyon. The water is taken through the tunnel and an aqueduct to Hydro #1, a hydroelectric power plant built by Homestake Mining Company in 1911. Just below the tunnel diversion site, the streambed crosses outcrops of the Madison Limestone and the Minnelusa Formation. While pre-diversion data is scanty, it appears that most of the water in Spearfish Creek continued past sinkhole zones in these formations to the present town of Spearfish. However, some recharge to these two aquifers most likely occurred in this reach prior to 1911. This recharge is believed to have supported the springs at the D.C. Booth Fish Hatchery. The tunnel diversion leading to Hydro #1 began in 1911, and the D.C. Booth Fish Hatchery springs failed approximately 5 years later, presumably because the springs were no longer being recharged in the upstream sinkhole zone. If, in the future, the tunnel diversion is abandoned and Spearfish Creek is returned to its natural condition, Spearfish Creek water should again recharge the two aquifers, and within approximately 5 years the D.C. Booth springs should flow again.

INTRODUCTION Spearfish Creek is one of the major streams in the Black Hills, and it is a popular tourist attraction. The perennial flow of Spearfish Creek originates from cold spring water primarily discharging from the Madison Limestone and Min-nelusa Formation. Route 14A is located in Spearfish Canyon, and is designated a “National Scenic Byway”. The 2003-2004 South Dakota Vacation Guide describes Spearfish Creek as “one of the prettiest streams in the Hills. Trout fish-ermen come to clear-running Spearfish Creek all year long.” However, through much of the lower Spearfish Canyon there is only a dry streambed. The water has been diverted into a 5.5 mile-long tunnel and aqueduct for the generation of electricity. There is a long and complex history of the diversion of water in the northern Black Hills by Homestake Mining Company (HMC). Water has been diverted from numerous watersheds for the Homestake gold mine at Lead, SD. The diversions are based on the doctrine of appropriative rights, the water law in the western USA. In the late 1900s, HMC began diverting water in a series of aqueducts and tunnels from numerous springs in the headwaters of Rapid, Elk, and Spearfish Creeks. At the headwaters of Spearfish Creek and Little Spearfish


Creeks, above Cheyenne Crossing, water today is still conveyed to the Hanna Pump Station and sent to Lead (Fig. 1). This water had been used by HMC for the operation of their gold mine. After metallurgical processing, the water and tailings were discharged into Whitewood Creek (Rahn et al., 1996). In 1977 enforcement of the Federal Water Pollution Control Act resulted in a cessation of direct disposal into Whitewood Creek. The Homestake Mine closed in 2001. Presently some water from the Hanna diversion is used by the towns and Lead and Deadwood as part of a contractual agreement between HMC and the Lead-Deadwood Sanitary District. Miles downstream from Hanna there are two hydroelectric power plants where the remaining water in Spearfish Creek has been diverted to produce electricity for the Homestake mine. The upstream diversion is at Savoy, where, until November 2003, water was collected from Little Spearfish Creek (just below Roughlock Falls) and from Spearfish Creek and transferred to a flume on the eastern side of Spearfish Canyon. The flume carried the water approximately 6.3 miles downstream (distances measured along Rt. 14A which closely follows the stream channel) to HMC Hydroelectric Plant #2 at Maurice, SD (Fig. 1). Hydro #2 was built in 1917. In 2003 the South Dakota Department of Game, Fish and Parks obtained the water rights for Hydro #2. Because of the termina-tion of these diversions to Hydro #2, water now flows down Spearfish Creek and Little Spearfish Creek to Savoy. Water in Little Spearfish Creek, famous for Roughlock Falls, now continues down into Spearfish Creek at “Spearfish Falls”, a scenic location made famous by early photographs of a train on a bridge over the falls. One mile below Hydro #2, the water in Spearfish Creek is diverted into a tunnel (Fig. 1). The tunnel intake and aqueduct diversion is 5.5 miles long and ends at HMC Hydroelectric Plant #1, built in 1911. Hydro #1 is still func-tioning, although presently much of the diverted water reportedly bypasses the turbines since the mine is closed and large amounts of electricity are no longer needed. Hydro #1 is 2,000 ft upstream from the D.C. Booth Fish Hatchery, located just above the town of Spearfish. The water exits Hydro #1 and provides a sustained flow for Spearfish Creek through the city park and on through Spear-fish. The average discharge in this reach is approximately 54 cfs (Driscoll and Carter, 2001). Above Hydro #1, the streambed of Spearfish Creek is essential dry all the way up to the tunnel diversion, a distance of approximately 7.3 miles. This reach includes the location of “Bridal Veil Falls”, where Rubicon Creek enters Spearfish Canyon.

PROPOSED TERMINATION OF DIVERSION TO HYDRO #1

It has been proposed (Rapid City Journal, October 19 and 23, 2003) that Hydro #1 be abandoned in a manner similar to Hydro #2. This would allow water to be returned to its natural streambed over the entire reach of Spearfish Canyon from Cheyenne Crossing to the town of Spearfish. The advantages of an additional 7.3 miles of spring-fed waters flowing in this beautiful canyon would be recreational use, including fishing, kayaking, and the increased value to tour-


ists who could see clear flowing water instead of bleached-white cobbles in a dry streambed. The mechanism whereby the termination of the diversion to Hydro #1 could be accomplished is not known. The Homestake Mine is now owned by Barrick

Figure 1. Topographic map showing the Spearfish Creek area from Cheyenne Crossing to Spearfish. Water diversions along Spearfish Creek are shown. Modified from U.S. Geological Survey 30 X 60 minute series topographic map of Rapid City, South Dakota. Original at 1:100,000 scale. Contour interval 50 meters.


Gold Corporation, a Canadian mining company. They have legal water rights for the diversion. But the mine is no longer operating; and Hydro #1 is 92 years old and its continued life is limited.

SINKHOLES

The water that is returned to Spearfish Creek after passing through Hydro#1 presently flows through the town of Spearfish. There is concern that if the diver-sion to Hydro#1 is terminated and the water returned to its natural streambed there could be no more water flowing through town because it would be lost into sinkholes in the “loss zone” immediately below the tunnel diversion site. An opinion expressed in the January 19, 2002, editorial page of the Rapid City Jour-nal is that water released from the tunnel diversion and allowed to flow down the natural streambed would be swallowed up by a fault. But a detailed geologic map (Lisenbee and Redden, 1991) shows there are no faults in this reach of Spearfish Canyon except a trivial one near Bridal Veil Falls. Based on the geologic map, if the tunnel diversion is terminated it is very unlikely that there would be any water loss for the first 3 miles below the tun-nel diversion site because the bedrock in this reach is the Cambrian Deadwood Formation to Devonian Englewood Formation. Swampy conditions exist in this reach of Spearfish Canyon and the water table seems to be higher than the streambed. Further downstream the Madison Limestone (Fig. 2) and Minnelusa Formation crop out and the hydrogeology changes from an effluent to influent conditions. As a result, some water may sink into these two aquifers in this loss zone. The U.S. Geological Survey (Hortness and Driscoll, 1998) conducted some stream gaging below the tunnel diversion site leading to Hydro #1. In June 1996, 54 cfs was discharged below the tunnel into the streambed of Spearfish Creek and 33 cfs came out below the loss zone. Thus, after approximately one month of study, 21 cfs was lost to the alluvium and Madison Limestone. It may be that nearly all the water will eventually flow all the way into the town of Spearfish, but it may take months or even years for this to happen. The tunnel diversion to Hydro #1 has been going on for 93 years. If the tunnel diversion ceases, it will take a while to recharge the alluvium and Madison Limestone and Minnelusa Formation back to their original condition. Because of the tunnel diversion, the channel has been deprived of water for 93 years and in a sense is “thirsty”. More study is needed, especially an evaluation of the discharge conditions when Spearfish Creek has low discharge as it flows into lower Spearfish Canyon. It is my opinion that in all probability after cessation of the tunnel diversion these aquifers will be eventually recharged to their original water table elevation and most of the water will cross the loss zone.


Figure 2. Geologic map of the lower Spearfish Canyon showing the altitude (ft) of the top of the Madison Limestone (Carter and Redden, 1999). The D.C. Booth Fish Hatchery is located. The dark color is the outcrop area of the Madison Limestone and Englewood Formation. Original at 1:100,000 scale. Contour interval 200 ft.


HISTORIC REFERENCE

Draining from the Precambrian core of the Black Hills, many streams naturally lose all but flood flows to the Madison Limestone in the Black Hills (Rahn and Gries, 1973; Hortness and Driscoll, 1998). Examples are Elk Creek, Boxelder Creek, Spring Creek, and French Creek. This water is a source of recharge to large springs at the outer periphery of the Paleozoic carbonate belt such as Cleghorn Spring on Rapid Creek, Cascade Springs, the spring at Ranch A along Sand Creek, the spring along Beaver Creek above Buffalo Gap, and the springs supporting the Fall River in Hot Springs. Two large streams in the Black Hills maintain their flow across the Paleozoic belt: Rapid Creek and most likely Spearfish Creek. The HMC diversions started so long ago that no one remembers if Spearfish Creek used to flow all the way through Spearfish Canyon. However, there is one published reference to the perennial flow of Spearfish Creek, described by Colo-nel Richard I. Dodge in 1875:

“Spear-Fish Creek. This creek is second only in volume to Red Water, being about three-fourths its size. It is purer, colder, clearer, softer, deeper, and much more rapid, rushing between its banks with the force of a cataract. It differs from all other streams which flow north through the mesa, in that it does not sink anywhere. Rising with a bound from the earth, not far from Crook’s Monument, it flows with the directness and force of a torrent, eating away it rocky bed until it has cut a canon for many miles of its course of not less than two thousand feet in depth. One of the surveying parties, getting into this canon from its head, found not a single place for more than thirty miles where its walls would have been scaled, and had to force a way through to where it comes out on the northern plain.”

D.C. BOOTH FISH HATCHERY

Many streams lose water to the Paleozoic limestone aquifers, but this water is not really lost because it helps recharge these aquifers. It helps support the big springs in the Black Hills such as Cleghorn Springs, Cascade Springs, Hot Springs, and many other springs (Rahn and Gries, 1973). Based on the hydro-geology of the lower Spearfish Canyon it is very likely that originally any water recharged in this loss zone helped sustain springs near or in the town of Spearfish. This would include the now defunct springs at the D.C. Booth Fish Hatchery as well as other springs that reportedly flowed in the 19th century such as Kroll’s spring, Randall spring, Saratoga spring, and a spring at the old grist mill. Figure 3 is a graphical plot of the discharge at the D.C. Booth Fish Hatchery. This data was supplied by the files from the D.C. Booth Historic National Fish Hatchery (now maintained by the U.S. Fish and Wildlife Service). The D.C. Booth springs originally belonged to John Johnston; these springs included the “Upper Spring” in “Ames Draw” which is a gully shown as “Fish Hatchery Gulch” on the 1:24,000 scale USGS topographic map of the Spearfish quad-


rangle. A reliable measurement made in 1892 by Evermann was 1,100 gpm (equivalent to 2.45 cfs). Later measurements in 1896 and 1897 are not as reli-able. There is a hiatus of information until comments by the U.S. Fish Commis-sion Annual Report in 1917 wherein “bad conditions” were observed. In 1919 “less discharge” was reported, and in 1920 the discharge was “as low as 22 gpm”. The main spring on Ames Draw was dynamited in a futile attempt to rejuvenate it. Essentially the D.C. Booth Fish Hatchery springs originally produced 1,100 gpm but by 1917 had all dried up.

Hydro #1 was built in 1911, and water was diverted from its normal course down Spearfish Canyon. The water originally crossed the Madison Limestone and the Minnelusa Formation at an elevation higher than the D.C. Booth Fish Hatchery springs. The Madison Limestone crosses under the streambed of Spearfish Canyon at an elevation ranging from 3,800 to 4,000 ft above sea level, whereas the “Upper Spring” at D.C. Booth Fish Hatchery is at 3,710 ft above sea level. Carter et al. (2003) show the potentiometric surface of the Madison

Figure 3. Discharge at the D.C. Booth Fish Hatchery(1) for the years 1890 to 1970. From U.S. Fish and Wildlife Service files. The dashed line is the estimated discharge for all of “Johnston’s Springs” which includes all springs between the “Upper Spring” in Ames’ Draw(2) all the way to Spearfish Creek. (1) In 1905 this property was acquired by the U.S. Fish Commission. Later it became the U.S. Fish and Wildlife Service “D.C. Booth Fish Hatchery”, and later the “D.C. Booth National Historic Fish Hatchery”.(2) “Ames Draw” is called “Fish Hatchery Gulch” on the U.S. Geological Survey 1:24,000 scale topographic maps of the Spearfish and Maurice quadrangles.


Limestone and the Minnelusa Formation; they are both approximately 3600 ft above sea level at the D.C. Booth Fish Hatchery. Based on the general hydrogeol-ogy of the Black Hills, water sinking in these two units in the canyon could be expected to reappear as “resurgent” springs at a low elevation of the outcrop of the Madison Limestone and Minnelusa Formation. An analogous hydrogeologic situation is Dark Canyon along Rapid Creek and the occurrence of Cleghorn Spring (Rahn and Gries, 1973). “Upper Spring” in Ames Draw is located approximately 600 ft south of the old D.C. Booth Fish Hatchery ponds. The reason for its precise location here may be due to a “breccia pipe”. There is a small sinkhole approximately 50 ft southeast of the (now dry) spring outlet. Karst features like this are believed to reflect the presence of a permeable conduit through the Minnelusa Formation (Epstein, 2000). Presumably “Upper Spring” was a resurgent spring, recharged by Spearfish Creek in the loss zone. The water then travelled downgradient from the loss zone and discharged through a breccia pipe. [It is unlikely that this spring was unrelated to Spearfish Creek, possibly due to some local perched water table. Perched water table springs do exist in the Black Hills; for example “Jones’ Spring along Elk Creek is related to Tertiary intrusives. But there are no Tertiary intrusives within 2 miles of “Upper Spring”.] The nearby valley of Sand Creek, Wyoming, provides another hydrogeologic analogy to lower Spearfish Creek and its relationship to the D.C. Booth springs. Upper Sand Creek, draining the Cement Ridge area, has a good perennial dis-charge but it sinks into the Madison Limestone just above its confluence with the (normally dry) Cold Springs Creek. Approximately 4 miles downvalley, at the lowermost outcrop of the Madison Limestone, a large spring occurs at Ranch A. There is no doubt that this resurgent spring is largely supplied by the recharge 4 miles up-valley. If a tunnel diversion circumvented the recharge site the magnifi-cent spring at Ranch A would be compromised. In all probability prior to 1911 some water from Spearfish Creek percolated into the Madison Limestone and the Minnelusa Formation in Spearfish Canyon. This recharge contributed to the D.C. Booth Springs. When the tunnel diversion began in 1911, this source of recharge essentially ceased and within a few years the D.C. Booth springs failed. If, in the future, the tunnel diversion to Hydro #1 ceases, and water again flows across the loss zone, it stands to reason that recharge to the two aquifers will again occur and sometime thereafter the D.C. Booth springs will flow again. The exact time interval between recharge and the rejuvenation of the springs is hard to predict with certainty, but Figure 3 can be used as a guide. It appears that within approximately 5 years after Hydro #1 was put on line, the springs failed. Therefore, it can be anticipated that within ap-proximately 5 years after the tunnel diversion is abandoned, the springs at D.C. Booth should produce their original discharge.

SUMMARY

If the tunnel diversion for Hydro #1 is abandoned the water will once again flow down the natural channel of Spearfish Creek. This action has the support of


many local landowners including the Spearfish Cabin Owners Association, and the Spearfish Canyon Preservation Trust. There is concern that if the tunnel diversion to Hydro #1 is terminated, Spearfish Creek would sink into the Madison Limestone and there would be no more water at the Spearfish city park or through the city of Spearfish. But I believe that in all probability nearly all of the water will eventually flow through. Further, after approximately 5 years any water lost in Spearfish Canyon below the tunnel diversion site will help revive the flow of springs at the D.C. Booth Hatchery as well as other springs in the town of Spearfish. Future research needs to determine if 21 cfs is a constant loss or to what degree this changes with time. During preliminary releases to the streambed, the D.C. Booth springs and other springs should be monitored, as well as the potentiometric levels. Future research should also evaluate the trout habitat in the reclaimed reach since there is some concern that engineering structures along Rt. 14A have detrimentally changed the streambed. The cessation of the tunnel diversion must be carefully conducted so as to maintain a live stream through the town of Spearfish. Presently Barrick Gold Corporation, a Canadian mining company, diverts Spearfish Creek to Hydro #1 for the purpose of generating electricity. The water right is not contingent on an operating mine. There is no longer mining, although electricity is being used for reclamation activities. This is an opportune time to restore Spearfish Creek back to its natural condition. Spearfish mayor Jerry Krambeck announced (Rapid City Journal, January 15, 2004) that city officials were trying to buy the water pipeline and Hydro #1. It would benefit South Dakotans to have a live stream once again. If the diversion to Hydro #1 were terminated, approximately 7.3 miles of a dry streambed could be turned into a beautiful flowing stream.

ACKNOWLEDGEMENT

Randi Sue Smith, Curator of the D.C. Booth Historic National Fish Hatch-ery, kindly allowed me access to the Archives of the D.C. Booth Fish Hatchery.

REFERENCES CITED

Carter, J. M., and J. A. Redden. 1999. Altitude of the top of the Madison Lime-stone in the Black Hills of South Dakota. U.S. Geological Survey Hydro-logic Atlas HA-744D.

Carter, J. M., D. C. Driscoll, and J. F. Sawyer. 2003. Ground-water resources in the Black Hills area, South Dakota. U.S. Geological Survey Water-Resources Investigations Report 03-4049.

Driscoll, D. G., and J. M. Carter. 2001. Hydrologic conditions and budgets for the Black Hills of South Dakota, through water year 1998. U.S. Geological Survey Water-Resources Investigations Report 01-4226.


Epstein, J. B. 2000. Gypsum karst and hydrologic evolution in the northern Black Hills. in Strobel, M. L., et al., eds. Hydrology of the Black Hills. South Dakota School of Mines and Technology Bulletin No. 20, p. 73-79.

Evermann, B. W. 1892. Report of investigations for the selection of a site of a fish-cultural station in Iowa, Nebraska, South Dakota or Wyoming. Barton W. Evermann, Ph.D., Assistant, U. S. Fish Commission.

Hortness, J. E., and D. G. Driscoll. 1998. Streamflow losses in the Black Hills of Western South Dakota. U.S. Geological Survey Water-Resources Investiga-tions Report 98-4116.

Lisenbee, A. L., and J. A. Redden. 1991. Unpublished geologic map of the Spearfish 1:24,000 scale quadrangle, South Dakota. South Dakota School of Mines and Technology, Governor’s Office of Economic Development.

Rahn, P. H., and J. P. Gries. 1973. Large springs in the Black Hills, South Dakota and Wyoming. South Dakota Geological Survey, Report of Investigation No. 107, 46 p.

Rahn, P. H., A. D. Davis, C. J. Webb, and A. D. Nichols. 1996. Water quality impacts from mining in the Black Hills, South Dakota, USA. Environmen-tal Geology 27: 38-53.


MOLECULAR DYNAMICS OF ATOMIC CLUSTERS: AN OBJECT ORIENTED APPROACH

John M. Schneiderman and Brian G. MooreChemistry Department

Augustana CollegeSioux Falls, SD 57197

ABSTRACT

The purpose of the project was to develop a graphical user interface (GUI) to simulate and display the interactions of a small amount of atoms in a clus-ter, specifically with interactions defined by a pairwise Lennard-Jones potential energy. Another goal of the project was to design the coding of the program in such a way that it would reflect how the various parts of the program interacted with each other. To this end four classes were developed: an atom, a cluster, a base simulation, and a molecular dynamics Lennard-Jones simulation. The atom class contained all the information that was necessary for the simulations, such as the potential energy of the atom. The program was designed with the ability to add new types of simulations, with no need to significantly modify the GUI code. In the GUI program the ability to watch the simulation as it takes place was introduced, as well as the ability to modify how the atoms themselves appear. Results are presented for a specific system, a cluster of rare gas atoms. A histo-gram of the potential energy is bimodal for a small (N appprox. 30-200 atom) cluster, a quantitative measure of the two qualitatively different regions--surface and bulk.

Keywords

Clusters, molecular dynamics, computational, simulations, C++, object-ori-ented, undergraduate, liquid, phase transition

INTRODUCTION

Involving undergraduate students in research in computational chemistry techniques can be very difficult because they must have extensive preparation before they can become effective contributors. In particular, for development of new computational methods, programming experience is required. For many problems, especially those of primarily a numerical nature, a wide variety of software and/or programming languages may be used to accomplish the desired result. Meanwhile, the programming languages and development environments which are commonly taught at the undergraduate level continue to change, re-flecting the changing situation in companies which use software and the software development industry. The current undergraduate computer science curriculum


typically introduces object-oriented programming (usually either C++ or Java) early on, and most development is done in a graphical integrated development environment, for example Microsoft Visual Studio.NET. In this paper, we discuss the adaptation of a classical molecular dynamics simulation program from a procedural language, (Fortran/C), to an object-ori-ented approach (specifically, using C++). This program, which was previously run only as a console application under UNIX (Linux), was redesigned to make use of the object-oriented capabilities available in C++. In addition, a graphical interface was developed for MS Windows, allowing real-time visualization of the atoms as they progress in the simulation. In undertaking a computational problem where writing new code is in-volved, an important consideration is always whether the questions at hand could be answered by an existing software product. Hence there is an important coupling between the scientific questions and the techniques one brings to bear on it. In this paper, we will discuss a specific system, an atomic cluster of rare-gas atoms, and use it as a basis to discuss a particular software approach and development environment that involves integrating the kinds of skills mentioned above--particularly object oriented programming. Our goal here was not merely to just convert the syntax of the program from C to C++, we were also interested in how the approach to the problem is differ-ent. Object-oriented programming is a completely different paradigm than the structured approach. In the work here, we have tried to completely re-think the problem from scratch, applying objective programming principles.

CLUSTERS AND PHASE TRANSITIONS: THE LIQUID STATE

Atoms or molecules expanded into a vacuum (such as in a molecular beam) will undergo expansion cooling and under certain conditions begin to form small clusters of atoms (Jellinek. 1999). Once formed, we can consider these clusters as isolated systems; typically the clustering occurs in only the initial stages of the expansion. These clusters can then be observed--for example by mass spectrom-etry, optical spectroscopy, or other means--and their properties studied (Johnson, et al. 2002). Not surprisingly, they can show very different behavior from that seen in the bulk. While the details of the clustering process are of course depen-dent on the particular nature of the species in the expansion, the phenomenon of cluster formation is quite general; dispersive (van der Waals) attractive interac-tions are present in any system. For this reason, we chose in our simulations to focus on the dispersive interac-tions alone, which may be modeled well by the Lennard-Jones potential (Maitland, et al., 1981); hence, when we speak of the kind of cluster we are modeling, one can perhaps best think of it as that of a rare-gas atom, such as argon or krypton. We have included the dispersive attraction and a short range repulsive interaction, but we have eliminated any possibility of actual “chemistry” occurring--no chemical reactions. The functional form of the Lennard-Jones potential is


where r is the interatomic separation, σ is the atomic diameter, and ε is an en-ergy parameter characterizing the depth of the attractive part of the potential well. In reporting results in this paper, we will use dimensionless reduced units: reduced temperature T* = kT/ε, reduced distance r* = r/σ, reduced density ρ* = (N/V)σ3, reduced time t* = (ε/mσ2)1/2t. The potential energy is taken to be pair-wise additive; the total potential energy of the system is the sum over all the pair interactions in the system. The typical range of sizes of atomic clusters studied in our work was from N=13 to about N=300. One of the most interesting features of a cluster of atoms interacting through primarily dispersive interactions is the observation of a phase transition (liquid/solid) in the cluster, with characteristics which are de-pendent on the size of the cluster. This phase transition has been seen clearly not only in computer simulations but also in experimental observations (Hahn and Whetton, 1988). One can think of the process mentally in the following way: an isolated group of atoms in a vacuum at a low enough temperature will stay “stuck” together for long enough to be characterized. Now imagine slowly heat-ing the cluster; this is not possible for a real cluster in a vacuum (nor is it the way that the simulations are conducted in this work), but this thought experiment serves to illustrate the idea. At some point, desorptions of atoms will begin to occur. However, for many cluster sizes in the ranges we studied, the cluster will melt before the evaporation of atoms becomes so rapid as to preclude obtaining structural information about the liquid cluster (Moore and Al-Quraishi, 2000). The liquid state of the cluster is particularly interesting. For example, if the cluster is of a size near N = 55, the melting will occur near a temperature of T* = 0.35, while in a bulk sample of the same atoms, the triple point is at about T* = 0.80. Thus the liquid in the cluster is existing at a temperature very much below where the liquid would cease to be stable in the bulk.

SIMULATION METHODS: GENERAL APPROACH

The methods for the molecular dynamics simulation were the same in both the older versions of the program (implemented with a procedural approach) and the newer versions developed for this work (implemented with an object-oriented approach). The velocity Verlet algorithm was used, with constant energy conditions (micro-canonical ensemble) (Verlet, 1967). A standard step size of dt* = 0.005 was used for all runs. Initial conditions for the clusters were face-centered cubic arrangements with all particles having zero velocities. An important adjustable parameter in our approach is the reduced lattice parameter a* = a/σ, where a is the edge length of the unit cell of the initial fcc lattice. Our goal is primarily to study liquid clusters, so a relatively large lattice parameter is typically used, in the range of a* = 1.7 to 2.0, corresponding to initial cluster densities in the range of ρ* = 0.80 to 0.50. By comparison, close-packed for the fcc lattice would occur at a* = 1.414 = √

— . The larger parameters typically lead to liquid clusters, with complete destruction of the initial fcc lattice. The fact that the simulation bears no “memory” of the initial fcc configuration is also confirmed by the radial distribution functions (RDFs) which were acquired and

2


qualitative observations of the cluster using the graphical user interface (GUI) which displays the positions of the atoms as the simulation progresses--developed for this paper. Both of these ways of observing the cluster (RDF and GUI) are discussed in more detail later in later sections.

PROGRAM DESIGN: PROCEDURAL APPROACH (C/FORTRAN)

The program to implement the molecular dynamics simulation was origi-nally developed in a mainframe environment (DEC/VAX and SGI systems) in Fortran 77 (f77). Subsequent to that, the program was ported to the C language for use in a Linux environment, most recently using the free GNU gcc compiler (Stallmann). Both versions of the program (C and f77) had very similar designs; the conversion to C was primarily of a syntactical nature. These older versions of the program were run exclusively as console applications, with data files for input and output. Data analysis consisted of using other applications (such as gnuplot) to visualize results after the program finished running. Both C and Fortran are procedural languages and the program designs reflected that choice. The procedural programming paradigm focuses on the procedures and sequences of steps required for the problem at hand. Functions are used to help organize the processing of the algorithm, etc. The program was organized roughly into three sections: input and pre-run processing, the dynamics run (including running data acquisition), and post-run operations and writing out of the final output files. Within the run loop (the central section of the program), there were three functions, move1, force, and move2, called in sequence for each move of the atoms, that contained the es-sence of the Verlet MD algorithm, patterned after routines of these same names provided in the text by Allen and Tildesley (1987). The force loop is the most time-consuming part of the program, as it contains the sum over all unique pairs of atoms, so as to calculate the force on each atom; the number of calculations in the force loop thus scales as roughly N2, where N is the number of atoms.

PROGRAM DESIGN: MODULAR/OBJECT-ORIENTED APPROACH (C++)

Our goal was to re-design the program with an object-oriented approach. This was accomplished in stages. The first stage was the relatively simple task of merely rewriting the program in C++, but preserving essentially all the original code, making only minor changes for syntax as required. The second stage was to make use of some of the classes provided in the C++ Standard Template Library (STL) (Stepanov and Lee, 1995)--in other words to make use of this standard set of pre-defined classes but not introducing any new user-defined ones. The third stage was to create a set of classes designed specifically for the problem at hand, yet general enough to be useful for future development of the code (for example to change the simulation method or to change the potential, etc.).


Using the Standard Template Library (STL)

The second stage in this process of moving from C to C++, the use of the STL, is worthy of some comment. The STL provides several very useful classes which can make program code easier to write and read and also provide more flexibility and some protection against errors. An example of this is the vector class. This class is one of several container classes provided in the STL. In C, data arrays must be dimensioned, and making an array larger or smaller is te-dious. Errors resulting from trying to access non-existent elements are common. The vector container provides methods to expand or contract the size of the array with a simple syntax; also, access may be checked for out-of-range condition. An example of the ease with which the vector class may be used is shown in the following code-snippet, where data is being read from a file (one value per line) into a vector.

vector<double> data;

double value;

while(input >> value) data.push_back(value);

int n = data.size();

cout << n << “ data points read\n”;

In the above example, data is the vector in which we are placing the num-bers, which are read one line at a time from the input stream. The methods push_back() and size() are examples of functions provided with the class, al-lowing for an automatic increase in the size of the vector as the numbers are put into it and interrogation of the number of elements when this process is finished. The corresponding code to achieve the same results takes up several more lines when rendered in C. Note also in this example that data is an example of an “object” in C++. Methods (or functions) are applied to the object using the postfix notation, thus data.size() is the operation of the class function size() on the object data.

Object-Oriented Design: User-Defined Classes

The power of object-oriented programming is that it allows the user to define new classes, similar to the vector class (defined in the STL); another way of thinking of a class is just as an abstract data type. Data types such as int or double are system provided; C++ allows the user to define completely new data types. However, the object-oriented programming paradigm is completely different than that for procedural programming. Rather than the main focus being the procedures, the first step is to design and create the classes which will be used. The classes should reflect the nature of the problem at hand. Ideally, the classes should reflect real objects if possible (Stroustrup 1997). Towards this end, the first main step in re-designing the simulation program in an object-oriented way was the creation of two classes, the atom class and the cluster class. The atom class contains the x, y, and z values for the position, veloc-


ity, and acceleration of the given atom, as well as the kinetic and potential energy of the atom. Accessors are provided to extract these variables, and method func-tions provided to change them. The class was designed to be flexible, with the idea that for further development, more variables could be added. For example, in our program, only one kind of atom is used. In future programs, one might want to have a mixture of different kind of atoms, so one might want to add variables such as the Lennard-Jones diameter and potential well depth, etc. An important consideration in the program design was the fact that the position, velocity, and acceleration values are frequently accessed. These variables are needed in the force loop, which is the most computationally intensive part. In the older C and Fortran programs, these arrays were declared as common or global variables, accessible to any function. This special status for these particular variables was dealt with by making the accessor functions to these variables in the atom class inline reference functions. The syntax in the class declaration for these functions is shown below. Thus, for example, in the molecular dynamics move1 routine, we can make a statement of the style below.

clus[i].vx() += 0.5*dt*clus[i].ax();

Note here that the ax() accessor is used to retrieve the value of the acceleration, and then the vx() function is used to change the value of the acceleration.

class Atom

{

private:

double rxval,ryval,rzval;//positions

double vxval,vyval,vzval;//velocities

double axval,ayval,azval;//accelerations

double Vatom,Katom;// potential energy, kinetic energy

public:

/* Class functions. */

Atom(); // constructor that initializes all values to zero

double& rx() {return rxval;}

double& ry() {return ryval;}

double& rz() {return rzval;}

double& vx() {return vxval;}

double& vy() {return vyval;}

double& vz() {return vzval;}

double& ax() {return axval;}

double& ay() {return ayval;}

double& az() {return azval;}

double getVatom() const;

double getKatom() const;


void move(double rxnew, double rynew, double rznew);

void push(double vxnew, double vynew, double vznew);

void jerk(double axnew, double aynew, double aznew);

void putVatom(double val);void putKatom(double val);

};

The cluster class is primarily a collection of atoms, plus related accessors and other methods. One way to implement this is to set up the cluster class to inherit from the STL vector class. Thus, methods such as push_back(), size(), etc. are automatically provided. The cluster object is just a vector of atoms, plus a few other related variables.

class Cluster: public vector<Atom>

{

private:

double xcm,ycm,zcm;//center of mass

double vxcm,vycm,vzcm;//velocity “center of mass”

double V,K;// total potential energy, kinetic energy

public:

/* Class functions. */

Cluster(); // constructor

/* fccgen is not a constructor, but it does kind of the

same thing. It generates an fcc lattice. */

void fccgen(int m, double rhostar);

/* Accessors. */

double getV() const;

double getK() const;

void pollVK(ostream& ostr) const;

double getxcm() const;double getycm() const;double getzcm() const;

double getvxcm() const;double getvycm() const;double getvzcm() const;

/* Calculators & such. */

void calcVK();

void calccm();

void putV(double value);void putK(double value);

void changecm(double x,double y, double z, double vx,double vy, double vz);

void readcn(string name);

void writecn(string name);

};

The atom and the cluster class are just data storage and retrieval objects. The actual algorithmic procedures to implement the simulation must act on these ob-jects. Toward this end, we developed two more classes, which are more abstract, but reflect the organization of tasks in the program. The main idea was to create a class that represented the simulation run itself--the central part of the original


C program. The specific method we chose to implement the molecular dynamics simulation involved using the Verlet method, Lennard-Jones potential, but these choices are only one specific case of many possibilities. Thus we decided to de-velop first a base simulation class, containing features common to any molecular dynamics simulation. Then we created a class called simulationmdlj containing the specific details of the method we have chosen (Lennard-Jones molecular dyn-mics). The simulationmdlj class inherits all the characteristics of the more general simulation class. The class declarations for these two classes are shown below.

class Simulation

{

public:

// class constructor

Simulation();

void setdefaults(string filename,int nstep);

void load(string currname, string prevname, int nstep);

void analyze();

void setstarttime(string base_name);

void writeout();

void summarize(int n,int nstep);

protected:

Cluster clus;

string curr_run_name,prev_run_name,defaults_file;

int nstep,iprint,igr,irho,iPKhist,iscreen;

double V,K,E,EN,VN,KN,Temp;

double t,dt,tstart;

/* Histograms */

Rdfhist rdfdata;

Rhohist rhodata;

PEhist PEdata;

KEhist KEdata;

};

class SimulationMDLJ : public Simulation

{

public:

// class constructor

SimulationMDLJ();

void run(int nstep);

// MD LJ core functions

void force();

void move1();

void move2();

};


Note the presence of the three functions move1, force, and move2 in the simu-lationmdlj class; these are essentially the same functions as those found in the original C/f77 versions of the program but adapted to the new programming environment.

Graphical User Interface (GUI)

The description of the program design so far applies only to the “number crunching” procedures of the simulation program. Data is read in, either from input files or the command line, the simulation progresses, and output is written out. This is a console application and can be compiled and run in any environ-ment where a C++ compiler is available. On a Linux system, the GNU gcc compiler is always available, and is the one we used solely for our UNIX runs. Under MS Windows, the console application can be developed in a variety of ways. Microsoft has a fully featured C++ compiler as part of their Visual Studio.NET development environment. Other compilers for MS Windows are also available, for the console application development. One can also use Windows ports of the free GNU gcc compiler; in our work, using gcc with Windows, we used either the mingw package or the cygwin environment. The MS Visual C++ approach to development provides an application framework which has support for development of programs with a graphical user interface (GUI). Fairly standard features, such as developing an input screen, with boxes for input parameters, are easily added. We decided as part of the program development to also add a screen which would show the positions of the particles on the screen as the simulation progressed. An example snapshot of the screen during a simulation is shown in Figure 1. The program has several features, not all of which will be described here. A few examples are presented in Figures 2 and 3. The development of the GUI provided a good example of a graphical development problem for the undergraduate student in this case and involved skills which built directly on those taught in programming courses. In addition, the ability to directly visualize the cluster during the simulation pro-vided a very real scientific benefit in terms of helping to verify whether or not the initial lattice structure was destroyed.


The main benefit from writing code to approach a computational problem is that one can ask questions which have never been asked before. Commercial and free programs are available to do molecular dynamics simulations on simple systems, but it is not easy to modify these programs to take data in a new way. We show in this section a few examples of the results from the program, with an emphasis on features which are hard to (or impossible to) realize in readily-avail-able existing programs. One important measure of cluster structure is the radial distribution func-tion (RDF). The idea of the RDF is a sort of “local” density, referenced to the particle frame. To acquire the RDF, we first pick a particular atom, then col-


Figure 1: A snapshot of the graphical user interface while the simulation is running. The cluster shown contained originally 108 atoms in a face-centered cubic lattice. In this run, several atoms were desorbed during the initial transient period of the run, and the lattice structure was completely destroyed, leaving a smaller liquid cluster. The desorbed atoms are discarded for subsequent runs and structural features are extracted on the remaining liquid cluster.

Figure 2: The file viewer screen included as part of the GUI. Shown are some of the summary numbers which are written out for the entire run. The units used are in re-duced form, defined in the text.


lect density information in radial shells centered on the chosen atom. Then we resolve this data into radial bins, normalize, and divide by the bulk density; the RDF shows deviations away from the bulk density in the environment nearby a given atom. An example of the RDF taken for a cluster of about N=55 atoms is shown in Figure 4. Note that the large radius limit goes to zero instead of one as in a bulk sample. This is because we are working with a finite system; for large enough radial distance, eventually we are outside the cluster. Shown in Figure 4 are two representative examples, for liquid and solid clusters of about the same size. The liquid cluster shows a typical solvation-type shell of a few atomic diameters, simi-lar to a bulk liquid, however, note the split second peak. This feature, which is not present in a bulk liquid, is ubiquitous for the cluster liquids we have studied. Such a feature is also often seen in glassy liquids. The solid cluster RDF has a few very distinct differences compared to the liquid cluster. First, the relative amplitudes of the two contributions to the split second peak are different; the larger radius contribution is enhanced. Secondly, the solid shows a clear peak at about r* = 1.7 which is not present at all in the liquid cluster. We believe this feature (the peak at r* = 1.7) to be a unique signature of the presence of solid ordering to some extent in the cluster, and we have used it to help identify when the cluster is in the pure liquid state. Another unique type of analysis we have performed is that of acquiring his-tograms of the kinetic energy and potential energy from the liquid cluster runs. Our initial intention was to try and use these histograms as an aid in quickly identifying particles which have desorbed from the cluster. These histograms have proved useful for that purpose, but they have turned out to be interesting in other ways as well. First, the kinetic energy histogram is unsurprising. Its form is close to that predicted by a Maxwell-Boltzmann distribution. However, the potential energy histogram appears qualitatively as shown in Figure 5. It is very strongly bimodal and appears to be dominated by two smooth peaks. An obvious interpretation is that we are seeing the two qualitatively different re-gions, surface and bulk, contributing to these histograms. This interpretation is reinforced by the dependence on size. When we compare the N=256 histogram

Figure 3: An example of the features available from the drop-down menus built into the GUI.


Figure 4: The radial distribution (RDF) of two clusters of approximately the same size, one in the liquid state, one in the solid state. The qualitative features of the RDF func-tion--especially the lack of a peak near r* = 1.7 in the liquid state--are typical.

Figure 5: Potential energy histograms of clusters in the liquid state, for two different sizes. The lower energy peak presumably represents the bulk contribution, the higher energy peak the surface contribution.


to the N=55 histogram, we find that the lower energy peak (the bulk part) has been enhanced at the expense of the higher energy peak (the surface part). These histograms have been normalized to one, so the total area under both curves is the same on the plot. It makes sense that the relative contribution of the bulk part will increase as the cluster size increases; the fraction of the atoms at or near the surface decreases with cluster size. We are in the process of performing more studies to confirm our interpretation of the potential energy histograms. In par-ticular, we hope to be able to spatially resolve the histogram acquisition process, thus conclusively proving exactly which part of the cluster gives rise to each peak in the histogram. To our knowledge, the potential energy histograms shown here are the first time such data has been collected for a simulation of a liquid cluster or droplet. This shows the power of being able to write unique code to analyze a problem of this nature.

PROGRAM AVAILABILITY

This project was not undertaken with the goal of reaching a program for general distribution, but rather to illustrate the general concept of incorporat-ing object-oriented features as part of program design, in part as a way to help integrate undergraduate student participation at an early stage. However, we will provide a copy of the program (under the general terms of a GNU license) at the corresponding author’s website, http://faculty.augie.edu/~bmoore, for those who are interested in seeing more details of the code developed relating to this paper. Both a console version and a Windows GUI version (with the same internal molecular dynamics methods) have been developed.

ACKNOWLEDGEMENTS

The development of the object-oriented approach to molecular dynamics was assisted by conversations with Dr. Chuck Yeung, Penn State Erie. This work was partially supported by an ARAF grant from Augustana College, through the Bush foundation.

LITERATURE CITED

Allen, M.P., and D.J. Tildesley. 1987. Computer Simulation of Liquids. Clar-endon, Oxford. 385 pp.

Hahn, M.Y., and R.L. Whetton. 1988. Rigid-Fluid Transition in Specific Size Argon Clusters. Phys. Rev. Lett. 61:1190-1193.

Jellinek, J., Ed., 1999. Theory of Atomic and Molecular Clusters: With a Glimpse at Experiments. Springer, Berlin. 429 pp.

Johnson, M.A., and M. Tilmann, Ed. 2002. Van der Waals Clusters and Their Ions. Elsevier, Amsterdam. 260 pp.


Maitland, G.C., M.Rigby, E.B.Smith, andW.A. Wakeham. 1981. Intermolecu-lar forces: their origin and determination. Clarendon, Oxford. 616 pp.

Moore, B.G., and A.A. Al-Quraishi. 2000. The Structure of Liquid Clusters of Lennard-Jones Atoms. Chemical Physics. 252:337-347.

Stallman, R. The GNU Project. http://www.gnu.org.Stepanov, A., and M. Lee. 1995. The Standard Template Library. HP Technical

Report HPL-95-11(R.1).Stroustrup, B. 1997. The C++ Programming Language. Addison-Wesley, Read-

ing, MA. 1019 pp.Verlet, L. 1967. Computer ‘Experiments’ on Classical Fluids. I. Thermody-

namical Properties of Lennard-Jones Molecules. Phys. Rev. 165:201-214.


PERCOLATION PARTITIONEDINTO PORE SIZE CLASSES

S.G. Wangemann and R.A. Kohl BIA-Yakama Agency, Toppenish, Washington

Plant Science DepartmentSouth Dakota State University

Brookings, SD 57007

ABSTRACT

The mathematical models we use to represent water potential and water flow through soil usually contain the independent variable of pore radius. We conducted an experiment to separate saturated water flux density into water flux density intervals between water potential values through a large undisturbed soil column. The measured values were compared to calculated values from the Poiseuille equation matched by the volume of drained pores. The measured flow rates were approximated by using 10-2x Poiseuille values for clean glass tubes.

Keywords

Percolation, Poiseuille equation, Pore flow.

INTRODUCTION

Water flow through soil or geologic strata can be visualized as water mov-ing through irregular spaces between primary soil particles or between the sand grains of an aquifer or water filtering system (Darcy, 1856). In addition, rapid water flow through soil is portrayed as flowing through pores left from decayed roots or worm burrows (Laws and Gilbert, 1887), or a combination of these. To mathematically represent this latter flow, models are used with the parameters of a cylindrical tube with a radius (r) or diameter (d). The word, pore itself, tends to invoke an image of a small circular opening in a surface. The terms of capillary rise or capillary potential also encourage this mental image of soil water flow by mathematically representing the flow and potential in terms of water in small cylindrical, glass capillaries. The word capillary is derived from the Latin word for hair, capilla. We mathematically represent capillary rise as:

hρgπr2 = σ2πr cos α [1]


where the symbols and their values for water at 20º C are: h = height of water rise in the capillary, cm ρ = density of water, 0.998 g cm-3

g = acceleration due to gravity, 981 ergs (g cm)-1

r = radius, cm σ = surface tension of water, 72.75 dynes cm-1

α = contact angle of the water with the surface.

Assuming a zero contact angle equation [1] reduces to the approximate equa-tion:

hr ≈ 0.15 cm2

or hd ≈ 0.30 cm2 [2] Capillary potential is equated to the height of rise, h.

Water flow through soil is often represented by a Darcy type equation:

Q = A K Δ H [3] t Δ L

where: Q = volume of water, cm3

t = time, s A = area, cm2

K = hydraulic conductivity, cm s-1

ΔH = capillary potential difference, cm ΔL = distance over which potential difference is measured, cm.

Equation [3] is used for both saturated and unsaturated water flow through soil and the hydraulic conductivity, K, is found to decrease by many orders of magnitude as the soil water potential decreases during soil drying. The hydrau-lic conductivity term embodies the reciprocal of all the factors that would slow water flow through the porous soil system. We accept the fact that soil water flow is too complicated to be represented by separate factors and their impact individually; thus K is a composite term determined experimentally for the specific conditions at the time of measurement. Also inherent in this equation is the fact that area, A, is the cross sectional area of the total mass of soil which includes the pores and not the flow area by itself. Thus, while K carries the units of velocity and would equal a mean flow rate if the potential gradient is due to gravity and straight down and equal to one, the actual flow velocity through the pores will be considerably greater than K as a majority of the area is occupied by soil solids and not permeable. In addition, if flow is much faster in larger pores that the smaller ones then the mean flow velocity would be some kind of average of a wide range of values. Under field conditions there are no natural boundaries as exists with water flow through pipes. Therefore, the Darcy equation is often written for water flow per unit area and the left side of the equation is termed water flux density (Jw).


Jw = Q = K ΔΨ [4] A t ΔL

Mean pore velocity, vp, is then given by (Biggar and Nielsen, 1980)

vp = Jw [5] Θv

where: Θv = Volumetric water content.

Tracers and soluble chemicals appear in leaching water much sooner than predicted by displacement flow. This rapid emergence is credited to macropore flow, rapid flow through the largest pores (Thomas and Phillips, 1979; Biggar and Nielsen, 1980; Bouma et al., 1983; White, 1985; and Gentry et al., 2000). As our attention becomes focused on pore flow, which we visualize as water flowing through cylindrical pores, then the Poiseuille equation written for water flow in cylindrical tubes seems appropriate.It is written as:

Q = gπ ρω r4 ΔH [6] 8 η L

where: ρw= density of water

η = viscosity of water (poise or dyne s cm-2)

L = length of the tube.

The Poiseuille equation is the standard used for viscosity measurements, thus the unit of poise for viscosity in the International System of Units. An added incentive for considering the Poiseuille equation with its viscosity term, are the studies which show water flow into soil varying on a diurnal cycle, mimicking changes of viscosity resulting from changes in temperature (Jaynes 1990, Musgrave 1955). The Poiseuille equation has also been used to represent the flow of water through xylem vessels of plants with moderate success. Published data resulted in measured xylem flow rates of 50% of the values calculated by the Poiseuille equation (Calkin et al. 1986), to 20% (Lewis and Boose 1995, Martre et al. 2001) and to 5% (Yamauchi et al. 1995). The Poiseuille equation was also used to represent water flow through ben-tonite and bentonite-sand mixtures for hazardous waste containment and stated to be within a half-order of magnitude of calculated values (Dixon et al. 1999). However, the pore sizes were measured by the mercury porosemetry method which leaves some doubt if the pore size values used were appropriate (Diamond 2000).


Our objective was to partition water flow through a large soil column into various pore size classes to determine if the Poiseuille pattern, where tube diam-eter is raised to the fourth power, would be followed.

METHODS

An undisturbed, round, grass covered soil core 0.6 m in diameter and 0.8 m high was carefully excavated in the field and banded with 20 mm wide metal banding at 0.15 m intervals to avoid cracking along natural cleavage lines as lateral pressure was being released. A 0.7 m diameter steel sleeve 0.9 m tall was placed around the core and the free space filled with urethane foam to maintain lateral pressure and provide water sealing. A steel plate was forced laterally under the column and used to lift and transport the column to a greenhouse where the column was set vertically. The base plates were sealed to the steel sleeve, and access holes were drilled for air entry and water removal from the base. Percolat-ing water was collected from the hole at the base of the column with a vacuum system. Percolation measurements were made into the grass covered Fordville loam (fine-loamy over sandy or sandy-skeletal, mixed, superactive, frigid Pachic Hap-ludoll). The soil consisted of about 0.65 m of loamy alluvium over sandy-skeletal glacial outwash. The soil bulk density of Fordville increased from 1.1 gm cm-3 at the 0.1 m depth to 1.5 gm cm-3 at the 0.5 m depth. At each depth of 140, 303, and 447 mm, four small (10 mm diameter by 30 mm long) tensiometer cups attached to water columns were installed through the sides of the column. These tensiometers were used to monitor soil water po-tential. Tension could be read to +/- 5 mm of water under the conditions in the column. When a unit water potential gradient (ΔΨ/ΔL=1)is achieved for equation [4] with a constant water application rate to the soil surface, the hydraulic conduc-tivity will be equal to the water flux density (Jw=K), (Hanks and Ashcroft, 1980 p 65-69). The apparatus shown in Figure 1 was constructed and used to achieve constant water application rates to the soil surface to determine percolation rates at specific soil water potentials. A reservoir (B) supplied a rotating water applicator head (H) fitted with calibrated syringe needles as drip nozzles. Varying needle sizes allowed for dif-ferent application rates to be uniformly applied to the soil surface. Equilibrium at a given application rate was reached when the matric potential at the different depth increments became constant. This resulted in a unit hydraulic gradient and equilibrium values for the data are plotted in Figure 2. A continuous water application of approximately 1.5 mm/h was maintained for 300 hours immediately prior to data collection to wet the the soil without causing pore blocking by air bubbles (Wangeman et al., 2000). Matric potential at depths of 140, 303, and 447 mm was measured during application rates from 0.7 mm/h to 39.5 mm/h over a further 17-day period of continuous water ap-plication.


The minimum application rate was based on the smallest needle nozzles that we had available. The application rate of 39.5 mm/h was not exceeded because the installed tensiometers reached an average matric potential value of approxi-mately -13 mm of water at this rate. Accurate matric potential values less nega-tive than -10 mm of water were not obtainable because the actual variability of the readings was about +/- 5 mm of water at a specific depth. The same value of water potential was obtained at a given water flow rate

Figure 1. Unsaturated-flow water application system.


through the column regardless of direction of change in the flow rate, thus hys-teresis did not affect our data collection.


Under ponded conditions, prior to installing the constant rate application unit, a long-term (12 to 47 h) infiltration rate of 61 mm/h was measured with a 20 mm head on the soil surface. From this the saturated hydraulic conductivity was estimated to be 59.5 mm/h by the Darcy Equation. Table 1 was developed to illustrate the water flow rate distribution through the soil column as a function of pore size. The regression equation developed from the data in Figure 2 was used to partition the flow rates between matric potential limits (column 1). Pore diameter (column 2) was calculated from the capillary rise equation assuming a zero contact angle. The flow rates listed in column 3 represent the flow through the pores for the largest pore size for this range and for each smaller pore size class. As lower matric potentials become the upper limit of the potential range, the larger pore sizes aerate and only smaller pore sizes are considered to be conducting.

Figure 2. Matric potentials at equilibrium with different constant flow rates through the soil column.


Table 1. Soil water flow rates through the soil column partitioned into matric potential ranges and associated capillary rise pore diameters.

Matric Potential Range mmwater (1)

Pore Diameter Range mm (2)

Total Flow Rate

mm h-1 (3)

Flow Rate by Size Class mm h-1 (4)

Flow Rate % (5)

Sum of Flow Rate

% (6)

flooded to -13 >2.3 59.5 20.0 33.6 33.6 -13 to -50 2.3 to 0.6 39.5 13.1 22.0 55.6

-50 to -100 0.6 to 0.3 26.4 12.0 20.2 75.8 -100 to -150 0.3 to 0.2 14.4 6.6 11.1 86.9 -150 to -200 0.2 to 0.15 7.8 3.5 5.9 92.8 -200 to -250 0.15 to 0.12 4.3 2.0 3.4 96.2 -250 to -300 0.12 to 0.10 2.3 1.0 1.7 97.9 -300 to -350 0.10 to 0.086 1.3 0.6 1.0 98.9 -350 to -400 0.086 to 0.075 0.7 0.3 0.5 99.4

The fourth column is the flow rate difference between the preceding rows and which illustrates the flow rate ascribed to a pore size range. The first dif-ferential flow rate represents the difference between the slightly ponded rate and the near zero matric potential flow rate which would include the largest pores available for flow. The remaining differential flow rates follow. The fifth column converts column four to percentages and the sixth column incrementally sums these percentages. Thus, 75% of the ponded flow through this soil profile can occur through pores with a diameter equal to or greater than 0.3 mm in diameter and a water potential of >-0.01 bar. The final value suggests that over 99% of gravity flow through this soil occurs through pores larger than 0.0075 mm in diameter and approximates a field capacity using a water potential of -0.04 bars. These data support two important concepts associated with soil water. First the large pores, those large enough to be seen with the unaided eye (>0.1 mm diameter), are associated with infiltration rates comparable to higher rainfall intensities on the order of 50 mm/h or more. Surface sealing on soil exposed to rainfall impact can obstruct these large pores, reducing infiltration rates and enhancing runoff and erosion. While we have observed this phenomenon in the past, the above table lends additional support to the validity of this concept. Secondly, rapid drainage and soil aeration are dependent on the large pores. These large pores are also structurally the weakest and are the first to collapse un-der compactive weights (Eriksson et al. 1974). Tire compaction eliminates these large pores at the surface. Wheel track inter-rows are often seen with standing water after a rain, while the non-wheel rows do not have standing water. How-ever, more insidious is the compaction below normal tillage depths caused by very heavy equipment (combines, grain carts, liquid manure tanks, etc.) passing over wet soil profiles (Hakansson and Petelkau 1994) . The collapse of the larger


pores, even to depths of one meter, hinders drainage, impedes root growth, and enhances the environment for denitrification (Eriksson et al. 1974). Table 2 is included to lend some support to the Poiseuille Equation type of model for soil water flow. A simplifying assumption was made that all of the flow through the soil profile for a matric potential range occurred through the largest pore size in that pore size class. Since the Poiseuille equation is dependent on the pore diameter raised to the fourth power, the larger pores in a class would carry most of the flow. These data were compared to calculated values using the Poiseuille Equation and found to be about 1% of the calculated values. Assum-ing that the smaller pores conducting water in this study follow a similar pattern, 1% of the calculated Poiseuille values are listed in column 3 of Table 2.

Table 2. Conducting pore volumes associated with the different flow rates through the profile. Fordville loam (fine-loamy over sandy or sandy-skeletal, mixed, superactive, frigid Pachic Hapludoll).

Matric Po-tential Range mm water (1)

PoreRadius mm (2)

Adjusted Poi-seuille Flow

RateA mm cm2 h-1(3)

PoresB # cm-2 for Observed Flow Rate(4)

Conducting Pore Volume

% (5)

Cumulative Conducting Pore Volume

% (6)

0 to -13 3 112,000 0.0002 0.01 0.01 -13 to -50 1.15 2422 0.005 0.03 0.04 -50 to -100 .3 11.2 1.07 0.30 0.34

-100 to -150 .15 0.70 9.4 0.66 1.00 -150 to -200 .10 0.14 25 0.80 1.80 -200 to -250 .075 0.044 46 0.81 2.61 -250 to -300 .06 0.018 56 0.63 3.24 -300 to -350 .05 0.0087 69 0.54 3.78 -350 to -400 .043 0.0047 63 0.37 4.15

A The values in column 3 are one percent of the flow rate calculated by the Poiseuille Equation.B Based on dividing the observed flow (Table 1) by the adjusted Poiseuille flow rate for a given matric potential range.

The fourth column contains the number of pores needed to carry the water in this matric potential range. The fifth column contains the values of the pore volume conducting the water in this potential range. Filling out the remainder of the table, we arrive at a total conducting volume slightly more than 4% for the sum of the volume of all pores that would drain down to a matric potential of -400 mm of water. This comes close to the drainable porosity of about 5% for the soil samples taken from these depths as measured on an Eijkelkamp ten-sion table. Using a value of 1% of the Poiseuille flow rate for the soil water flow rates through the pores may be fortuitous in producing a drained pore volume so close to the 5% drainable volume from the tension since using 2% instead of 1% would have halved the draining pore volume. And using half the value (ie. 0.5%) would have doubled the drainage volume compared to the 1% value.


When comparing this 1% value for soil with the 5 to 50% values for xylem measurements we need to remember that xylem tube cross sections, tube length, and flow rates can be reasonably measured and a direct comparison with the Poiseuille Equation can be made. Soil pore sizes are estimated from the capillary rise equation [2] and pore continuity, pore tortuosity, and pore wall roughness are unknowns. Even so, the flow estimates and the drained pore volume succes-sion appear to follow a Poiseuille Equation concept with pore radius to a power of four.

CONCLUSIONS

When we consider the mathematical representations that we use for water flow through soil, they are dominated by the concept of cylindrical conducts. As such, the Poiseuille equation should logically be considered though little material has been published considering its use. The experimental results published in this paper does follow a Poiseuille equation spectrum when the Poiseuille values for clean glass tubes are multiplied by 10-2. Since pore radius is raised to the fourth power in the Poiseuille equation small irregularities in pore diameter or pore restrictions would cause large differences in flow rates between soil pores and glass tubes.

LITERATURE CITED

Biggar, J.W., and D.R. Nielsen. 1980. Mechanisms of chemical movement in soils. p. 213-227. In. A. Banin and U. Kafkafi (eds.), Agrochemicals in soils. Pergamon, Oxford. UK

Bouma, J., C. Belmans, L.W. Dekker, and W.J.M. Jeurissen. 1983. Assessing the suitability of soils with macropores for subsurface liquid waste disposal. J. Environ. Qual. 12:305-311.

Calkin, H.W., A.C. Gibson, and P.S. Nobel. 1986. Biophysical model of x y-lem conductance in tracheids of the fern Pteris vittata. J. Exp. Bot. 37:1054-1064.

Darcy, H. 1856. “Les Fontaines Publique de la Ville de Dijon.” Dalmont, Paris. as cited in M. Muskat. 1946. The Flow of Homogeneous Fluids Through Porous Media. p. 55-120.

Diamond, S. 2000. Mercury porosimetry; An inappropriate method for the measurement of pore size distributions in cement-based materials. Cement and Concrete Res. 30:1517-1525.

Dixon, D.A., J. Graham, and M.N. Gray. 1999. Hydraulic conductivity of clays in confined tests under low hydraulic gradients. Can. Geotech. J. 36:815-825.

Eriksson, J., I. Hakansson, and B. Danfors. 1974. The effect of soil compaction on soil structure and crop yields. Swedish Institute of Agricultural Engi-neering. Bul. 34. English Translation by J. K. Aase, UADA-Agricultural Research Service. 101pp.


Gentry, L.E., M.B. David, K.M. Smith-Starks, and D.A. Kovacic. 2000. Ni-trogen fertilizer and hericide transport from tile drained fields. J. Environ. Qual. 29:232-240.

Hakansson, I. and H. Petelkau. 1994. Benefits of limited axle load. p.479-499.In B.D. Soane and C. van Ouwerkerk (eds.) Soil compaction in crop pro-duction. Elsevier Science BV.

Hanks, R.J. and G.L. Ashcroft. 1980. Applied Soil Physics. Springer-Verlag. New York. p. 65-69.

Jaynes, D.B. 1990. Temperature variations effect on field-measured infiltration. Soil Sci. Soc. Am. J. 54:305-312.

Lawes, J.B., J.H. Gilbert, and R. Wattington. 1882. On the amount and com-position of the rain and drainage water collected at Rothamstead. Williams Clowes and Sons LTD., London, England. p.169 In. J. B. Lawes and J. H. Gilbert. Rothamsted Memoirs Vol.V. Field Experiments &c &c.

Lewis, A.M., and E.R. Boose. 1995. Estimating volume flow rates through xylem conduits. Am. J. Bot. 82:1112-1116.

Martre, P., H. Cochard, and J.L. Durand. 2001. Hydraulic architecture and water flow in growing grass tillers (Festuca arundinacea Schreb.). Plant Cell and Environ. 24:65-76.

Musgrave, G.W. 1955. How much of the rain enters the soil. p. 151-159. In Water. The Yearbook of Agriculture. The United States Department of Ag-riculture. U. S. Gov. Printing Office. Washington 25, D. C.

Thomas, G.W. and R.E. Phillips. 1979. Conseqences of water movement in macropores. J. Environ. Qual. 8:149-152.

Wangemann, S.G., R.A. Kohl, and P.A. Molumeli. 2000. Infiltration and per-colation influenced by antecedent soil water content and air entrapment. Transactions of the ASAE 43(6):1517-1523.

White, R.E. 1985. The influence of macropores on the transport of dissolved and suspended matter through soil. p. 95-120. In B.A. Stewart (ed.) Ad-vances in Soil Science. Vol.3 Springler-Verlag, New York.

Yamauchi, A., H.M. Taylor, D.R. Upchurch, and B.L. McMichael. 1995. Axial resistance to water flow of intact cotton taproots. Agron. J. 8 7 : 4 3 9-445.


ROAD CULVERTS ACROSS STREAMSWITH THE ENDANGERED TOPEKA SHINER,

NOTROPIS TOPEKA, IN THE JAMES,VERMILLION, AND BIG SIOUX RIVER BASINS.

Steven S. WallDepartment of Wildlife and Fisheries Science


605-688-5124

Charles R. Berry, Jr.U.S. Geological Survey

South Dakota Cooperative Fish and Wildlife Research UnitSouth Dakota State University

Brookings, SD 57007

ABSTRACT

We evaluated 232 installed corrugated pipe culverts at 81 sites where roads cross streams that have a high potential for Topeka shiner presence. Culvert con-ditions were characterized by the amount of perching, embeddedness, blockage, gradient and water velocity, and rated for potential as a barrier to upstream fish migration. Seven sites were classified as high priority for maintenance or mitiga-tion, 22 were classified as medium priority for mitigation, and 52 were classified as low priority for mitigation. The data allows the Department of Transportation to plan for road crossing maintenance and the conservation of the rare species and associated stream habitat.

Keywords

Stream, culvert, Topeka shiner, Notropis topeka, road crossings, barriers, highway, corrugated culvert, stream habitat, fish migration, barrier

INTRODUCTION

Road construction plans for stream crossings usually include manage-ment practices that conserve the stream habitat and biota. A free-span bridge eliminates problems associated with culverts (Clay 1995), but culverts are more economical for crossing small streams. During culvert installation, conservation practices protect stream habitat, but maintenance is needed afterwards because culverts can become a barrier to fish migration if they become blocked, perched, or inadequate for flows (Gebhards and Fisher 1972, Lauman 1976, WDFW 2001).


Culverts become inadequate for stream discharge when discharge increases because of land use changes upstream (e.g., wetland drainage, urbanization). High flows can be a velocity barrier to fish migration when velocity through long culverts exceeds fish swimming ability (Lauman 1976, Adams et al. 2000). During low flows the culvert may have inadequate water depths for fish passage, or may be blocked by debris. A perched culvert develops when scouring at the downstream end causes channel incision and a vertical water drop from the cul-vert to the stream (Gebhards and Fisher 1972). The listing of the Topeka shiner (Notropis topeka) as endangered in 1999 has caused highway departments to review the impacts of construction activities on this rare fish. Topeka shiners inhabit small tributaries to the James, Vermillion, and Big Sioux rivers (Wall et al., 2001). These tributaries are crossed many times by roads over concrete box culverts or corrugated steel pipe culverts that may influence Topeka shiner movements. Topeka shiners undertake localized movements before spawning from mid-May to early August (Kerns 1983, Barber 1986, Stark et al. 1999) and in response to flow variation (Layher 1993, Dahle 2001, Mammoliti 1994). Topeka shiners are small minnows (family Cyprinidae) that reach 6 cm in length (Blausey 2001). Adults can swim up to about 35 cm/second (Adams et al. 2000) but may be blocked by small waterfalls that require leaping. Our objectives were to assess the condition of culverts at road crossings over streams that were suitable habitat for the Topeka shiner, and to prioritize crossing sites for mitigation or maintenance.

METHODS

Stream and road maps were overlaid for each 11-digit Hydrological Unit watershed in the James, Vermillion, and Big Sioux basins where Topeka shin-ers have been found or are predicted because of the presence of suitable habitat (Wall et al. 2001). We focused on crossings with corrugated culverts and stream segments with a high or moderate potential for Topeka shiner presence because steel corrugated culverts have the most impact on fish migration in small streams (Warren and Pardew 1998). Road crossings were given high, medium or low priority for maintenance by combining a culvert condition index with a stream habitat suitability index (Table 1).

Culvert Condition Indices

We measured culvert slope, length, perch, counter sink, bed width, and wa-ter velocity (Figure 1) as well as many other measures reported elsewhere (Wall and Berry 2002). We indexed culvert condition by scoring the five features for their potential impact on Topeka shiner movement and summed the scores (Table 2). An index of < 6 indicated a poor condition, whereas a score of > 8 indicated a good condition (Table 1). Culvert features were scored as follows.


Table 1. Culvert condition and habitat indices and method of combining two indices to determine maintenance priorities.

CULVERT CONDITION INDEX HABITAT SUITABILITY INDEX

Index score Condition Index score Suitability MaintenancePriority

< 6 Poor ≥ 6 High High≥ 6 and < 8 Medium ≥ 6 High Moderate

≥ 8 Good ≥ 6 High Low≥ 8 Good < 6 Low Low

≥ 6 and < 8 Medium < 6 Low Low< 6 Poor < 6 Low Low

Perch: We speculated that a perch > 6 cm might block Topeka shiner migration. Using seasonal stage data for a wet year and dry year for a stream with Topeka shiners, we calculated that a perch of 6 cm would not occur during wet years but would occur 87% of the time during dry years (Wall and Berry 2002). So, a culvert perched > 6 cm at either the upstream or downstream end scored zero; one perched < 6 cm was scored 2. Sites with multiple culverts are planned with

Figure 1. Schematic of a corrugated pipe highway culvert showing long and cross sec-tions, and measurements to index culvert condition in relation to fish passage.


one or more embedded culverts to convey low flows, so we used the best culvert score to represent the crossing. Embeddedness: A culvert that is embedded 30% has little impact on stream hydrology (WDFW 2001), therefore a culvert embedded ≥ 30% scored 2; one embedded 30 - 15% scored 1.5; one embedded 1 - 15% scored 1; and one not embedded scored zero. Scores for each end of the culvert were averaged.

Table 2. Culvert condition index scores for seven sites classified as high priority for mitigation. All sites had culverts in poor condition (i.e. scores < 6). Site identification number refers to sites in the Big Sioux (BS), James (JR), and Vermillion (VR) river basins (locations given in Wall and Berry, 2002).

Site (ID) Culvert (#)

Perch (m)

Embeded (%)

Gradient (cm/m)

Velocity (cm/s)

Blocked (Yes/No)

Score(Index)

BS007 1 0.54 0.0 0.5 52 no

0 0 2 0 2 4

2 0.25 9.7 0.7 59 no

0 1 2 0 2 5

3 0.34 0.0 0.2 58 no

0 0 2 0 2 4

BS018 1 0.50 3.5 1.2 73 no

0 1 2 0 2 5

2 0.50 7.0 1.1 33 yes

0 1 2 0 0 3

JR019 1 0.09 0.0 -0.4 0 yes

0 0 2 2 0 4

2 0.11 22.4 0.2 0 yes

0 1.5 2 2 0 5.5

3 0.09 0.0 -1.0 0 yes

0 0 2 2 0 4

4 0.07 0.0 -0.2 0 yes

0 0 2 2 0 4

JR047 1 0.85 0.0 0.4 56 no

0 0 2 0 2 4

VR003 1 0.16 0.4 0.0 0 yes

0 0.5 2 2 0 4.5

VR005 1 0.24 0.0 *** 0 yes

0 0 2 2 0 4

2 0.23 0.0 *** 0 yes

0 0 2 2 0 4

VR0111 * * * * * no culvert 01At VR011 there was no culvert in place and stream was blocked by the road


Blockage: A culvert was scored zero if either the upstream or downstream end was completely blocked by material without spaces allowing fish passage. Open or partially blocked culverts scored 2. Velocity: We measured velocity with a current meter. The culvert scored 2 if the mean velocity was ≤ 35 cm/s, which is the sustained swimming performance for Topeka shiners (Adams et al. 2000). Scores were zero if velocity was > 35 cm/s. However, a score of 2 was given when velocity was > 35 cm/s in early spring (i.e. April to May 6) but we expected velocity to be < 35 cm/sec by the mid-May to August spawning season.Gradient: Surveying equipment was used to determine slope. A culvert scored 2 if gradient was ≤ 3 cm/m, which is equivalent to a low stream gradient (WDFW 2001); scores were zero for gradients > 3 cm/m.

Stream Reach Indices

We assessed stream habitat features up and downstream from crossings using standard methods (Wall et al., 2001). Habitat features that are associated with Topeka shiner presence included: low bank height, low bank incision, streambed substrate materials of fine gravel to cobble, presence of pool habitat and sub-merged macrophytes, riparian zones with grass or pasture and low livestock use, and overhanging vegetation (other than trees) along the stream bank (Blausey 2001). We recorded whether each physical habitat feature was present (score = 1) or absent (score = 0) based on planned criteria (Wall and Berry 2002). Upstream and downstream habitat scores at each site were summed (maximum score of 18 was possible), and the average score was assigned to the site (Table 3). Maximum score was 9 for sites with all desirable habitat features. We classified crossings with scores ≥ 6 as good habitat for Topeka shiners, and crossings with scores < 6 were classified as having low habitat suitability.

RESULTS

We surveyed 36 watersheds with Topeka shiners. In 25 watersheds we found 81 crossings over corrugated steel culverts (n = 232) in reaches of high or moder-ate Topeka shiner habitat. Of the 81 crossings, 28 had good culvert conditions, 40 had medium conditions, and 13 were in poor condition. Good stream habi-tat surrounded 49 crossings, whereas 32 sites had habitat features that were not indicative of Topeka shiner presence. Combined culvert and stream reach index scores classified seven crossings as high priority sites for mitigation (Table 2, Table 3). High priority sites had poor culvert conditions with scores ranging from 3 to 5.5, and reach habitat associated with Topeka shiner presence (scores = 6 and 7). Of the remaining sites, 22 were classified as medium priority for mitigation, and 52 were classified as low priority for mitigation according to combined index scores shown in Table 1. An example of a high priority mitigation crossing (culvert index = 5) is shown in Figure 2. The culverts are perched 0.25 m above stream level (score = 0). The culvert had no blockage (score = 2) but there was no streambed mate-

130 Proceedings of the South Dakota Academy of Science, Vol. 83 (2004)Ta

ble

3. S

trea

m r

each

inde

x sc

ores

for

str

eam

hab

itat

mea

sure

s at

sev

en s

ites

cla

ssifi

ed a

s hi

gh p

rior

ity

for

mit

igat

ion.

All

site

s ha

d ha

bita

t in

dica

-ti

ve o

f To

peka

shi

ner

pres

ence

(i.e

. sco

res

≥ 6)

. US

= U

pstr

eam

fro

m c

ross

ing,

DS

= d

owns

trea

m f

rom

cro

ssin

g.

Site

IDU

S Be

d M

atD

S Be

d M

atU

S H

d C

utD

S H

d C

utU

S Lv

Stk

Use

DS

LvSt

k U

seU

S O

H

Veg

DS

OH

Ve

gU

S Tr

eeD

S Tr

eeU

S Ri

p Zo

neD

S Ri

p Zo

ne

BS00

7SI

SIlo

wye

slo

wlo

wye

sye

sye

sye

sye

sye

s

00

10

11

11

00

11

BS01

8SI

/CB

SI/C

Bye

slo

whi

ghm

edno

yes

yes

noye

sye

s

11

01

00.

50

10

11

1JR

019

SI /

SASA

/ CL

noye

sm

edm

edye

sye

sno

noye

sye

s

00

10

0.5

0.5

11

11

11

JR04

7SA

/CB

SA/C

Bye

sye

sm

edm

edye

sye

sno

yes

yes

yes

1

10

00.

50.

51

11

01

1VR

003

SISI

nono

med

low

yes

yes

yes

yes

yes

yes

0

01

10.

51

11

00

11

VR00

5G

rass

BD/S

I/CB

nono

low

high

yes

noye

sno

yes

yes

0

11

11

01

00

11

1VR

011

FG/C

BFG

/CB

noye

sm

edhi

ghye

sye

sno

noye

sye

s

11

10

0.5

01

11

11

11

Bed

Mat

= b

ed m

ater

ial,

Hd

Cut

= H

ead

cutt

ing,

LvS

tk U

se =

liv

esto

ck u

se,

OH

Veg

= o

verh

angi

ng v

eget

atio

n, R

ip Z

one

= r

ipar

ian

zone

con

-si

stin

g of

gra

sses

and

for

bes,

Sub

Mac

= s

ubm

erge

d m

acro

phyt

es,

Bnk

Ht

= b

ank

heig

ht,

Bnk

Inc

is =

ban

k in

cisi

on,

USS

= u

pstr

eam

rea

ch s

core

, D

SS =

dow

nstr

eam

rea

ch s

core

, AVS

= m

ean

reac

h sc

ore

(rou

nded

), C

B =

cob

ble,

BD

= b

ould

er, S

I =

silt

, CL

= c

lay,

FG

= fi

ne g

rave

l.


Tabl

e 3

cont

inue

d. S

trea

m r

each

ind

ex s

core

s fo

r st

ream

hab

itat

mea

sure

s at

sev

en s

ites

cla

ssifi

ed a

s hi

gh p

rior

ity

for

mit

igat

ion.

A

ll si

tes

had

habi

tat

indi

cati

ve o

f To

peka

shi

ner

pres

ence

(i.e

. sco

res

≥ 6)

. US

= U

pstr

eam

fro

m c

ross

ing,

DS

= d

owns

trea

m f

rom

cro

ssin

g.

Site

s ID

US

Sub

Mac

DS

Sub

Mac

US

Bnk

Ht

DS

Bnk

Ht

US

Bnk

Inci

sD

S Bn

k In

cis

US

pool

sD

S po

ols

USS

DSS

AVS

BS00

7ye

sye

slo

wlo

wlo

wlo

wno

no

1

11

01

10

07

56

BS01

8ye

sno

med

med

med

med

noye

s

1

00.

50.

50.

50.

50

14

7.5

6JR

019

yes

yes

low

med

low

med

nono

11

10.

51

0.5

00

7.5

5.5

7JR

047

nono

med

med

med

med

noye

s

0

00.

50.

50.

50.

50

15.

55.

56

VR00

3no

nolo

wlo

wlo

wlo

wno

no

0

01

11

10

05.

56

6VR

005

nono

low

med

low

low

nono

00

10.

51

10

06

5.5

6VR

011

nono

low

low

low

low

yes

yes

00

11

11

11

7.5

67

1 B

ed M

at =

bed

mat

eria

l, H

d C

ut =

Hea

d cu

ttin

g, L

vStk

Use

= l

ives

tock

use

, O

H V

eg =

ove

rhan

ging

veg

etat

ion,

Rip

Zon

e =

rip

aria

n zo

ne c

on-

sist

ing

of g

rass

es a

nd f

orbe

s, S

ub M

ac =

sub

mer

ged

mac

roph

ytes

, B

nk H

t =

ban

k he

ight

, B

nk I

ncis

= b

ank

inci

sion

, U

SS =

ups

trea

m r

each

sco

re,

DSS

= d

owns

trea

m r

each

sco

re, A

VS =

mea

n re

ach

scor

e (r

ound

ed),

CB

= c

obbl

e, B

D =

bou

lder

, SI

= s

ilt, C

L =

cla

y, F

G =

fine

gra

vel.


rial (low bed width) in the culvert. The velocity (52-59 cm/s) exceeded Topeka shiner swimming ability (score = 0).

DISCUSSION

We chose measurements of culvert conditions that others have suggested were important (McKinnon and Hnytka 1985, Belford and Gould 1989, War-ren and Pardew 1998, Toepfer et al. 1999, WDFW 2001, Wellman et al. 2000). We introduced the scoring and index protocol to synthesize the many parameters we measured for culverts and stream habitat to create a decision support pro-tocol to help prioritize sites for inspection by highway engineers or for possible maintenance. All crossings would require some degree of mitigation according to our classification system (i.e. all culverts scored < 10). We prioritized the cross-ings so that maintenance might start at the high priority sites using guidelines for proper culvert placement in general (Clay 1995, Odeh 1999) and in South Dakota specifically (Cunningham 2000, NRCS 2001). The problem of passage by small fishes in low gradient streams has not been given much attention. Movement of small fishes (ie., Notropis sp.) through properly installed box culverts was comparable with or higher than movement through natural stream reaches, whereas corrugated culverts restricted movement somewhat (Warren and Pardew 1998). Two general conclusions have been made for insuring passage of juvenile salmonids: 1) “the low velocities and level of turbulence required for passage of juvenile fish are so low, they are impractical to

Figure 2. Downstream view of a stream crossing that is a high priority for maintenance because of excessive perch (> 0.2 m) and high water velocities (> 0.35 m/s) that present a potential passage problem for Topeka shiners.


achieve in design,” and 2) “if juvenile passage is desired, it is recommended that a natural channel be built within the culvert” (Bates 2001). When mitigating or replacing culverts, construction should minimize im-pacts to the stream and thus the fish community. Standard best management practices need to be clearly stated and planned before road construction begins. Erosion control is important, especially during the spawning season, because sedimentation can reduce the reproductive success of the Topeka shiner and may have contributed to its extirpation in some parts of its range (Minckley and Cross 1959, Tabor 1998). At a crossing with multiple culverts, one culvert should be embedded in the streambed to allow fish passage during low flows. There are many streams with Topeka shiners in South Dakota (Wall et al., 2001), so it is important that guidelines and best management practices for road construction at stream crossings be practiced and improved. The Topeka shiner is a pioneering species (Blausey 2001) that can be found in intermittent streams (Minckley and Cross 1959, Barber 1986, Dahle 2001). Thus there are times (e.g., drought, late summer) when a stream segment clas-sified as good habitat may be dry at a road crossing where a culvert needs to be mitigated. Current guidelines prohibit in-stream disturbances during the spawning season, which is late-May to early August in headwater streams of the Big Sioux River in Minnesota (Hatch 2001). Dry conditions at road crossings provide the opportunity to maintain culverts with minimum impact on stream habitat. The Endangered Species Act regulations have caused us to examine our management of streams. The willingness of the South Dakota Department of Transportation to take partial responsibility for fish conservation is an important aspect of South Dakota’s conservation plan for this species (Shearer 2003). As construction goes on at dozens of road crossings in the future, an effort will be made to conserve this rare fish and in doing so, conserve stream habitat for all species, including landowners. Erosion, sedimentation, improper installation of culverts or lack of maintenance of road crossings are not only problems for the fish community, but also for landowners upstream because of flooding and downstream because of channel incision and bank sloughing.

ACKNOWLEDGEMENTS

The South Dakota Cooperative Fish and Wildlife Research Unit, is sup-ported by the US Geological Survey, South Dakota Department of Game, Fish and Parks, U. S. Fish and Wildlife Service, South Dakota State University, and Wildlife Management Institute. The South Dakota Department of Transporta-tion provided funding. Jeff Shearer, B. Kopplin and Nathan Harris helped with fieldwork, and S. Kopplin helped with the final report design and map making.


LITERATURE CITED

Adams, S.R., J.J. Hoover, and K.J. Killgore. 2000. Swimming performance of the Topeka shiner (Notropis topeka) an endangered Midwestern minnow. American Midland Naturalist 144: 178-186.

Barber, J.A. 1986. Ecology of Topeka shiners in Flint Hills streams. M.S. Thesis, Emporia State University, Emporia, Kansas. 86 pp.

Bates, K. 2001. Fish passage design at road culverts: a design manual for fish passage at road crossings. Washington Department of Fish and Wildlife, Environmental Engineering Division. Olympia, Washington, 41 pp.

Belford, D.A., and W.R. Gould. 1989. An evaluation of trout passage through six highway culverts in Montana. North American Journal of Fisheries Management 9: 437 – 445.

Blausey, C.M. 2001. The status and distribution of the Topeka shiner Notropis topeka in eastern South Dakota. M.S. Thesis, South Dakota State University, Brookings. 112 pp.

Clay, C.H. 1995. Design of Fishways and Other Fish Facilities. Second edition. Lewis Publishers, Ann Arbor. 248pp.

Cunningham, G.R. 2000. Road and bridge construction best management practices for stream sites inhabited by Notropis topeka (Topeka shiner). Eco-Centrics, Omaha, NE. 21 pp plus appendices.

Gebhards, S. and J. Fisher. 1972. Fish passage and culvert installations. Idaho Fish and Game Department. 12pp.

Hatch, J.T. 2001. What we know about Minnesota’s first endangered fish species: The Topeka shiner. Journal of the Minnesota Academy of Science 65(1): 39-46.

Kerns, H.A. 1983. Notropis topeka in Kansas: Distribution, habitat, life history. Kansas Fish & Game Commission, Non-game Program, final report.

Lauman, J.E. 1976. Salmonid passage at stream crossings: A report with de-partmental standards for passage of salmonids. Environmantal Manage-ment Section, Oregon Department of Fish and Wildlife, Portland Oregon. 78pp.

Layher, W.G. 1993. Changes in fish community structure resulting from a flood control dam in a Flint Hills stream, Kansas, with emphasis on the Topeka shiner. University of Arkansas at Pine Bluff. Cooperative Fisheries Research Project AFC-93-1. 30pp.

Mammoliti, C. S. 1994. The effects of watershed impoundments on the Topeka shiner (Notropis topeka). Kansas Department of Wildlife and Parks, Pratt, KS. Unpublished Report. 13pp.

McKinnon, G.A. and F.N. Hnytka. 1985. Fish passage assessment of culverts constructed to simulate stream conditions on Liard River tributaries. Ca-nadian technical report of fisheries and aquatic sciences No. 1255. Western Region, Department of Fisheries and Oceans, Winnipeg, MB. 121 pp.

Minckley, W.L. and F.B. Cross. 1959. Distribution, habitat, and abundance of the Topeka shiner, Notropis topeka (Gilbert), in Kansas. American Midland Naturalist 61: 210-217.


NRCS (Natural Resources Conservation Service). 2001. Fish Passage. Conser-vation Practice Standard Code 396, NRCS, Huron, SD. 3 pp.

Odeh, M. editor, 1999. Innovations in Fish Passage Technology. American Fisheries Society, Bethesda, Maryland. 212pp.

Shearer, J. 2003. Conservation plan for the Topeka shiner. South Dakota De-partment of Game, Fish and Parks, Pierre, 34 pp.

Stark, W., J. Luginbill, and M.E. Eberle. 1999. The status of the Topeka shiner (Notropis topeka) in Willow Creek, Wallace County, Kansas. Unpublished report, Dept. of Biological Sciences, Fort Hays State University, Hays, Kan-sas. 13 pp.

Tabor, V.M. 1998. Final rule to list the Topeka shiner as endangered. Federal Register 63 (240): 69008-69021.

Toepfer, C.S., W.L. Fisher, and J.A. Haubelt. 1999. Swimming performance of the threatened leopard darter in relation to road culverts. Transactions of the American Fisheries Society 128: 155 – 161.

Wall, S.S., C.M. Blausey, J.A. Jenks and C.R. Berry Jr. 2001. Topeka shiner (Notropis topeka) population status and habitat conditions in South Dakota. Final report for research work order # 73. South Dakota State University, Brookings, SD. 300 pp.

Wall, S.S. and C.R. Berry Jr. 2002. Inventory and mitigation of culverts crossing streams inhabited by the Topeka shiner (Notropis topeka) in South Dakota. Final Report to South Dakota Department of Transportation. Available from SD Coop Unit, SDSU Box 2140b, Brookings, SD 57007. 92 pages plus appendices.

Warren, M.L. Jr. and M.G. Pardew. 1998. Road crossings as barriers to small-stream fish movement. Transactions of the American Fisheries Society 127: 637 – 644.

Wellman, J.C., D.L. Combs, and S. Bradford Cook. 2000. Long-term impacts of bridge and culvert construction or replacement on fish communities and sediment characteristics of streams. Journal of Freshwater Ecology 15 (3): 317 –328.

WDFW. 2001. Fish passage design at road culverts: A design manual for fish passage at road crossings. Washington Department of Fish and Wildlife http://wdfw.wa.gov/hab/engineer/habeng.htm (accessed August 3, 2004).


HABITAT USE AND POPULATION BIOLOGYOF THE NORTHERN REDBELLY SNAKE

AT OAKLAKE FIELD STATION, SD

Regina D. Cahoe and Nels H. Troelstrup, Jr.Department of Biology & Microbiology


ABSTRACT

Habitat use and population biology of the northern redbelly snake (Storeria occipitomaculata occipitomaculata) were studied at the Oak Lake Field Station to enhance existing knowledge and future management efforts for this species in eastern South Dakota. The redbelly snake is listed as a monitored species in South Dakota, species of special concern in Nebraska and protected species in Iowa. Forty three snakes were caught by hand and from drift fences from May-August 2003. Three snakes (7%) were recaptured during the sampling period. Time duration between recaptures ranged from 1-63 days and all three recaptures occurred within 50m of their first collection. Nearly twice as many females (65%) as males (35%) were collected. Snout to vent length ranged from 97 to 245 mm (mean=185 mm) and body mass (WW) ranged from 1.0-10.7 g (mean =4.4 g). Air and ground temperatures at collection locations ranged from 17.8ºC to 29.6ºC. Over half (65.8%) of captured snakes were found on northerly aspects with soil moisture at collection sites ranging from 0%-100% (mean=56%). Many snakes (63%) were captured on unpaved roads consisting of 90% or greater sand and gravel. Vegetation cover at collection sites ranged from 0%-100% (median=38.8%). Those sites with vegetation were dominated by grasses and forbs. No snakes were found between 1200 hrs-1600 hrs through-out the summer and more than half of snakes were captured late during the growing season (58.1% during August). This study enhances our understanding of redbelly snake habitat use and demographics in eastern South Dakota. Future efforts will focus on descriptions of hibernacula and population estimates for the Oak Lake Field Station.

INTRODUCTION

Storeria occipitomaculata occipitomaculata, the northern redbelly snake, is a rare, non-venomous species. It is one of three subspecies found throughout much of the United States and parts of southern Canada (Ashton and Doud 1991). Redbelly snakes belong to the largest snake family, Colubridae. Snakes within this family have one lung and no limb remnants (Stidworthy 1974). S.o. occipi-tomaculata has been documented in Brookings, Codington, Day, Deuel, Grant, Hamlin, Lincoln, Minnehaha, and Roberts Counties (Ashton and Doud 1991).


The northern subspecies is currently listed as a monitored species in South Dakota, but previously carried a threatened status (Ashton and Doud 1991). It is currently listed as protected in Iowa and is a species of special concern in Nebraska. No special status has been granted in North Dakota or Minnesota. Adults of S.o. occipitomaculata are slender, small-bodied and can range from 20-25 cm in total length. Ground color varies from reddish brown to gray with faint brown stripes on the sides and a lighter stripe along the mid-dorsum (Ball-inger et al 2000). They are reported to feed on annelids, mollusks and insects in wooded and wet habitats (Smith and Stephens 2003; Thompson and Backlund 2000; Blanchard 1937). While the northern subspecies is widely distributed throughout the eastern United States, it does not appear to be overly abundant anywhere (Conant 1975). The species was first described by Storer in 1839 and the first detailed ecological study was conducted by Blanchard (1937) at the University of Michi-gan Biological Station on Douglas Lake. Blanchard’s survey from 1930-1936 resulted in the collection of only 157 individuals with an average of 22 collected per year. He described the diet, size distribution, sex ratio and general habits of the snake in northern Michigan. His study is still one of the most cited life history studies of S.o. occipitomaculata. In other states, S.o. occipitomaculata make use of rock crevices, abandoned ant mounds and animal burrows as hibernacula (Carpenter 1953). Females ap-parently emerge in the spring already gravid with eggs. Offspring are born alive between the months of July and August (Brodie and Ducey 1989). While few young have been captured, brood size appears to range up to 21 young per year (Oldfield and Moriarty 1994; Wright and Wright 1957) and varies with paren-tal female size (Brodie and Ducey 1989). S.o. occipitomaculata approach sexual maturity by the end of their first year and first reproduction is believed to take place the following spring or summer (Semlitsch and Moran 1984). Years of previous observation suggest that S.o. occipitomaculata are locally abundant within the boundaries of the South Dakota State University Oak Lake Field Station. Little is known regarding hibernacula, brood size, popula-tion size distribution, diet and specific habitat requirements of these snakes in eastern South Dakota. Thus, long-term observations and a locally abundant snake population provided the opportunity to (1) establish baseline natural history information and (2) begin population studies of S.o. occipitomaculata within eastern South Dakota. Oak Lake Field Station provides a good location for collecting this data because it provides optimal habitat, snakes are frequently observed and the station is near the western edge of the known distribution for this subspecies.

STUDY AREA

S.o. occipitomaculata were collected over the period May – August 2003 at Oak Lake Field Station, Brookings County, SD. Oak Lake Field Station is a 232 ha facility managed by South Dakota State University (Figure 1). This facility includes tall grass prairie, oak woodland, grazed pastures and pothole wetland


environments. Observations over several years suggest that S.o. occipitomaculata are locally abundant.

METHODS

S.o. occipitomaculata were collected using a combination of manual searches (Blanchard 1937) and drift fences with associated pitfall traps (Floyd et al. 2002; Semlitsch and Moran 1984). Sites frequented by redbelly snakes were manually searched at intervals throughout the day and evening five days a week through-out the summer. Six drift fences were constructed and placed in areas that snakes had previously been observed and also in a varying array of habitat surroundings (Table 1). Drift fences consisted of three - 3.0 m sections of flashing separated by 11 L bucket + funnel pitfalls. Four pitfalls were buried to the ground surface in each fence. Pitfall traps were checked every 24 hours Habitat attributes were evaluated within a 1 m2 quadrat placed at the site of each collection. A global positioning system was used to define each collection point. Air temperature, humidity and wind speed were evaluated with a Kestrel meteorological instrument. Light intensity reaching the surface was measured with an Extech Lux meter. Soil temperature and moisture were measured with a Rheotemp soil thermometer and Lincoln soil moisture probe. Percent of ground surface with clay-silt, sand-gravel and cobble-boulder substrate was visually esti-mated. Percent total vegetative cover and cover of grasses, forbs, shrubs and trees at the point of collection were also visually estimated.

Figure 1. S. occipitomaculata were collected from Oak Lake Field Station in northeast-ern Brookings County, South Dakota during summer 2003.


Table 1. Location of drift fences and pitfall traps placed at Oak Lake Field Station, Brookings County, SD.

Fence Latitude Longitude

1 44º 30.82’N 096º 32.47’W2 44º 30.34’N 096º 32.00’W3 44º 30.53’N 096º 31.96’W4 44º 30.65’N 096º 31.96’W5 44º 30.65’N 096º 31.75’W6 44º 30.35’N 096º 31.76’W

Any snakes sampled from manual surveys and pitfall traps were measured (snout-vent length, SVL), sexed and marked by clipping a single ventral scale. Canadian herpetologists suggest clipping subcaudal scales as there is less risk of injury to the snake from penetration into the abdominal cavity (Ministry of Environment, Lands and Parks 1998). The clipping of the scale does not harm the snake but the mark will remain with the snake even as it sheds its skin. Scale clippings were used to mark and identify snakes. A code was devised identifying the side and number of the clipped scale (e.g., 1L). Right and left were assigned with the snake ventral side-up and the head facing away from the researcher. Snakes were clipped until scales were deemed too close to the tip of the tail. This was repeated for the right side. From this point on, snakes were double clipped (e.g., 1L1R). Clipping was performed with small hand scissors. A triangular notch was cut, which revealed a dark, almost black underside. This darkened por-tion contrasted greatly with the red belly and made recaptures easy to identify. Snakes were sexed by (1) manual eversion of the hemipenal sacs and (2) breadth, length and shape of tail. Hemipene eversion can sometimes be difficult because older or large males can be difficult to evert and excess pressure may cause injury (Ministry of Environment, Lands and Parks 1998). Male snakes have a noticeably broader tail due to the presence of the retracted hemipenes in that portion of the tail (Ministry of Environment, Lands and Parks 1998). Also, male S.o. occipitomaculata have been found to have significantly longer tails than those of females (Semlitsch and Moran 1984). The use of a probe was considered unsafe because of small body size. Care was taken to insure snakes were released where they had been collected. All data were entered onto spreadsheet and submitted to the Oak Lake Field Station database. A one-way ANOVA was used to compare differences in length and mass among sexes (Snedecor and Cochran 1980). Linear regression was utilized to quantify the relationship between snake mass and length. Due to the small number of recaptures, a population estimate was not calculated.

RESULTS

Forty three S.o. occipitomaculata were captured between the months of May and August 2003 by hand or with pitfalls at Oak Lake Field Station. Live females


(65.0%) outnumbered live males (35.0%) throughout the study period. Three snakes were found dead. These specimens were preserved and donated to the Oak Lake Field Station collection. Of the 40 snakes which were captured alive and clipped, three were recaptured later during the summer (Table 2). Two of the three recaptured snakes were collected within a week of their original capture. All three were collected within 50 meters of their original capture point.

Table 2. Date, time, body mass and length of S.o. occipitomaculata captured and re-captured at Oak Lake Field Station, Brookings County, SD.

Specimen Julian Date Time (hrs) Weight (g) Length (mm)

3L 155 17:20 3.2 1913L Recapture 218 21:05 4.7 195

1R 181 9:43 2 1271R Recapture 188 20:20 2 127

8R 217 9:20 2.9 1798R Recapture 218 20:42 2.1 179

Snake SVL varied from 97 mm to 245 mm and averaged 185 mm (Table 3). No significant difference in SVL was observed between males and females (ANOVA, p = 0.228). Snake wet weight varied from 1.0 g to 10.7 g and aver-aged 4.4 g (Table 3). Again, no significant difference was observed between males and females (ANOVA, p = 0.409). However, the body mass of several individuals appeared to be below what might be expected based upon the mass-SVL relationship (Figure 2). In addition, the body mass of four individuals was far above that predicted by this relationship. These specimens may have been gravid or may have recently fed. Snake SVL was found to explain 44% of snake mass when all snake measurements were considered. However, the high body mass of four individuals were found to be statistical outliers (p < 0.05). When these individuals were eliminated from the analysis, snake SVL explained 72% of the variability in snake mass (Figure 2).

Table 3. Morphology of male and female S.o. occipitomaculata collected (minimum, median, maximum, mean and coefficient of variation (%)) at Oak Lake Field Station, Brookings County, South Dakota.

Dimension n Min Med Max Mean C.V.

Male Length (mm) 14 100 203 245 195 19.2Male Mass (g) 14 1.0 4.7 10.7 4.7 48.4Female Length (mm) 27 97 180 225 180 18.7Female Mass (g) 27 2.0 3.8 10.2 4.1 45.4

Snake collections were conducted at set and random times during the day from 0800 hrs to 2300 hrs. Fences were also checked during each search. S.o.


occipitomaculata were collected during the early morning and evening hours. No snakes were collected from 1200 hrs to 1600 hrs throughout the sampling period (Figure 3). Snakes were not found until after 1700 hrs from mid-July to late August. At this time during the summer, snakes were captured from dusk (around 2000 hrs) to near midnight (2300 hrs) using hand-held flashlights and spotlights. On August 7th (2000 hrs to 2200 hrs), nine individual snakes were caught during one search. This concentration of snakes had not previously been seen nor was again seen during the project period. Habitat measurements were made from each snake collection point. Many snakes were found on gravel roads and trails, presumably to sun or cross from one habitat range to the other. A positive correlation was found between air and ground temperatures where snakes were collected. Nearly 80% percent of collected snakes (30 of the 38 recorded) were found at air temperatures between 21º C and 26º C. A similarly high percentage of collected snakes were found on soils within this same range. More snakes were collected from habitat with northerly aspects than any other direction (65.8%). Most of the captured snakes (63%) were collected on unpaved roads consisting of 90% or more sand and gravel (Table 4). Soil moisture was typically greater than 50% of saturation and vegetative cover at collection points was typically less than 50%. Those sites with vegetative cover were dominated by grasses and forbs.

Figure 2. Relationship between S. occipitomaculata mass (WW) and snout to vent length from collections at Oak Lake Field Station, Brookings Co., South Dakota. Circled points are statistical outliers (p < 0.05).


Table 4. Summary of habitat conditions at S.o. occipitomaculata collection points, Oak Lake Field Station, Brookings County, SD.

Dimension n Min Med Max Mean C.V.

Air Temperature (C) 40 17.8 22.8 29.6 23.1 9.8Ground Temperature (C) 38 20.1 25.0 27.2 24.6 7.2Clay & Silt (%) 42 0.0 0.0 1.0 0.2 205.6Sand & Gravel (%) 42 0.0 0.9 1.0 0.8 43.5Cobble & Boulder (%) 42 0.0 0.0 0.5 0.1 186.5Soil Moisture (%) 41 0.0 80.0 100.0 56.4 71.3Vegetative Cover (%) 42 0.0 1.5 100.0 22.6 165.0Grasses (%) 42 0.0 0.0 100.0 16.0 187.4Forbs (%) 42 0.0 0.0 60.0 5.6 222.4Shrubs (%) 42 0.0 0.0 20.0 <0.1 478.3Trees (%) 42 0.0 0.0 10.0 <0.1 648.1

DISCUSSION

Forty individual snakes were caught within a three month sampling period during this effort. It is unknown if this number is high or low for a population of S.o. occipitomaculata because numbers captured vary greatly from one study to another. Blanchard (1937) captured 157 individuals in Michigan over the period 1930-1936 (average of 22 per year) while Semlitsch and Moran (1984) collected 249 individuals in South Carolina over a one year period. Trapido

Figure 3. Variation in snake collections throughout the growing season and during the course of the day at Oak Lake Field Station, Brookings Co., South Dakota.


(1944) concluded that S.o. occipitomaculata may be locally abundant in some areas and totally absent in others with seemingly similar habitat. A low recapture rate (4.7%) prevented estimation of S.o. occipitomaculata population size at Oak Lake Field Station. However, others have witnessed similar difficulty estimating population size using mark-recapture methods. Blanchard (1937) recaptured only 2 of 157 marked S.o. occipitomaculata. This low recapture success was attributed to the “interhabitat wanderings” of this spe-cies as other snake species sampled within the same study were more frequently recaptured. Semlitsch and Moran (1984) recaptured 4 of 61 marked snakes in South Carolina. They attributed low recaptures to rapid turnover of individuals within their population. We observed a sex ratio of nearly 2F:1M at Oak Lake Field Station through-out the summer collecting period. This ratio was similar to that observed by Blanchard (1937) in Michigan. In contrast, Semlitsch and Moran (1984) col-lected roughly equal numbers of adult male and female snakes in South Carolina. Over half the snakes we collected were found on gravel roads in August. Female snakes give birth to live young late in the summer (Ashton and Doud 1991), so it is possible that female snakes may seek warmer, exposed areas to (1) enhance development of their young or (2) find optimal birthing locations (Blanchard 1937). Snake size (snout to vent length and body mass) observed from this eastern South Dakota population appears to fall within the range reported from other studies. However, others have reported sexual dimorphism in body size with males exceeding females in total length (longer tails) and females exceeding males in body mass (Semlitsch and Moran 1984; Blanchard 1937). Drift fences and pitfall traps are frequently used to collect semi-fossorial snakes (Floyd et al. 2002; Semlitsch and Moran 1984) and we employed use of this equipment in this study. However, we collected only 2 (4.7%) snakes in pit-falls and 1 snake was captured under a pitfall trap. High prairie winds frequently destroyed fences and rainfall events sometimes flooded and filled pitfall buckets with water and sediment. In addition, wildlife frequented the pitfalls and often pulled the buckets from their holes and removed the funnels. Thus, fences and pitfalls required continuous maintenance with little return. While fences and pitfalls were somewhat difficult to maintain and snakes were infrequently caught, other animals were captured throughout the sampling season. Small mammals, (esp. Least Shrew (Cryptotis parva) and Prairie Vole (Microtus ochrogaster)) were frequently caught and released. In addition, several Silphidae, Carabidae, and Gryllidae species (all Insecta) were trapped almost daily within pitfalls and re-leased. Such frequent captures support the assumption that our fences and traps were working properly, but were not particularly effective for collecting redbelly snakes. Smith and Stephens (2003) recently recommended timed visual encoun-ter surveys to monitor the status of the Black Hills subspecies (S.o. pehasapae). S.o. occipitomaculata were found most frequently from northerly aspects. This orientation may have been related to snake preference for reduced sun ex-posure and lower temperatures during mid-summer. Most snakes collected from this effort were found in semi-mesic habitats with air and ground temperatures ranging between 21º C and 26º C. However, experimental studies of redbelly


snake thermal tolerance suggest that the critical thermal temperature for this species may be nearly 38º C (Brattstrom 1965). As summer air and ground temperatures commonly exceed 30º C in eastern South Dakota, one might con-clude that S.o. occipitomaculata may be avoiding areas of direct sunlight and hot temperatures. Elick and Sealander (1972) found S.o. occipitomaculata to have a high resistance to desiccation compared with other small colubrid snakes. How-ever, their study also suggested that these snakes may seek out places with high humidity to avoid dessication. S.o. occipitomaculata is widely distributed in the eastern half of the United States but does not appear to be overly abundant anywhere (Conant 1975; Trapido 1944). This snake is currently listed as a protected species in Iowa and a species of special concern in Nebraska. Listed previously as threatened in South Dakota (Ashton and Doud 1991), this status has now been downgraded to a monitored species. No special status has been granted in North Dakota or Minnesota. Even so, factors influencing mortality are of primary importance to non-game managers. Prior to this study, several S.o. occipitomaculata were found partially eviscerated but otherwise intact along the roads and in tallgrass prairie habitat of Oak Lake Field Station. While these snakes had obviously been attacked by other animals, they were not even partially consumed. Two observations made during the course of this study seemed to shed light on this phenomenon. Four barn swallows (Hirundo rustica) were observed to attack one snake within short, grass habitat. These birds would swoop down, pick up the snake and release it in mid-flight. The birds continued to do this until the snake was rescued. When recovered, the snake appeared unhurt (albeit stunned) with no puncture wounds evident. Another attack from an American Robin (Turdus migratorius) was observed in similar short grass habitat. Again, the snake was recovered before the bird had managed to cause mortal injury. Perhaps the deceased snakes observed previously had been similarly attacked by avifauna. Curiously, the victims of these attacks were not consumed. Barrett and Villar-roul (1992) reported an S.o. occipitomaculata inside a Kestrel’s (Falco sparverius) nest box. Smith and Stephens (2003) listed ruffed grouse (Bonasa umbellus), American robin (T. migratorius) and American kestrels (F. sparverius) among the known predators of redbelly snakes. Vehicle traffic may also pose considerable threat to redbelly snakes in high traffic areas. Oak Lake Field Station roads have limited vehicle access. How-ever, even this low traffic area witnessed mortality of three redbelly snakes from vehicles during this study. S.o. occipitomaculata’s tendency to “freeze” rather than “flee” from danger may increase chances of mortality from vehicular traffic. Oak Lake Field Station appears to contain a fairly large and possibly isolated population of S.o. occipitomaculata. South Dakota government documents sug-gest that this subspecies is only found in small pockets found on the eastern edge of South Dakota (Ballinger et al. 2000; Ashton and Dowd 1991). Oak Lake appears to be an optimal study site for this species due to an abundance of mesic habitat and population size. This study has enhanced our understanding of redbelly snake habitat use and demographics in eastern South Dakota. Future efforts will focus on hibernacula use and population estimates at Oak Lake Field Station.


ACKNOWLEDGMENTS

The authors thank South Dakota State University and the Oak Lake Field Station for the use of its property, facilities and equipment. Thanks are also extended to K. Okins, J. Rust, A. Gronke, and P. Lorenzen for their help in the field. Funding for this project was provided by a Joseph F. Nelson scholarship through South Dakota State University.

LITERATURE CITED

Ashton, D.E. and E.M. Dowd. 1991. Fragile legacy. Endangered, threatened and rare animals of South Dakota. South Dakota Department of Game, Fish and Parks, Report No. 91-04, Pierre, SD. 56pp.

Ballinger, R., J. Meeker and M. Thies. 2000. A checklist and distribution maps of the amphibians and reptiles of South Dakota. Transactions of the Ne-braska Academy of Science 26: 29-46.

Barret, G.C. and M.R. Villarroul. 1994. Storeria occipitomaculata occipitomacu-lata predation. Herpetological Review 25: 29-30.

Blanchard, F.N. 1937. Data on the natural history of the red-bellied snake, Store-ria occiptomaculata (Storer), in Northern Michigan. Copeia 3: 151-162.

Brattstrom B.H. 1965. Body temperatures of reptiles. American Midland Natu-ralist 73: 376-422.

Brodie, E.D. and P.K. Ducey 1989. Allocation of reproductive investments in the Redbelly Snake Storeria occipitomaculata. American Midland Naturalist 122: 51-58.

Carpenter, C. 1953. A study of hibernacula and hibernating associations of snakes and amphibians in Michigan. Ecology 34: 74-80.

Conant, R. 1975. A field guide to reptiles and amphibians of eastern and central North America. 2nd Edition, Houghton Mifflin Company, Boston. 429pp.

Elick, G.E. and J.A. Sealander. 1972. Comparative water loss in relation to habitat selection in small colubrid snakes. American Midland Naturalist 88: 429-439.

Floyd T.H, K.R. Russell, C.E. Moorman, D.H. Van Lear, D.C. Guynn, Jr. and J.D. Lanham. 2002. Effects of prescribed fire on herpetofauna within hard-wood forests of the Upper Piedmont of South Carolina. Proceedings of the Eleventh Biennial Southern Silvicultural Research Conference. U.S. Depart-ment of Agriculture. 622 pp.

Ministry of Environment, Lands and Parks. 1998. Inventory Methods for Snakes. Standards for Components of British Columbia’s Biodiversity No.38. Resources Inventory Branch for the Terrestrial Ecosystems Task Force, Resources Inventory Committee. Vancouver, B.C.50pp.

Oldfield, B. and J.J. Moriarty. 1994. Amphibians and reptiles native to Minne-sota. University of Minnesota Press, Minneapolis, MN. 237pp.

Semlitsch, R.D. and G.B. Moran. 1984. Ecology of the redbelly snake (Storeria occipitomaculata) using mesic habitats in South Carolina. American Midland Naturalist 111: 33-40.


Smith, B. and N. Stephens. 2003. Conservation assessment for the redbelly snake in the Black Hills National Forest South Dakota and Wyoming. United States Department of Agriculture, U.S. Forest Service, Custer, SD. 18pp.

Snedecor, G.W. and W.G. Cochran. 1980. Statistical methods. 7th Edition, Iowa State University Press, Ames, IA. 507pp.

Stidworthy, J. 1974. Snakes of the world. Grosset & Dunlap Publishers, New York. 160pp.

Thompson, S. and D. Backlund. 2000. South Dakota snakes. A guide to snake identification. South Dakota Department of Game, Fish and Parks, Pierre, SD. 28pp.

Trapido H. 1944 The snakes of the genus Storeria. American Midland Natural-ist 31:1-83.

Wright, A.H. and A.A. Wright. 1957. Handbook of snakes of the United States and Canada. Comstock Publishing Associates, Ithaca, NY. 564pp.


OVERVIEW AND RESULTS OF THEBLACK HILLS HYDROLOGY STUDY

Daniel G. Driscoll and Janet M. CarterU.S. Geological SurveyRapid City, SD 57702

ABSTRACT

The Black Hills Hydrology Study was a long-term study that was initiated in 1990 and was completed in 2002, following completion of a series of 21 reports and 11 published maps. The study was a regional assessment of water resources that had a purpose of assessing the quantity, quality, and distribution of surface water and ground water in the Black Hills area of South Dakota, which has com-plex hydrogeology. A major focus of the study was on describing the hydrologic significance of the major bedrock aquifers in the area, which are the Inyan Kara, Minnekahta, Minnelusa, Madison, and Deadwood aquifers. An overview of the study and a summary of major results are provided herein.

Keywords

Black Hills, hydrology, ground water, surface water, water quality, Madison aquifer, Minnelusa aquifer, precipitation, recharge

INTRODUCTION

The Black Hills area is an important resource center that provides an eco-nomic base for western South Dakota through tourism, agriculture, timber, and mineral resources. Water originating from the area is used for municipal, indus-trial, agricultural, and recreational purposes throughout much of western South Dakota. Population growth, resource development, and periodic droughts have the potential to affect the quantity, quality, and availability of water within the Black Hills area. The hydrogeology of the Black Hills area is extremely complex. A vertical se-quence of bedrock aquifers is contained within a series of geologic units that have been uplifted and exposed in the Black Hills area. Several important regional aquifers in the northern Great Plains receive substantial recharge in the area. Ground-water and surface-water hydrology are highly influenced by geologic conditions, which can be extremely heterogeneous and have large spatial vari-ability throughout the area. Ground water and surface water interact extensively in the area, and both aquifer recharge and streamflow are influenced by climatic conditions, which have large spatial and temporal variability. The Black Hills Hydrology Study was initiated in 1990 to assess the quan-tity, quality, and distribution of surface water and ground water in the Black


Hills area of South Dakota. This long-term study was a cooperative effort among the U.S. Geological Survey (USGS), the South Dakota Department of Environ-ment and Natural Resources, and the West Dakota Water Development District, which represented various local and county cooperators. This paper provides an overview of the Black Hills Hydrology Study and summarizes major results.

OVERVIEW OF STUDY

The Black Hills Hydrology Study was initiated in 1990 and was completed in 2002. The Black Hills Hydrology Study was designed as a regional assessment of the water resources and was not designed to address site-specific issues. The study focused on describing the hydrologic significance of the major bedrock aquifers in the Black Hills area, which are the Inyan Kara, Minnekahta, Minn-elusa, Madison, and Deadwood aquifers. The highest priority was placed on the Madison and Minnelusa aquifers, which are widely used and interact extensively with the surface-water resources of the area. The study consisted of two primary phases—data collection and interpreta-tion. An extensive network consisting of 71 observation wells, 94 precipitation gages, and 60 streamflow-gaging stations was used during the data collection phase, which ended in 1998. Critical components of this network (primarily observation wells and selected streamflow-gaging stations) are being maintained for long-term purposes through cooperative programs between the USGS and various local, State, and Federal cooperators. The interpretive phase of the study was completed in 2002, and resulted in the publication of a series of 21 reports and 11 maps.

DESCRIPTION OF STUDY AREA

The study area for the Black Hills Hydrology Study includes the topographi-cally defined Black Hills and portions of six counties in western South Dakota. The generalized outer extent of the outcrop of the Cretaceous-age Inyan Kara Group approximates the outer extent of the Black Hills area (Figure 1). The Black Hills are situated between the Cheyenne River and the Belle Fourche Riv-er, which is the largest tributary to the Cheyenne River. The study area includes most of the larger communities in western South Dakota and contains about one-fifth of the State’s population. Land-surface altitudes range from about 7,242 feet (2,207 meters (m)) above National Geodetic Vertical Datum of 1929 (NGVD 29) at Harney Peak to about 3,000 feet (914 m) above NGVD 29 in the adjacent plains. The overall climate of the Black Hills area is continental, which is charac-terized by low precipitation amounts, hot summers, cold winters, and extreme variations in both precipitation and temperatures. Local climatic conditions are affected by topography, with generally decreasing temperatures and increasing precipitation with increasing altitude. Average annual precipitation for water years (WY) 1950-98 ranged from 16 inches (41 centimeters (cm)) in the south-


ern and northern parts of the study area to greater than 28 inches (71 cm) near Lead and Deadwood (Driscoll and Carter 2001). Long-term trends in precipitation for WY1931-98 for the study area are il-lustrated in Figure 2A. Annual precipitation for the study area averaged 18.61 inches (47 cm) and ranged from 10.22 inches (26 cm) in WY 1936 to 27.39 inches (70 cm) in WY 1995. The cumulative trends (Figure 2B) show that sustained periods of generally deficit precipitation occurred during 1931-40 and 1948-61. The middle to late 1990s was the wettest period since 1931, which caused potential for bias towards wet conditions for hydrologic data collected during this period. This potential bias was addressed in various analyses per-formed as part of the study, and has been balanced to some extent by relatively dry conditions during the late 1980s and early 1990s.

Figure 1. Study area.


Throughout geologic time, the Black Hills area has experienced frequent periods of inundation by seas, extended erosion, mountain building, and intru-sion by igneous rocks; thus, the hydrogeology of the study area is very complex. The Black Hills uplift formed as an elongated dome about 60 to 65 million years ago during the Laramide orogeny (Darton and Paige 1925). The oldest geologic units in the study area are the Precambrian crystalline (igneous and metamorphic) rocks, which are exposed in the central core of the Black Hills, extending from near Lead to south of Custer. The Precambrian rocks generally have low permeability; however, localized aquifers occur in many locations in the crystalline core of the Black Hills where secondary porosity and permeability have resulted from fracturing and weathering of the rocks. Surrounding the Precambrian crystalline core is a layered series of sedi-mentary rocks including limestones, sandstones, and shales that are exposed in roughly concentric rings around the uplifted flanks of the Black Hills, as shown by the outcrops of the Madison Limestone and Minnelusa Formation (Figure 1). The more permeable of these sedimentary rocks—the Cambrian- and Ordovi-cian-age Deadwood Formation, Mississippian-age Madison Limestone, Penn-sylvanian- and Permian-age Minnelusa Formation, Permian-age Minnekahta Limestone, and Cretaceous-age Inyan Kara Group—contain major aquifers that

Figure 2. Long term trends in precipitation for Black Hills area, water years 1931-98 (from Driscoll, Hamade, and Kenner 2000).


are able to store and transmit large quantities of water and are used extensively for water supplies within and beyond the study area. Minor bedrock aquifers occur in other units, including confining units, due to fracturing and interbed-ded permeable layers. Various unconsolidated units, including alluvium and colluvium, are considered aquifers, where saturated. The hydrologic setting of the Black Hills is schematically illustrated in Figure 3. Individually, the major aquifers generally are separated by confining layers, which are composed of less permeable rocks, or by relatively impermeable layers within the individual units. The aquifers and confining units generally dip away from the flanks of the Black Hills (Figure 3). In general, ground-water flow in these aquifers is radially away from the central core of the Black Hills. The aqui-fers primarily receive recharge from infiltration of precipitation on outcrops, and several aquifers also receive recharge from streamflow losses. Confined (artesian) conditions generally exist within the major aquifers, in locations where an upper confining layer is present. Flowing wells and artesian springs that originate from confined aquifers are common around the periphery of the Black Hills.

Surface water in the Black Hills area is highly influenced by geologic condi-tions, and five hydrogeologic settings have been identified that have distinctive influences on streamflow and surface-water quality. The five settings, which are shown in Figure 4, are represented by four areas because two of the settings—loss zone and artesian spring—are considered to share a common area. Numerous headwater (water-table) springs, originating primarily from the Madison and Minnelusa aquifers, occur in the limestone headwater setting, which is a high-altitude area of generally low relief in the western part of the

Figure 3. Schematic showing simplified hydrologic setting of the Black Hills area.


study area known as the “Limestone Plateau” (Figure 4). Most of the headwater springs occur near the eastern edge of the Limestone Plateau and provide base flow for many Black Hills streams. These streams flow across the Precambrian igneous and metamorphic rocks in the crystalline core setting. The loss zone setting (Figure 4) consists of areas that are heavily influenced by streamflow losses. Most streams generally lose all or part of their flow as they cross the outcrop of the Madison Limestone (Rahn and Gries 1973; Hortness and Driscoll 1998). Karst features of the Madison Limestone, including col-lapse features, sinkholes, and solution features such as caves, are responsible for the Madison aquifer’s large capacity to accept recharge from streamflow. Large streamflow losses also occur in many locations within the outcrop of the Min-nelusa Formation, and limited losses probably also occur within the outcrop of the Minnekahta Limestone (Hortness and Driscoll 1998).

Figure 4. Hydrogeologic settings for the Black Hills area.


Large artesian springs originating primarily from the Madison and Min-nelusa aquifers occur in many locations downgradient from these loss zones. Thus, the loss zone and artesian spring settings are represented by the same area (Figure 4). These artesian springs provide an important source of base flow in many streams beyond the periphery of the Black Hills. No artesian springs are known to be located beyond the outcrop of the Inyan Kara Group; thus, the area beyond this outcrop is considered to be the exterior setting (Figure 4).

RESULTS OF STUDY

Detailed results of the Black Hills Hydrology Study have been reported in a variety of publications, including data reports, map reports, topical reports, and summary reports. Water-level data through WY 1998 for the entire network of observation wells were summarized in a project data report (Driscoll, Bradford, Moran 2000) that also included water-quality data for a variety of surface-and ground-water sampling sites. Previous project data reports (Driscoll and Brad-ford 1994; Driscoll et al. 1996) provided similar data. Streamflow and precipi-tation data were published annually in Water Resources Data for South Dakota (U.S. Geological Survey 1991-99). A summary of precipitation data available for the study area was provided by Driscoll, Hamade, and Kenner (2000). The report summarized monthly data for precipitation gages operated within the study area during 1931-98 and examined long-term trends for precipitation. The report also provided an isohyetal map showing the spatial distribution of precipitation for WY 1961-90. A number of topical reports addressing a variety of subject matter were pub-lished throughout the course of the study. One of the earliest topical reports was a compilation of streamflow loss thresholds (maximum sustainable loss rates) to bedrock aquifers for major streams within the study area (Hortness and Driscoll 1998). Unique loss thresholds were quantified for 24 streams in the study area and ranged from negligible (no loss) for several streams to as much as 50 cubic feet per second (ft3/s) (1.4 cubic meters per second (m3/s)) for Boxelder Creek. Quantification of losses to individual aquifers was possible for only about one-third of the streams. Loss thresholds for individual streams were concluded to be relatively constant, and without measurable effect from streamflow rates or duration of flow through the loss zones. A summary of data for all continuous-record gaging stations within the study area was assembled by Miller and Driscoll (1998). This report included detailed statistical summaries for gaging stations with 10 or more years of stream-flow data and also provided a preliminary characterization of geologic influences on streamflow. A more detailed analysis of the distinctive influences of the five different hydrogeologic settings (Figure 4) on streamflow characteristics was provided by Driscoll and Carter (2001). Relations between annual streamflow and annual precipitation were quantified for “representative” gaging stations for selected hydrogeologic settings. Relative strong correlations between annual streamflow and precipitation were demonstrated for the crystalline core and exterior settings. Similar correlations were extremely weak for the limestone


headwater setting because of the influence of long-term ground-water storage; however, correlations for this setting were substantially improved by using “mov-ing-average” precipitation as an explanatory variable. A series of 1:100,000-scale maps showing ground-water resources were produced, and included maps showing the distribution of hydrogeologic units (Strobel et al. 1999), potentiometric surfaces of the major bedrock aquifers (Strobel et al. 2000), and altitudes of the tops of the geologic formations that contain the major aquifers (Carter and Redden 1999). Accompanying data re-ports documenting selected data for specific wells and springs were published for the potentiometric maps (Galloway 2000) and structure-contour maps (Carter 1999). Water-level records from 71 observation wells indicated that there is no long-term decline in water levels in any of the bedrock aquifers in the Black Hills area. Water levels in a large percentage of the observation wells completed in the Madison and Minnelusa aquifers respond especially quickly to climatic condi-tions. The total volume of recoverable water stored in the major aquifers (including aquifers in the Precambrian rocks) within the study area was estimated as 256 million acre-feet (316.1 billion cubic meters) (Carter et al. 2002). Most of the recoverable water is stored in the Inyan Kara, Minnelusa, and Madison aquifers (Table 1).

Table 1. Estimated volumes of recoverable water in storage for major aquifers in study area (modified from Carter et al. 2002).

AquiferEstimated Amount Of Recoverable Water In Storage

Million acre-feet Billion cubic meters

Inyan Kara 84.7 104.5Minnekahta 4.9 6.0Minnelusa 70.9 87.5Madison 62.7 77.3

Deadwood 30.5 37.6Precambrian 2.6 3.2

COMBINED STORAGE FOR MAJOR AQUIFERS 256.3 316.1

Hydraulic connection between the Madison and Minnelusa aquifers at some locations in the Rapid City area has been confirmed through aquifer testing and dye tracing (Greene 1993, 1997). In the Spearfish area, aquifer testing provided no indication of hydraulic connection in the vicinity of tested wells (Greene et al. 1999). Potential for hydraulic connection is indicated in several locations by similarities in hydrographs for paired observation wells completed in the Madison and Minnelusa aquifers; however, hydraulic connection generally can-not be confirmed or refuted because aquifer testing has not been performed at most locations (Driscoll et. al. 2002). Hydraulic connection probably occurs at


many artesian spring locations, and artesian springflow probably accounts for the largest percentage of leakage that occurs between the Madison and Minnelusa aquifers Carter, Driscoll, and Hamade (2001) used a variety of methods in estimat-ing recharge to the Madison and Minnelusa aquifers. Recharge estimates for the aquifers were combined because recharge from streamflow losses could not be quantified separately for most streams. Estimated relations between annual recharge and precipitation were used in deriving estimates of recharge from infiltration of precipitation. Large outcrops of the Madison Limestone and Minnelusa Formation occur in the Black Hills of Wyoming and were considered in estimating recharge to the Madison and Minnelusa aquifers. Recharge to these aquifers for WY 1931-98 in South Dakota and Wyoming averaged about 344 ft3/s (9.7 m3/s). Annual recharge rates were highly variable (Figure 5) and ranged from about 62 ft3/s (1.8 m3/s) in 1936 to about 847 ft3/s (24.0 m3/s) in 1995. About three-quarters of the total recharge to the Madison and Minnelusa aquifers is from infiltration of precipitation on outcrops, and about one-quarter is from streamflow losses. The largest amount of precipitation recharge occurs in the Limestone Plateau area of the Black Hills.

The recharge estimates developed by Carter, Driscoll, and Hamade (2001) were used extensively in development of detailed hydrologic budgets for the Madison and Minnelusa aquifers (Carter, Driscoll, Hamade, and Jarrell 2001).

Figure 5. Annual recharge to the Madison and Minnelusa aquifers, in the Black Hills of South Dakota and Wyoming, water years 1931-98 (from Carter, Driscoll, and Hamade, 2001).


The budgets for these aquifers were combined because several of the budget com-ponents, including recharge and springflow, could not be quantified individually. Estimates of average combined budget components for WY 1987-96 were: 395 ft3/s (11.2 m3/s) for recharge, 78 ft3/s (2.2 m3/s) for headwater springflow, 189 ft3/s (5.4 m3/s) for artesian springflow, 28 ft3/s (0.8 m3/s) for well withdrawals, and 100 ft3/s (2.8 m3/s) for net ground-water outflow from the study area. The recharge estimates for WY 1987-96 are substantially higher than for WY 1931-98 because of substantially wetter climatic conditions. Artesian springflow is the single largest discharge component for the Madison and Minnelusa aquifers. Various hydrologic budgets (WY 1950-98) for the study area (excluding Wyoming) were developed by Driscoll and Carter (2001), including ground-water budgets for all of the major bedrock aquifers, surface-water budgets, and combined ground- and surface-water budgets. Ground-water budgets for the study area are dominated by the Madison and Minnelusa aquifers, with recharge to these aquifers comprising about 84 percent of total recharge to all bedrock aquifers, which averaged about 348 ft3/s (9.9 m3/s). Springflow was estimated as 219 ft3/s (6.2 m3/s), of which 94 percent originates from the Madison and Minnelusa aquifers. Well withdrawals were estimated as 40 ft3/s (1.1 m3/s), of which 70 percent was withdrawn from the Madison and Minnelusa aquifers. Ground-water outflow from the study area was estimated as 89 ft3/s (2.5 m3/s), of which 65 percent occurs in the Madison and Minnelusa aquifers. Surface-water inflows to the study area have averaged about 252 ft3/s (7.1 m3/s) and outflows have averaged about 552 ft3/s (15.6 m3/s). Total consumptive usage from both ground-water and surface-water sources was estimated as 218 ft3/s (6.2 m3/s), which includes well withdrawals of 40 ft3/s (1.1 m3/s), reservoir evaporation of 38 ft3/s (1.1 m3/s), and consumptive streamflow withdrawals of 140 ft3/s (4.0 m3/s). Of the average annual precipitation in the study area, about 91.6 percent is returned to the atmosphere through evapotranspiration, about 3.5 percent recharges aquifers in the study area, and about 4.9 percent becomes runoff from the land surface. Water-quality characteristics for ground water and surface water were sum-marized by Williamson and Carter (2001). Water quality of the major aquifers generally is very good in and near outcrop areas but deteriorates progressively with distance from the outcrops. In the Minnelusa aquifer, concentrations of dissolved sulfate vary markedly over short distances, especially near a zone where active anhydrite dissolution occurs. Most limitations for the use of ground water are related to aesthetic qualities associated with hardness and high concentrations of chloride, sulfate, sodium, manganese, and iron. Very few health-related limita-tions exist for ground water; most limitations are for substances associated with radioactive decay, such as radon and uranium. In addition, high concentrations of arsenic have been detected in a few samples from the Madison and Minnelusa aquifers. Surface-water quality is distinctly influenced by the hydrogeologic settings (Figure 4). For most streams, concentrations of dissolved solids increase as streamflow decreases. However, for streams in the limestone headwater and artesian spring settings, which are dominated by ground-water discharge, con-centrations of dissolved solids have little variability. Most streams generally meet


water-quality standards established for designated beneficial uses. The primary exceptions are streams in the exterior setting, which frequently fail to meet stan-dards for suspended sediment (South Dakota Department of Environment and Natural Resources 1998) and occasionally fail to meet standards for temperature and dissolved oxygen during low-flow conditions. Water-quality characteristics for selected streams in Lawrence County, which includes many mineralized areas, were described by Williamson and Hayes (2000). Separating influences of mining activities from natural water quality generally is difficult because historical data are sparse. Degradation of water quality from large-scale mining activities has been documented, however, in An-nie, Bear Butte, Squaw, and Whitewood Creeks. A major focus of the Black Hills Hydrology Study was to obtain a better understanding of flow systems within the Madison and Minnelusa aquifers, which are extremely complex due to heterogeneity and anisotropy related to karst features and fractures and to interactions between the aquifers and surface-water resources. Geochemical analyses by Naus et al. (2001) showed that regional flowpaths in the Madison and Minnelusa aquifers are largely deflected around the Black Hills. The dominant proportion of water in these aquifers within the study area is recharged within the Black Hills area. Age-dating analyses indicated that ground-water flow velocities in the Madison and Minnelusa aquifers are extremely variable. A technical summary (Driscoll et al. 2002) and a layreader summary (Carter et al. 2002) of the results of the Black Hills Hydrology study were published. Most reports published as part of the study are available online at URL http://sd.water.usgs.gov/projects/bhhs/table1.html. More information about the Black Hills Hydrology Study, including digital data, is available online at URL http://sd.water.usgs.gov/projects/bhhs/Intro.html. The Black Hills Hydrology Study has provided an abundance of information regarding the water resources of the Black Hills area, which is widely used by a variety of resource managers. This information base also will serve as a founda-tion for addressing future needs for additional hydrologic information. Ongoing research by various organizations continues to address developing needs.

REFERENCES CITEDCarter, J.M. 1999. Selected data for wells and test holes used in structure-con-

tour maps of the Inyan Kara Group, Minnekahta Limestone, Minnelusa Formation, Madison Limestone, and Deadwood Formation in the Black Hills area, South Dakota. USGS OFR 99-260. 51 pp.

Carter, J.M., Driscoll, D.G., and G.R. Hamade. 2001. Estimated recharge to the Madison and Minnelusa aquifers in the Black Hills area, South Dakota. USGS WRIR 00-4278. 66 pp.

Carter, J.M., Driscoll, D.G., Hamade, G.R., and G.J. Jarrell. 2001. Hydrologic budgets for the Madison and Minnelusa aquifers in the Black Hills of South Dakota and Wyoming, water years 1987-96. USGS WRIR 01-4199. 53 pp.


Carter, J.M., Driscoll, D.G., Williamson, J.E., and V.A. Lindquist. 2002. Atlas of water resources in the Black Hills area, South Dakota. USGS HA-747, 120 pp.

Carter, J.M., and J.A. Redden. 1999. Altitude of the top of the Inyan Kara Group, Minnekahta Limestone, Minnelusa Formation, Madison Limestone, and Deadwood Formation in the Black Hills area, South Dakota. USGS HA-744-A to HA-744-E. Scale 1:100,000.

Darton, N.H., and Sidney Paige. 1925. Central Black Hills [quadrangle], South Dakota. USGS Atlas of the United States. Folio 219. 34 pp.

Driscoll, D.G., and W.L. Bradford. 1994. Compilation of selected hydrologic data, through water years 1992, Black Hills Hydrology Study, western South Dakota. USGS OFR 94-319. 158 pp.

Driscoll, D.G., Bradford, W.L., and M.J. Moran. 2000. Selected hydrologic data, through water year 1998, Black Hills Hydrology Study, South Dakota. USGS OFR 00-70. 284 pp.

Driscoll, D.G., Bradford, W.L., and K.M. Neitzert. 1996. Selected hydrologic data through water year 1994, Black Hills Hydrology Study, South Dakota. USGS OFR 96-399. 162 pp.

Driscoll, D.G., and J.M. Carter. 2001. Hydrologic conditions and budgets in the Black Hills area of South Dakota, through water year 1998. USGS WRIR 01-4226. 143 pp.

Driscoll, D.G., Carter, J.M., Williamson, J.E., and L.D. Putnam. 2002. Hydrol-ogy of the Black Hills area, South Dakota: USGS WRIR 02-4094. 150 pp.

Driscoll, D.G., Hamade, G.R., and S.J. Kenner. 2000. Summary of precipita-tion data compiled for the Black Hills area of South Dakota, water years 1931-98. USGS OFR 00-329. 151 pp.

Galloway, J.M. 2000. Selected hydrologic data for the Inyan Kara, Minnekahta, Minnelusa, Madison, and Deadwood aquifers in the Black Hills area, South Dakota. USGS OFR 99-602. 60 pp.

Greene, E.A. 1993. Hydraulic properties of the Madison aquifer system in the western Rapid City area, South Dakota. USGS WRIR 93-4008. 56 pp.

Greene, E.A. 1997. Tracing recharge from sinking streams over spatial dimen-sions of kilometers in a karst aquifer. Ground Water. 35:898-904.

Greene, E.A., Shapiro, A.M., and J.M. Carter. 1999. Hydrogeologic charac-terization of the Minnelusa and Madison aquifers near Spearfish, South Dakota. USGS WRIR 98-4156. 64 pp.

Hortness, J.E., and D.G. Driscoll. 1998. Streamflow losses in the Black Hills of western South Dakota. USGS WRIR 98-4116. 99 pp.

Miller, L.D., and D.G. Driscoll. 1998. Streamflow characteristics for the Black Hills of South Dakota, through water year 1993. USGS WRIR 97-4288. 322 pp.

Naus, C.A., Driscoll, D.G., and J.M. Carter. 2001. Geochemistry of the Madi-son and Minnelusa aquifers in the Black Hills area, South Dakota. USGS WRIR 01-4129. 118 pp.


Rahn, P.H., and J.P. Gries. 1973. Large springs in the Black Hills, South Dakota and Wyoming: S.D. Geological Survey Report of Investigations 107. 46 pp.

South Dakota Department of Environment and Natural Resources. 1998. The 1998 South Dakota 303(d) waterbody list. South Dakota Department of Environment and Natural Resources, Pierre, S.D. Various pagination.

Strobel, M.L., Galloway, J.M., and G.R. Hamade. 2000. Potentiometric surface of the Inyan Kara, Minnekahta, Minnelusa, Madison, and Deadwood aqui-fers in the Black Hills area, South Dakota. USGS HA-745-A to HA-745-E. Scale 1:100,000.

Strobel, M.L., Jarrell, G.J., Sawyer, J.F., J.R. Schleicher, and M.D. Fahrenbach. 1999. Distribution of hydrogeologic units in the Black Hills area, South Dakota. USGS HA-743. Scale 1:100,000.

U.S. Geological Survey. 1991-99. Water resources data for South Dakota, water years 1990-98. USGS Water Data Reports SD-90-1 to SD-98-1 (published annually).

Williamson, J.E., and J.M. Carter. 2001. Water-quality characteristics for the Black Hills area, South Dakota. USGS WRIR 01-4194. 196 pp.

Williamson, J.E., and T.S. Hayes. 2000. Water-quality characteristics for selected streams in Lawrence County, South Dakota, 1988-92. USGS WRIR 00-4220. 131 pp.


STRATIGRAPHY AND ANALYTICPALEONTOLOGY OF THE LOWER

PIERRE SHALE AT BROWN RANCH,SOUTHWESTERN SOUTH DAKOTA

Marcus R. RossDepartment of GeosciencesUniversity of Rhode Island

Kingston, RI 02881

ABSTRACT

Paleontologists, both staff and student, from the Museum of Geology at the South Dakota School of Mines and Technology have been actively prospecting for fossils in the Campanian-age exposures of the Lower Pierre Shale on the Kenneth Brown Ranch since 1994. A measured section of the Pierre Shale at Brown Ranch is presented, and consists of 27.36 m of strata, including outcrops of the Gammon Ferruginous, Sharon Springs, and Mitten Black Shale Members. Field prospecting was combined with an assessment of the Museum of Geology collections at the South Dakota School of Mines and Technology to produce a database of 53 vertebrate located within an interval of 0 to 18 m above the base of the section. Fossil distributions within this interval were grouped and tested in 0.5- and 1 m thick stratigraphic intervals and compared against random (Pois-son) distributions. At a resolution of 1 m, fossils display random distribution, whereas at a resolution of 0.5 m fossils display an aggregated distribution. Partic-ularly fossiliferous horizons at a resolution of 0.5 m are identified at the quadrats 5.0 and 14.5 m (and perhaps at 12.0 m) above the measured base. These results enable predictions to be made concerning the locations of potentially fossilifer-ous areas at Brown Ranch.

Keywords

Late Cretaceous, marine stratigraphy, Poisson test

INTRODUCTION

The Kenneth Brown Ranch is located in southwestern South Dakota, ap-proximately fifteen miles (32 km) to the southeast Rapid City, seven miles (10 km) east of Highway 79 along Lower Spring Creek Road (Figure 1). The study area is in Sec. 7, T25, R9E, in the Hermosa NE USGS topographic map (7.5 minute/1:24,000 scale). Since 1994, staff and students from the Museum of Ge-ology at the South Dakota School of Mines and Technology have been actively prospecting and collecting the Brown Ranch for vertebrate and invertebrate


remains of marine organisms from exposures of the lower members of the Pierre Shale. This study aims to: 1) document the geology and stratigraphy of the Lower Pierre Shale at Brown Ranch, and 2) determine what the pattern of fossil distribution is within the stratigraphy of Brown Ranch.

METHODS

A measured section of the strata exposed at Brown Ranch (see Appendix I) was constructed using a five-foot Jacob staff. Measurements were then converted to metric for construction of the stratigraphic column presented here. Local dip is 10º, with a dip direction of S57ºE (strike of N33ºE). Fossils at Brown Ranch were located by prospecting in the field, consulting Museum of Geology field notes, and with the generous help of Kenneth Brown, who identified many of the sites where the Museum of Geology had collected specimens in the past. The stratigraphic positions of fossils were documented in the field and taken from information recorded in field notes of previously collected specimens. These data were correlated with the measured section recorded in Appendix I. Finally, statistical analysis of the stratigraphic distribution of fossils was con-ducted using Microsoft Excel. Fossils were placed into 0.5 and 1 m thick inter-

Figure 1. Location of study area. State of South Dakota inset.


vals of strata according to their position above the base of the measured section. These distributions were then compared against random (Poisson) distributions to determine if the distribution pattern of the fossils is random, uniform, or ag-gregated for each interval thickness.

THE PIERRE SHALE

Depositional Setting

The Pierre Shale is a thick sequence of black marine shale, mudstone, and claystone, deposited during the late Cretaceous (early Campanian through late Maastrichtian). The Pierre Shale was deposited as sediments settled in the north-south trending epicontinental Cretaceous Western Interior Seaway. This seaway connected the Gulf of Mexico with the Arctic Ocean during the late Cretaceous, with probable connection with the Hudson Bay area as well (Dyman et. al, 1994, Kauffmann, 1984). The morphology of the Cretaceous Western Interior Seaway was constantly changing due to the effect of tectono-eustatic third-order marine cycles coupled with smaller, fourth-order marine cycles (see Kauffman ,1984). The Cretaceous Western Interior Seaway was an asymmetric basin, bounded by the Cordilleran thrust belt to the west, and a stable craton on to the east (Dyman, et. al, 1994, see Kauffmann, 1984 for diagram). Deposits of terrig-enous-derived clastics were thickest in the west in a subsiding foreland basin, and thinned to the east (Dyman et. al, 1994). Kauffmann (1984) estimated maximum water depth as between 200 and 500 m along the western and central portions of the Western Interior Seaway, with a shallower region over a hinge zone in the east. Pierre Shale was deposited in this shallower region (Kauffmann, 1984).

General Stratigraphy of the Lower Pierre Shale

The Pierre Shale was originally named by Meek and Hayden (1861) based on outcrops along the Missouri River near Fort Pierre, South Dakota. The unit cov-ers much of South Dakota west of the Missouri River, excluding the Badlands, the Black Hills, and northwestern South Dakota (Gries and Martin, 1985). A reference section for the Pierre Shale in the southern Black Hills was described at Red Bird, Wyoming, by Gill and Cobban (1966). At Red Bird, the Pierre Shale is 3,137.2 feet thick (956.2 m) and is subdivided into seven members (from oldest to youngest): Gammon Ferruginous Member, Sharon Springs Member, Mitten Black Shale Member, Red Bird Silty Member, Lower Unnamed Member, Kara Bentonitic Member, and the Upper Unnamed Member (Gill and Cobban, 1966). Pertinent to the present study are the lower three members of the Pierre Shale: the Gammon Ferruginous Member, Sharon Springs Member, and Mitten Black Member. Each of these three units is exposed at Brown Ranch. The Gammon Ferruginous Member was deposited during in the Scaphites hippocrepis, Baculites sp. (smooth), and Baculites sp. (weak flank ribs) range zones during the late early Campanian (Kauffman et al., 1993, Obradovich, 1993,


Dyman et al., 1994). Also, the Gammon Ferruginous Member was the last unit deposited during the fourth marine cycle (Niobrara Cyclothem of Kauffmann, 1984) of the Cretaceous Western Interior Seaway (Dyman et al., 1994). The Sharon Springs Member was deposited during in the Baculites obtusus range zone during the early middle Campanian (Kauffman et al., 1993, Obradovich, 1993). The Ardmore bentonite is located at the base of the Baculites obtusus range zone, and has been dated at 80.5 mya +/- 0.5 Ma (Obradovich, 1993). The Mitten Black Shale Member was deposited during the middle Campanian, from the Baculites mclearni through the Baculites perplexus (upper) biostratigraphic range zones (Gill and Cobban, 1966, Kauffman et al., 1993, Dyman et al., 1994). Both the Sharon Springs Member and the Mitten Black Shale Member were de-posited during the fifth marine cycle (Clagget Cyclothem of Kauffmann, 1984) of the Cretaceous Western Interior Seaway (Dyman et al., 1994).

Stratigraphy of the Lower Pierre Shale at Brown Ranch

A measured section of the members exposed at Brown Ranch is recorded in Appendix I, and illustrated in Figure 2. The total thickness of strata recorded from Brown Ranch is 27.36 m (nearly 90 feet).

Gammon Ferruginous Member

The Gammon Ferruginous Member was described by Gill and Cobban (1966, p. A8) as a “dark hard platy-weathering noncalcareous shale that contains a few layers of widely separated red-weathering siderite concretions and several thin beds of pale-yellow non-swelling bentonite…[and] generally weathers to a smooth, rounded bald outcrop.” At Brown Ranch, few exposures of the Gam-mon Ferruginous Member crop out in the base of gullies and as low profile, sparsely vegetated surfaces. The shale is massive, dark olive-gray in color, and weathers into small light gray chips and dust. There are no siderite concretions in the Gammon Ferruginous Member exposed on Brown Ranch; however, there are three small, pale yellow bentonites near top of the unit (G1-G3). The base of the measured section (Figure 2) presented in Appendix I was established at the site of a partial skeleton of a subadult mosasaur, Tylosaurus pro-riger (field number MRR-01-01) recovered by staff and students of the Museum of Geology during the summer of 2002. This specimen, found 2.10 m below the base of the Ardmore bentonite “series” (see below), is at present the lowest reliable stratigraphic datum within the Gammon Ferruginous Member at Brown Ranch and the only fossil specimen recovered from the Gammon Ferruginous Member at Brown Ranch. Thus the Gammon Ferruginous Member is not fully exposed at Brown Ranch. An unconformity and a thick sequence of bentonites (see below) mark the contact of the Gammon Ferruginous Member with the Sharon Springs Member (Martin 1996).

Sharon Springs Member

The Sharon Springs Member was first recognized and described in the Black Hills by Darton (1902, p. 4) for outcrops of “a very distinct black, splintery,


fissile shale about 150 feet thick.” Elias (1931) affixed the name Sharon Springs Member to deposits in Kansas (Wallace and Logan Counties), and Searight ap-plied the name to deposits in the Black Hills in 1938. One year later, Moxon, Olson, and Searight (1939) gave a full description of what they considered to be the Gammon Ferruginous Member, Sharon Springs Member, and Mitten Black Shale Member of the lower Pierre Shale. In their work on the Pierre Shale, Gill and Cobban (1966) suggested that the description given by Moxon, Olson, and Searight (1939) for the “lower” and “upper” zones of the Sharon Springs Mem-ber were actually descriptions of the Gammon Ferruginous Member Ferruginous and Mitten Black Shale Member, respectively, leaving the “middle” zone as the Sharon Springs Member. In his study on the elevated uranium content of the Sharon Springs Mem-ber, Kepferle (1959) discussed the unique depositional setting that produced this unit. He noted that the elevated organic content of the shale (comprised largely of macerated plant material and capable of producing “about four gallons of oil per ton of shale”, p. 581) indicated that the Sharon Springs Member was deposited slowly, under reducing conditions. Furthermore, the sediment input

Figure 2. Stratigraphic column of the lower Pierre Shale at Brown Ranch. See text and Appendix I for discussion. GFM = Gammon Ferruginous Member; BBS = Basal Bentonite Series.


was from the west, based on 1) interfingering sandstones and shales in Montana and Wyoming, 2) bentonites thicken westward, and 3) the presence of westward thickening of the Sharon Springs Member (p. 600). Kepferle (1959, p. 600-601) also proposed a maximum water depth of “less than 250 feet.” This estimate was based on three lines of evidence: 1) excellent preservation of Baculites specimens, indicating shallow, quiescent bottom condi-tions; 2) the marl of the overlying Gregory Member at the Missouri River in South Dakota contain neritic-zone foraminifera; and 3) thin lag deposits pre-sumed to represent erosion by wave action. Kauffmann (1984, fig. 2) appears to agree with this water depth estimate, though Kauffmann did not discuss a precise depth for the Sharon Springs Member (or the Pierre Shale in general), nor did he discuss his methods for estimating water depth. The total thickness of the Sharon Springs Member at Brown Ranch is 13.78 m (Figure 2), and unconformably overlain by the Mitten Black Shale Member. In contrast, the thickness of the Sharon Springs Member at Red Bird is 126.6 feet (30.59 m), and conformably overlain by the Mitten Black Shale Member (Gill and Cobban, 1966). The section preserved at Brown Ranch represents less than half of the stratigraphic thickness at Red Bird. At the base of the Sharon Springs Member is a 2.97 m thick sequence of bentonites (Figure 2). Gill and Cobban (1966) defined this unit as the base of the Sharon Springs Member in the southern Black Hills. At Brown Ranch, the unit consists of nine individual bentonite layers separated by shale. Based on geochemical considerations, Bertog (1997a, 1997b, 1998) grouped the lower seven bentonites (AB1-7 Appendix I) together and collectively referred to them as the Ardmore bentonite “succession.” Bertog (1997a, 1997b, 1998) separated the two thick bentonites at the top of this thick bentonite series (SB1 and SB2 in Appendix I), arguing that they are geochemically distinct from the Ardmore bentonite “succession” just below. Within the Ardmore bentonite “succession” is the Ardmore bentonite, origi-nally named by Spivey (1940) for a 1-meter thick bentonite layer being mined commercially at Ardmore, SD. At Brown Ranch the Ardmore bentonite (A3 in Appendix I) is 76.2 cm thick. Obradovich (1993) sampled the Ardmore benton-ite from the Clagget Shale in the Elk Basin of Wyoming, and using 40Ar/39Ar laser fusion, established a radiometric age of 80.54 mya +/- 0.55 Ma. The 40Ar/39Ar date provides a radiometric age at Brown Ranch that is located within the Bacu-lites obtusus range zone (Obradovich, 1993, Kauffman et al., 1993). Above the basal bentonite series (the Ardmore bentonite “succession” plus S1 and S2), the bulk of the Sharon Springs Member is 10.81 m thick. In the southern Black Hills, the shale directly above the basal bentonite series is typi-cally enriched with uranium (Kepferle, 1959). The shale of the Sharon Springs Member is dark gray, fissile and organic rich. The shale weathers to thin, silvery, papery chips, and forms steep-sided bluffs. The Sharon Springs Member, above the basal bentonite series, contains numerous small bentonites (29 bentonites above the Ardmore bentonite “succession”). Bentonites thicker than 2.5 cm above the basal bentonite series (SB6, SB21, SB26, and SB29 in Figure 2) serve as excellent local stratigraphic markers because they can be readily identified and traced throughout the exposures at Brown Ranch.


Four concretion layers (SC1-4 in Figure 2) are present in the Sharon Springs Member at Brown Ranch. All of the concretions within the Sharon Springs Member are pale gray, argillaceous limestone. The lower concretion layer (SC1) contains small (10-15 cm), sparsely exposed concretions just 2.31 m above the basal bentonite series. The remaining three concretion layers are situated in close association with each other. Concretion layer SC2 is composed of medium-sized (25-30 cm) concretions situated one meter below concretion layer SC3. Concre-tion layers SC3 and SC4 are separated stratigraphically by 60 cm of shale, and combine to form a vast sheet of pale gray concretions at the top of numerous cliffs dotting the landscape at Brown Ranch. The concretions are large (60 cm), and weather by fracturing into numerous small pieces, which may have the ap-pearance of a fractured turtle shell. Vertebrate and invertebrate marine fossils are abundant in the Sharon Springs Member (see Carpenter, 1996). Numerous marine vertebrate fossils have been recovered from this unit, and include fossils from the thick bentonite series, the black shale, and from the concretions (particularly SC3 and SC4). Invertebrates were not recorded from the Sharon Springs Member in this study, but they are known to occur, particularly in concretions (Kenneth Brown, pers. comm.).

Mitten Black Shale Member

Rubey (1930) first named the Mitten Black Shale Member for outcrops at Driscoll Creek in the northern Black Hills of Wyoming. At the type locality, the Mitten Black Shale Member unconformably overlies the Gammon Ferruginous Member, is 145 feet (44.2 m) thick, and is divided into lower and upper sec-tions (Rubey, 1930). The Sharon Springs Member is not represented at the type locality. At the southern Black Hills reference section in Red Bird, Wyoming, the Mitten Black Shale Member is 938 feet (285.9 m) thick, and conformably overlies the Sharon Springs Member (Gill and Cobban, 1966). The unit thins eastward, and is only 80 feet (24.4 m) thick along the Cheyenne River at Hot Springs, South Dakota (Gill and Cobban, 1966). At Brown Ranch the Mitten Black Shale Member is 11.48 m thick, and rests unconformably above the Sharon Springs Member. The unconformity is identi-fied in the field as an iron-stained, fossil-hash horizon 10-20 cm thick, lying 27 cm above the bentonite SB29 (about 1 m above the top of SC4). This uncon-formity has been recognized at other outcrops of the Lower Pierre Shale and can be correlated throughout the southern Black Hills (Martin, 1996). At Brown Ranch, the Mitten Black Shale Member is capped by a prominent tan cone-in-cone concretion layer, which is typically exposed along hilltops and covered with vegetation. The exposed portions of the Mitten Black Shale Member at Brown Ranch can be divided into lower and upper sections. The lower section is 2.63-2.73 m thick and is comprised of dark gray, dusty weathering shale. The lower section contains five bentonites (MB1-MB5; Figure 2), and invertebrate fossils have been observed. These flattened body fossils with preserved nacreous material differ from the stein kern preservation typically encountered in the upper part of the Mitten Black Shale Member. Staff and students at the Museum of Geology


have collected a number of marine vertebrate fossils from the lower portion of the Mitten Black Shale Member. The upper part of the Mitten Black Shale Member is 8.65 m thick and is comprised of dark gray shale with numerous iron-rich concretions comprising five distinct layers (MC1-5). These concretions are thin and broad (most are less than 10 cm thick and 15 to 75 cm wide), the weathering products of which stain the shale rust-red within the concretionary zone. The lowest concretion layer, MC1 forms a horizon of distinct concretions at some locations, whereas at other locations the concretions are interconnected and form a singular, tabular ironstone unit. There may be an unconformity at the base of this unit. None of the other four iron-rich concretion layers form a tabular mass such as MC1. A layer of thick, tan cone-in-cone concretions (MC6) caps the upper portion of the Mitten Black Shale. Near the top of the section, grass and other vegetation overgrows much of the shale, and the upper portion of the Mitten Black Shale Member is not well exposed at Brown Ranch. Internal casts of invertebrate fossils such as ammonites and inoceramid clams are extremely abundant in the upper section. Though most of the specimens are of poor quality, some of these fossils are identifiable, and specimens of the straight ammonite Baculites mclearni, the coiled ammonite Placenticeras cf. pla-num, and the clam Inoceramus have been identified and collected.

Tertiary Stream Deposits

Along the southwestern margins of exposures at Brown Ranch, a broad sheet of rounded cobbles covers two of the Pierre Shale members described above. The cobbles lie with angular unconformity across both the Sharon Springs Member and the Mitten Black Shale Member, and cobble size decreases slightly to the southeast. The cobbles in this unit appear to be derived from source rocks in the Black Hills.

PALEONTOLOGIC RESOURCES FROM BROWN RANCH

Carpenter (1996) compiled a faunal list of fossil vertebrate specimens from museums and the field observations in the Sharon Springs Member of the Pierre Shale (outcrops in Colorado, Kansas, South Dakota, and Wyoming). Carpenter’s faunal list includes 28 genera recorded from 539 museum specimens and 871 field observations. This list provides paleontologist working at Brown Ranch with a group of organisms with which to compare the local fauna in terms of completeness. Fossils recovered from the members exposed at Brown Ranch consist of ma-rine taxa representative of other locations of the Pierre Shale and are similar to that described by Carpenter (1996), though the diversity is predictably smaller. Table 1 presents a comparison of the fossils as documented by Carpenter versus those recovered from Brown Ranch. The total number of genera recognized from Brown Ranch is 13. The differences between the two studies are not sur-prising, in that the data set compiled by Carpenter is nearly 23 times larger, and


specimens examined by Carpenter were collected from a larger area, over longer periods of time, than has been done (or can be done) at Brown Ranch.

Table 1: Comparison of fossils recovered from the Sharon Springs Member at various localities (according to Carpenter, 1996) and those recovered from the Sharon Springs Member at Brown Ranch.

Taxon NUMBER OF GENERA

Sharon Springs Member (from Carpenter, 1996)

Sharon Springs Member at Brown Ranch

Class Chondrichthes Family Cretoxyrhinidae 1 0 Family Anacoracidae 1 1Class Osteichthyes Family Protosphyraenidae 1 0 Family Ichthyodectidae 3 1 Family Saurocephalidae 2 0 Family Plethodidae 1 0 Family Pachyrhizodidae 1 0 Family Crossognathidae 1 0 Family Cimolicthyidae 1 1 Family Enchodontidae 1 1 Family Dercetidae 1 0 Family incertae sedis 1 0Class Reptilia Family Toxochelidae 1 1 Family Polycotylidae 1 1 Family Elasmosauridae 2 1 Family Mosasauridae 5 4 Family Pteranodonidae 1 1 Family Hadrosauridae 1 0Class Aves Family Hesperornithidae 2 1Total Genera: 28 13

The relative frequencies of Brown Ranch fossils are shown in Figure 3. Fos-sils were grouped into the following nine categories: fish, turtles, mosasaurs, plesiosaurs, pterosaurs, birds, and indeterminate vertebrate remains. Given this distribution, four groups have a relative abundance above 10%: fish (25%), mosasaurs (35%), and birds (15%). The remaining four categories (turtles, ple-


siosaurs, pterosaurs, and indeterminate vertebrates) collectively represent 35% of the total number of fossils studied. Moreover, the combination of the two most abundant groups (fish and mosasaurs) comprises well over half (60%) the fossils recorded in this study. At this point, it should be noted that the database used to generate Figure 3 is heavily skewed by fossils collected by staff and students at the Museum of Ge-ology. For example, a survey of Museum of Geology collections includes a large number of mosasaurs, but very few fish. Carpenter (1996) recognized this same bias, noting that, for example, unidentifiable fish make up approximately 34% of bones observed in the field, while making up only 1% of museum specimens. Therefore, collection procedures must be recognized, and relative abundances of fossils as presented here should be considered biased approximations rather than absolute statements of abundance.

STATISTICAL ANALYSIS OF THESTRATIGRAPHIC DISTRIBUTION OF FOSSILS

Introduction

Fossils recovered from Brown Ranch have been documented as occurring in a large number of geographic as well as stratigraphic positions. Stratigraphic positions were taken from a total of fifty-three (53) fossils collected and/or iden-tified from Brown Ranch. The bulk of fossils (58%) have their stratigraphic occurrence between approximately 10 and 16 m above base, or only 22% of

Figure 3: Relative frequencies of all fossils recorded from Brown Ranch.


stratigraphic thickness. This portion of strata also represents a large amount of surface expression on Brown Ranch. Fossils recovered from above and below this package of rock are from units that have limited surface expression at Brown Ranch. This variation in the amount of surface expression of particular intervals of strata brings an element of bias into the data set presented here. In order to mediate this bias, the following method is employed. Rather than performing an analysis of the distribution of fossils among the entire strati-graphic column at Brown Ranch, the analysis is restricted only to that inclusive portion of strata that has produced fossil material to date. This results in an analysis of the strata from 0 m to 18 m above base. This is particularly useful because of the limited exposure and intense vegetation of the upper portion of the Mitten Black Shale Member (see above), which has thus far not produced any vertebrate fossil material.

Poisson Distribution Test for Randomness

The null hypothesis (H0) is that fossils from Brown Ranch are located randomly within the stratigraphic column of the units exposed. In contrast, a uniform or aggregated distribution of fossils indicates that fossils are distributed in some kind of pattern. A Poisson distribution test is employed to determine which distribution applies to the fossils at Brown Ranch. In order to statistically test the null hypothesis, the 53 fossils used in this study are placed into intervals (statistically termed quadrats) each representing a particular thickness of strata. Of the 27.36 meters of strata exposed at Brown Ranch, vertebrate fossils have been recorded from the base of the section to a height of 17.85 meters. Figure 4 illustrates the distribution of fossils according to 0.5 m quadrats, with reference to the stratigraphic column at Brown Ranch. The mean number of fossils (m) is calculated by dividing the total number of fossils by the number of intervals in which they have been placed. To compare these data against a Poisson (random) distribution, we first cal-culate the Poisson probability using the formula:

P(x) = (m x) [e-m] / x!

where: P(x) = probability of observing x individuals in an interval x = an integer counter; 0, 1, 2 … m = mean of the distribution x! = (x) (x-1) (x-2) … (1) and 0! = 1 by definition

The Poisson distribution is calculated by multiplying the Poisson probability by the number of intervals in each study.

Test for Goodness-of-Fit

To test the goodness-of-fit of the observed data against the expected results of the Poisson distribution, we use the index of dispersion test. This test was chosen over a Chi-square test on the basis of the greater sensitivity of the index


of dispersion test, which allows for fewer Type-II errors (Krebs 1999). The index of dispersion (I) is defined as the variance (s2) divided by the mean (m). If the distribution of organisms is truly random (a theoretical Poisson distribution), then the variance will equal the mean, and the dispersion will be equal to 1. To continue the test, a chi-squared test is then performed on the dispersion:

X2 = I (n - 1)

where: X2 = value of chi-squared with (n – 1) degrees of freedom I = index of dispersion n = number of intervals

The null hypothesis of random distribution is not rejected if:

X2.975 > Observed chi-squared value > X2

.025

Figure 4. Distribution of fossils at Brown Ranch according to 0.5 m quadrats. GFM = Gammon Ferruginous Member; BBS = Basal Bentonite Series; MBS = Mitten Black Shale Member.



The results of the Poisson distribution and index of dispersion tests are shown in Table 2. The data grouped in 0.5 m intervals have a X2 value of 351.59 with 36 degrees of freedom, and fall outside the X2

.975 and X2.025 critical values (30.57

and 53.2, respectively). The null hypothesis, random stratigraphic distribution of fossils, is therefor rejected. Because the variance is greater than the mean and the index of dispersion is greater than one, it can be concluded that the pattern of distribution of fossils is aggregated. The data grouped into 1 m intervals reveal a different pattern. The index of dispersion tests result in a X2 value of 17.66 with 18 degrees of freedom. This value falls squarely between the X2

.975 and X2.025 criti-

cal values (7.56 and 27.59, respectively). Here the null hypothesis is retained. At a resolution of 1 m, the fossils display a pattern of random distribution.

Table 2. Values of variables and result of index of dispersion and X2 test.

Interval Size 0.5 meter 1 meter

Number of Intervals (n) 36 18Mean (μ) 1.47 2.94Variance (s2) 14.77 3.05Standard Deviation 3.84 1.75Index of Dispersion 10.05 1.04X2 351.59 17.66X2

.975 20.57 7.56X2

.025 53.20 27.59

The seemingly contradictory results of the analysis of 0.5 and 1 m intervals can be explained by the differences of resolution. In each analysis, the total thickness of fossil-producing strata at Brown Ranch remains constant (18 m). As quadrat size is increased, the total number of quadrats decreases within the stratigraphic package, and the resolution of the analysis also decreases. Thus a resolution of 1 m may be too coarse to elucidate fossil distribution patterns at Brown Ranch. A resolution of 0.5 m, however, is capable of detecting aggregated patterns within this stratigraphic package. Further collections will determine which of these two possibilities is correct. The Poisson distribution test is only capable of determining whether or not a pattern of data is likely aggregated, random, or uniform in its distribution in space. If, as is the case for the data analyzed in 0.5 m quadrats, the Poisson test reveals an aggregated pattern, the significance is that there is an aggregated pat-tern. Clusters of data are as significant to this test as are absences of data. As a result, after determining this pattern, we must here enter a more qualitative analysis of the data set. At a resolution of 0.5 m, two quadrats show significant spikes in fossil content: 5.0 and 14.5 m (see Figure 4). Both of these quadrats have fossil contents greater than one standard deviation from the mean (Table 2). Less significant, but perhaps also promising, is the quadrat at 12.0 m above base,


whose fossil content is lies just within one standard deviation from the mean. Based upon this analysis, it is recommended that researchers working at Brown Ranch appropriate prospecting energies preferentially to these horizons that are significantly enriched in fossil material. Fortunately for the researcher, these quadrats, all in the Sharon Springs Member, are easily identified in the field: the interval at 5.0 m is located at and above the thick bentonite series; the interval at 12.0 m is located 0.5 m above SB26; the interval at 14.5 m is located within concretion zones SC4. As the database of fossil material continues to grow, more patterns may emerge from the data, and further studies could help to explain the patterns de-scribed here. For example, as the size of the database expands, a greater number of patterns of distribution may be recognized. Conceivably, a large enough data set may be able to determine particular distribution patterns among different taxonomic groups. A continued systematic and analytic search for fossil material at Brown Ranch may result in a greater understanding of the sedimentological processes and paleontology of the lower Pierre Shale.

CONCLUSIONS

Exposures of the lower Pierre Shale (Gammon Ferruginous, Sharon Springs, and Mitten Black Shale Members) at Brown Ranch are 27.36 m thick. A total of 53 vertebrate fossils have been documented from these exposures. Statistical analysis reveals that at a resolution of 0.5 m, two significant fossiliferous horizons exist at Brown Ranch, as well as a third possibility. These horizons are located at the following intervals: 5.0 m, 14.5 m, and 12.0 m, all within the Sharon Springs Member. These horizons should be targeted in field prospecting, since they have a greater likelihood of producing fossil material than other horizons.

ACKNOWLEDGEMENTS

The author wished to thank Kenneth Brown, who gave permission to con-duct this study of his land, and added his enthusiastic support and help in the field. Gale Bishop, Arden Davis, Jim Helsche, and Alvis Lisenbee provided help-ful comments on drafts of this paper. Corinna Ross greatly aided in fieldwork. The Museum of Geology at the South Dakota School of Mines and Technology permitted use of its collections and field notes.

LITERATURE CITED

Bertog, J.L., 1997a, Bentonite correlation by using trace element analysis of the Sharon Springs and Mitten Black Shale Members of the Pierre Shale (late Cretaceous), South Dakota. [Master’s thesis]: South Dakota School of Mines and Technology, 85 p.


Bertog, J.L., 1997b Bentonite correlation of the Sharon Springs and Mitten Black Shale Members of the Pierre Shale (Late Cretaceous), South Dakota, by trace element analysis: Society of Vertebrate Paleontology Abstracts of Papers, v. 17 no. 3, p. 32.

Bertog, J.L., 1998, High-resolution stratigraphic events in the Sharon Springs Member of the Late Cretaceous Pierre Shale, U. S. Western Interior; prelimi-nary investigations: Geological Society of America Abstracts with Programs, v. 30, no. 7, p. 338.

Carpenter, K., 1996, Sharon Springs Member, Pierre Shale (lower Campanian): Depositional environment and origin of its vertebrate fauna, with a review of North American Cretaceous plesiosaurs: Unpublished Ph.D. Dissertation, University of Colorado, Boulder, 251 p.

Darton, N.H., 1902, Description of the Oelrichs quadrangle [South Dakota-Nebraska]: USGS Geological Atlas, Folio 85, 6 p.

Dyman, T., Cobban, W.A., Fox, J.E., Porter, K.W., Hammond, R.H., Nichols, D.J., Perry, W.J. Jr., Porter, K.W., Rice, D.D., Setterholm, D.R., Shurr, Tysdal, R.G., G.W., Haley, J.C., and Campen, E.B. 1994, Cretaceous rocks from southwestern Montana to southwestern Minnesota, northern Rocky Mountains and Great Plains Region, in Shurr, G.W., Ludvigson, G.A., and Hammond, R.H., (eds.), Perspectives on the Eastern Margin of the Cre-taceous Western Interior Basin: Boulder, Colorado, Geological Society of America Special Paper 287.

Elias, M.K., 1931, The geology of Wallace County, Kansas: Kansas Geological Survey Bulletin, number 18, 254pp., 42 plates.

Gill, J.R., and Cobban, W.A., 1966, The Red Bird section of the Upper Creta-ceous Pierre Shale in Wyoming: USGS Professional Paper 393-A, 73 p., 12 plates.

Gries, J.P. and Martin, J.E., 1985, Composite outcrop section of the Paleozoic and Mesozoic strata in the Black Hills and surrounding areas, in Rich, F.J. (ed.), Geology of the Black Hills, South Dakota and Wyoming, second edi-tion: Rocky Mountain Geological Society of America Field Trip Guide, p. 261-292.

Kauffmann, E. G., 1984, Paleobiogeography and evolutionary response dynamic in the Cretaceous Western Interior Seaway of North America: in Wester-mann, G. E. G., Jurassic-Cretaceous Biochronology and Paleogeography of North America: Geological Association of Canada Special Paper 27, p. 273-306.

Kauffmann, E. G., Sageman, B. B., Kirkland, J. I., Elder, W. P., Harries, P. J., and Villamil, T., 1993, Molluscan biostratigraphy of the Cretaceous West-ern Interior Basin, North America, in Caldwell, W and Kauffman, E. (eds.) Evolution of the Western Interior Basin: Geological Association of Canada Special Paper, v. 39, 397-434.

Kepferle, R. C., 1959, Uranium in the Sharon Springs Member of the Pierre Shale South Dakota and northeastern Nebraska: U.S. Geological Survey Bulletin 1046-R, p. 577-604.

Krebs, C. J., 1999, Ecological Methodology, Fourth Ed.: Menlo Park, Benjamin Cummings, 620 p.


Meek, F. B. and Hayden, F. V., 1861, Descriptions of new Lower Silurian (Pri-mordial), Jurassic, Cretaceous, and Tertiary fossils collected in Nebraska territory, with some remarks on the rocks from which they were obtained: Proceedings of the Philadelphia Academy of Science, v. 13, p. 415-447.

Martin, J.E., 1996, Disconformities in the lower Pierre Shale (Cretaceous) of South Dakota, Geological Society of America Rocky Mountain Region Ab-stracts with Programs, v. 28, no. 4, p. 16.

Moxon, A. L., Olson, O. E., Searight, W. V., 1939, Selenium in rocks, soils, and plants: South Dakota Agricultural Experimental Station Technical Bulletin, No. 2, 94 p.

Obradovich, J., 1993. A Cretaceous time scale” in Caldwell, W and Kauffman, E. (eds.) Evolution of the Western Interior Basin: Geological Association of Canada Special Paper, v. 39, 389-396.

Rubey, W.W., 1930, Lithological studies of fine-grained Upper Cretaceous sedi-mentary rocks of the Black Hills region: USGS Professional Paper 165-A, p. 1-54.

Searight, M. V., 1938, The microfauna of the Sully member of the Pierre: Iowa Academy of Science Proceedings, V. 45, p. 135-137.

Spivy, R.C., 1940, Bentonite in southwestern South Dakota: South Dakota Geological Survey Report Investigations, v. 36, 56 p.

APPENDIX I

The following is a complete stratigraphic section of the Pierre Shale as ex-posed at Brown Ranch. The section records strata from the upper section of the Gammon Ferruginous, the entirety of the Sharon Springs, and the lower section of the Mitten Black Shale Members. The Sharon Springs is incomplete at Brown Ranch, this section records its thickness at Brown Ranch. Base is taken as position of MRR-01-01, along west side of creek bed to west of slope in Zone 2. This is the lowest recognizable data point within the Gam-mon Ferruginous Member of the Pierre Shale as exposed at Brown Ranch.

0.00 m Position of MRR-01-010.00 m Olive gray shale 1.14 m thick.1.14 m Pale yellow bentonite 0.6 cm thick. G11.15 m Olive gray shale 55 cm thick.1.70 m Pale yellow bentonite layer 0.3 cm thick. G21.70 m Olive gray shale 33 cm thick.2.03 m Pale yellow bentonite stringer 0.3 cm thick. G32.03 m Olive gray shale 7 cm thick.2.10 m Base of the Sharon Spring Member coincides with the base of the

Ardmore Bentonite series; 15.2 cm-thick bentonite layer, cream-tan when fresh, slightly rusty at base. A1

2.25 m Massive, black shale/claystone parting, 2.5-5.0 cm thick2.30 m Cream-tan bentonite layer 5.0-10.0 cm thick with an undulating

upper contact. A2


2.38 m Massive, black shale/claystone parting 7.5-12.5 cm thick; undula-tory contact with bentonite below.

2.48 m Cream-tan bentonite layer 76.2 cm thick; weathers off-white and yellow. This unit is the Ardmore bentonite of Spivey (1940). A3

3.24 m Small black shale/claystone parting 2.5 cm thick.3.27 m Bentonite layer 7.6 cm thick; weathers off-white. A43.34 m Black shale/claystone parting 10.1 cm thick.3.45 m Bentonite layer 10.1 cm thick; weathers off-white. A53.55 m Small black shale/claystone parting 3.5 cm thick.3.59 m 61 cm thick bentonite unit; lower 25.5 cm is orange and rusty, and

forms a somewhat resistant unit at top, chalky in consistency; the upper 35.5 cm is cream-tan, and weathers yellow. A6

4.20 m Small black shale/claystone parting 1.0-2.0 cm thick.4.22 m Rusty and tan-cream bentonite 20.3 cm thick. Top of Ardmore

Bentonite Series. Total thickness of the Ardmore Bentonite Series is 2.12 m. A7

4.43 m Black shale/claystone parting 4.0 cm thick.4.47 m Bentonite layer 25.5 cm thick; pale yellow when weathered, cream-

colored when fresh. SB14.72 m Black shale/claystone parting 7.6 cm thick.4.80 m Bentonite layer 28.0 cm thick; pale yellow when weathered, cream

and yellow colored when fresh; rusty and gypsumiferous at top. Top of thick bentonite series. SB2

5.07 m Black fissile shale 53 cm thick.5.61 m Tan-orange bentonite 2.5 cm thick. SB35.63 m Black fissile shale 86.5 cm thick6.50 m White-yellow bentonite 0.6 cm thick. SB46.50 m Black fissile shale 15 cm thick.6.65 m White-yellow bentonite 0.6 cm thick. SB56.66 m Black fissile shale 54.4 cm thick.7.20 m White-yellow bentonite 7.5 cm thick. SB67.28 m Black fissile shale 2.5 cm thick7.30 m Yellow-orange bentonite 0.6 cm thick. SB77.31 m Black fissile shale 7.4 cm thick.7.38 m Intermittent concretion layer; concretions are approximately 30 cm

in diameter, 10-15 cm thick, and weather a dusty tan-gray. SC17.48 m Yellow-orange bentonite 1.3 cm thick. SB87.50 m Black fissile shale 14 cm thick. 7.64 m Yellow bentonite 0.6 cm thick. SB97.65 m Black fissile shale 22.4 cm thick.7.88 m Black fissile shale 21.4 cm thick.8.09 m \ |------Bentonite couplet; bentonites are pale yellow and 0.3 cm thick SB11&128.12 m /8.13 m Black fissile shale 25.4 cm thick.8.38 m Pale yellow bentonite 0.6 cm thick. SB13


8.39 m Black fissile shale 12.4 cm thick.8.51 m Pale yellow bentonite stringer 0.3 cm thick. SB148.51 m Black fissile shale 6.7 cm thick.8.58 m \ |------Bentonite couplet; bentonites are pale yellow and 0.6 cm thick. SB15&168.64 m /8.65 m Black fissile shale 25.8 cm thick.8.91 m Pale yellow bentonite 0.6 cm thick. SB178.92 m Black fissile shale 27.4 cm thick.9.19 m \ |-----Bentonite couplet; bentonites are pale yellow and 0.6 cm thick. SB18&199.30 m /9.30 m Black fissile shale 5.8 cm thick.9.36 m Pale yellow bentonite stringer 0.3 cm thick. SB209.36 m Black fissile shale 42.7 cm thick.9.79 m Yellow bentonite 2.5 cm thick. SB219.82 m Black fissile shale 58.5 cm thick.10.40 m Pale yellow bentonite 0.6 cm thick. SB2210.41 m Black fissile shale 11.4 cm thick.10.52 m \ |----Bentonite couplet; bentonites are yellow-orange and 0.6 cm thick. SB 23&2410.56 m /10.56 m Black fissile shale 66.8 cm thick.11.23 m Yellow bentonite stringer 0.3 cm thick. SB2511.23 m Black fissile shale 25.7 cm thick.11.49 m Yellow/rust orange bentonite (cream-gray when fresh) 5.0 cm thick.

SB2611.54 m Black fissile shale 84 cm thick.12.38 m Base of concretion layer; concretions are 25-30 cm in diameter, 15-

20 cm thick, and are light gray and rust-colored. SC212.57 m Yellow bentonite 0.6 cm thick, located near top of concretions at

12.38 m. SB2712.58 m Black fissile shale 99 cm thick.13.57 m Base of concretion layer; concretions are approximately 60 cm in

diameter, 25 cm thick, and are light tan. These concretions fracture into small pieces, and form the top of small hills and rises along the flanks of the cliffs at Brown ranch; they also crop out as an extensive sheet of small mounds to the east of the cliff area. SC3

13.82 m White-yellow bentonite 0.6 cm thick. Located at top of above con-cretionlayer. SB28

13.83 m Black fissile shale 59.4 cm thick.14.42 m Base of concretion layer; same description as concretions at 13.57

m. SC414.67 m Black fissile shale 90 cm thick.


15.57 m Pale yellow bentonite (cream-gray when fresh) 4.5 cm thick. SB29

15.62 m Black fissile shale 26.5 cm thick.15.88 m Sharon Springs/Mitten Black Shale Member contact. This contact

is marked by an iron-rich zone of variable (10-20 cm) thickness. The top of the unit is composed of a fossil hash consisting of both invertebrate (largely ammonite fragments) and vertebrate material.

15.98 m Dark gray shale 54-64 cm thick.16.62 m Yellow-tan bentonite stringer 0.4 cm thick. MB116.62 m Dark gray shale 37.6 cm thick.17.00 m Orange-yellow bentonite 0.6 cm thick. MB217.01 m Dark gray shale 24.4 cm thick.17.25 m Pale white bentonite (cream-colored when fresh) 1.9 cm thick.

MB317.27 m Dark gray shale 5 cm thick17.32 m Variably white, yellow, and orange bentonite (gray-cream when

fresh) 3.8 cm thick. MB417.36 m Dark gray shale 1.31 m thick.18.67 m Rust-orange bentonite stringer 0.3 cm thick. MB518.67 m Dark gray shale 3.7 cm thick.18.71 m Iron-cemented, rust-colored resistant unit 5.0-10.0 cm thick; unit

is composed of tabular concretions approximately 75 cm wide, typi-cally in contact one with another (or very close). In some locations this forms a broad singular unit, in others the concretions are more noticeably individual. MC1

18.76 m Rusty gray shale 1.35-1.40 m thick.20.16m Concretion layer; concretions are 10-13 cm thick, dark rust red/ma-

roon; concretions vary in width from 50-85 cm. MC220.26 m Rusty gray shale 2.10-2.13 m thick.22.39 m Concretion layer; concretions are smaller (10 cm thick, 30 cm

wide), more orange in color. MC322.49 m Rusty gray shale 41 cm thick.22.90 m Concretion layer; concretions are thin and flat, rust-red on surface

and whitish inside; 7.5 cm thick and 15-25 cm wide. MC422.98 m Rusty gray shale 1.40 m thick.24.38 m Concretion layer; concretions are pale orange and small; 5 cm thick

and 15 cm wide. MC524.43 m Rusty gray shale 2.93 m thick.27.36 m Base of cone-in-cone concretion layer; this is an extensive unit

which caps the two eastern hills on Brown Ranch at the fossil local-ity. The concretions are orange-tan, approximately 20-30 cm thick, and are of varying widths (though they tend to form large sheets at the tops of the hills. MC6


ENVIRONMENTAL EFFECTS ON XYLEM CAVITA-TION IN SEEDLESS COTTONWOOD

Steven L. Matzner and Laura C. GoochDepartment of Biology


ABSTRACT

Water stress can cause xylem cavitation (aspiration of an air bubble into xylem conduits) which reduces the ability of plants to conduct water, limiting stomatal conductance and carbon gain. Environmental factors that alter xylem structure have the potential to alter drought tolerance in plants. The objectives of this study were to determine if various environmental factors affected resis-tance to xylem cavitation in seedless cottonwood (sterile crosses of Populus bal-samifera, deltoids, tristis, nigra, angulata, and petroskyana). One year old hybrid male clones were received from the Minnehaha conservation district in March of 2003. The trees were transplanted into control, high soil porosity, low soil porosity, shade, and low nitrogen treatments. Cavitation was induced in stems by centrifuging stem sections to specific tensions and measuring hydraulic conduc-tance (a measure of water flow) at 0, -0.1, -1.0, -1.5, -2.0, -2.5, -3.25, and -4.0 MPa of tension. All treatments exhibited a decline in hydraulic conductance with increasing tension. Plants from the various treatments did not differ in cavitation resistance. The results of this study are in contrast with several other studies where environmental effects caused changes in cavitation resistance (including another study on Populus). In this study, the variation between individual trees (perhaps related to the trees being crosses of six species) was greater than the variation between treatments and may have masked any potential treatment differences.

Keywords

Xylem cavitation, water stress, hydraulic conductance, Populus

INTRODUCTION

Xylem cavitation induced by water stress is common and can result in sub-stantial seasonal losses in xylem conductance (Alder et al. 1996, Tyree and Sperry 1989). Several studies have provided experimental evidence for the air-seeding hypothesis as the explanation for the mechanism of drought-induced xylem cavitation (Sperry and Tyree 1988, Tyree et al. 1994, Zimmerman 1983). Air seeding occurs when an air bubble is pulled into a water-filled conduit from an adjacent air filled conduit through pores in interconduit pit membranes in the


xylem conduit wall. The primary factor determining cavitation resistance is the size of pores in the interconduit pit membrane (Sperry and Tyree 1988). The size of the pores in the pit membrane determines the negative pressure at which bubbles would be aspirated, causing cavitation (Tyree et al. 1994). It is possible to induce cavitation by spinning a stem section within a centrifuge to specific tensions (Alder et al. 1997). Several recent studies have indicated that environmental factors can alter xylem cavitation resistance. A study by Harvey and van den Driese (1997) found that fertilizing Populus with nitrogen decreased cavitation resistance, presumably due to increased vessel sizes and larger intervessel pit pore diameters. In this same study, low phosphorus was found to decrease resistance by increasing the diameter of the intervessel pit pores. In a study by Cochard et al. (1999), trees growing under normal light were compared with shade acclimated individuals and it was found that the shade acclimated individuals were less resistant to cavitation. One problem with the study by Cochard et al. (1999) was that trees growing in the light might also have had greater water stress. All of the these studies were conducted “in the field” and as the Cochard et al. (1999) study illus-trated, it is sometimes difficult to assess whether changes in cavitation resistance are due to just one factor. The goal of this study was to determine if changes in cavitation resistance could be seen in trees grown under growth chamber condi-tions and whether it was possible to eliminate any “confounding effects” like those observed in the Cochard et al. (1999) study.


Experimental Design

In March of 2003, hybrid male clones of Populus (sterile crosses of Populus balsamifera, deltoids, tristis, nigra, angulata, and petroskyana) were planted in three-liter pots and grown in a growth chamber (Enconair, Winnipeg, Canada). The trees were transplanted into control, high soil porosity, low soil porosity, shade and low nitrogen treatments. For control treatments, growth chamber temperatures were maintained at approximately 25ºC, daytime and 23ºC at night. Humidity was maintained between 85-95%. A 14-hour light period was used with light levels maintained at 1000 μmoles*m-2-*s-1 at the top of the canopy. Soil media for control and shade treatments was composed of a potting soil, black soil, peat moss, sand, and perlite (1:1:1:0.5:0.5, by volume) mixture. Shade cloth was used to achieve a low light treatment of approximately 500 μmoles*m-2-*s-1. The low porosity treatment used only the standard black soil without potting soil, peat moss, sand, or perlite. Soil media for the high poros-ity soil was composed of fritted clay (DRI-RITE, Crestwood, Illinois, USA; van Bavel et al. 1978). All treatments were watered daily alternating between 1/10 Hoagland’s solutions and de-ionized water.


Cavitation Curves

Stem segments of Populus were cut under water to prevent embolism caused by the procedure. Shoot segments (approximately 14 cm in length) were flushed for 20 minutes at 0.1 MPa to refill any cavitated vessels. Stem preparation fol-lowed the protocols described by Sperry and Saliendra (1994). One end of the stem was hooked to tubing connected to a solution source and flow was induced using a raised medical IV bag. A filtered (0.2 mm) 10 mmolar KCl solution was used to perfuse the stem. The flow rate through the stem was measured as the amount of effluent collected during a timed interval. Measurement of hydraulic conductance was defined as the mass flow rate through the segment divided by the pressure difference, and all measurements of flow rate were expressed as a percent of the maximum flow rate. After measuring the maximum flow rate with all xylem cells conducting, the stems were exposed to a tension by spinning the stems in a centrifuge as described by Alder et al. (1997). The flow rate was again measured using the height of the solution bag to generate a small pressure difference (0.01 MPa). This procedure was repeated until the flow rate dropped by 95% to generate a complete cavitation curve. Cavitation was induced in stems by centrifuging stem sections to specific tensions and measuring hydraulic conductance (a measure of water flow) at 0, -0.1, -1.0, -1.5, -2.0, -2.5, -3.25, and -4.0 MPa.

Statistics

Treatment differences in vulnerability to xylem cavitation were analyzed using a repeated measures analysis of variance. Hydraulic conductance measure-ments on single stems at different levels of applied pressure was the repeated factor. Level of applied pressure was the within subject effect.

RESULTS

The loss in hydraulic conductance with increasing tension for all treatments exhibited an almost linear decline (Figure 1). For all treatments, conductance dropped to between 87-94% at tensions of -0.1 MPa. By -1.0 MPa, conduc-tance had decreased to approximately 65-74%. At -1.5 MPa of tension, conduc-tance for all treatments ranged between 50-61% and by -2.0 MPa 32-51%. At -2.5 MPa of tension, conductance ranged between 15-30% and by -3.25 MPa, conductance had dropped to 7-13%. By -4.0 MPa, conductance was very close to zero (1-2%) for all treatments. Although there did seem to be a tendency for the low nitrogen treatment plants to maintain a higher level of conductance at increased tension compared with the shade plants (implying greater cavitation resistance), this tendency was not statistically significant. In fact, none of the treatments exhibited significant differences in cavitation resistance. For all treat-ments variability was high, as can be seen by the large standard error bars. Large standard error bars can be the result of low replication, but 12 to 15 trees were sampled for all treatments in this experiment and there was not a significant


decrease in the standard error bars with increasing replication. This would seem to indicate that the high variability was not simply a function of low replication. One possible source of the observed variation could have been the fact that the trees were crosses of six different Populus species. The Minnehaha Conservation district makes the crosses and mixes them up, so we could not account for this potential variation source.

DISCUSSION

Within the literature, several environmental factors have been shown to alter xylem cavitation resistance on a number of different plants (including Populus). A study by Harvey and van den Driessche (1997) looked at the effects of nitrogen and phosphorus nutrition on xylem cavitation in P. trichocarpa and P. deltoides. They found that high nitrogen levels decreased resistance, which they suggested was correlated with vessel diameter and interconduit pore size. They also found that higher phosphorus nutrition increased cavitation resistance and suggested this was due to greater pore membrane strength. In contrast, a study by Ewers et

Figure 1. Effect of applied tension (MPa) on the percent of maximum conductance on five treatments in Populus clones. Control (circle symbol), High Soil Porosity (square symbol), Low Nitrogen (erect triangle symbol), Low Soil Porosity (diamond symbol), and Shaded (inverted triangle symbol). Each point represents the mean percent loss of con-ductance (±SE) for all measurements in that treatment (n = 12-15).


al. (2000) showed that Pinus taeda exhibited increased cavitation resistance when fertilized with both phosphorus and nitrogen. The combined fertilization made it difficult to evaluate whether nitrogen or phosphorus was more important in the effect. The effects of light on cavitation resistance in this study was not consistent with most other studies. In this study, light levels had no effect on resistance. Conversely, Cochard et al. (1999) reported increased resistance in Fagus sylvatica due to higher light levels while Lemoine et al. (2002) reported decreasing resis-tance with decreasing light (for both, higher light was related to higher resis-tance). It should be noted, however, that water stress might have been a factor in the Cochard et al. (1999) study since the high light treatment plants had more negative water potentials. One other study did match the results of this study. Lipp and Nilsen (1997) also did not see a change in cavitation resistance in Rhodendendron maximum with varying light levels. The effect of either high (increased) or low (decreased) soil porosity on the cavitation resistance of Populus in this study was not significant. This did not match the study by Hacke et al. (2000) which reported that increased soil poros-ity decreased cavitation resistance in Pinus taeda. Various environmental factors did not affect xylem cavitation resistance in seedless cottonwood (Populus spp.) in this study. In general, this did not match most of the results reported in the literature including a study looking at other Populus species. The high variation between individual trees may have contrib-uted to masking any potential treatment differences.

ACKNOWLEDGEMENTS

This work was supported through Augustana College Research and Artist Fund, and a USDA NRICGP 00-35106-9403 grant. The authors would also like to thank Kari Pabst and Jaime Phipps for help and support during this study.

REFERENCES

Alder, N.N., W.T. Pockman, J.S. Sperry, and S. Nuismer. 1997. Use of centrifu-gal force in the study of xylem cavitation. Journal of Experimental Botany 48:665-74.

Alder N.N., S.S. Sperry, and W.T. Pockman. 1996. Root and stem embolism, stomatal conductance and leaf turgor in Acer grandidentatum population along a soil moisture gradient. Oecologia 105:233-301.

Cochard H., D. Lemoine, and E. Dreyer. 1999. The effects of acclimation to sunlight on the xylem vulnerability to embolism in Fagus sylvatica L. Plant Cell and Environment 22:101-108.

Ewers, B.E., R. Oren, and J.S. Sperry. 2000. Influence of nutrient versus water supply on hydraulic architecture and water balance in Pinus taeda. Plant, Cell and Environment 23:1055-1066.


Hacke, U.G., J.S. Sperry, B.E. Ewers, D.S. Ellsworth, K.V.R. Schafer, and R. Oren. 2000. Influence of soil porosity on water use in Pinus taeda. Oeco-logia 124:495-505.

Harvey H.P., and R. van den Driessche. 1997. Nutrition, xylem cavitation and drought resistance in hybrid poplar. Tree Physiology 17:647-54.

Lemoine, D., H. Cochard, and A. Granier, 2002. Within crown variation in hydraulic architecture in beech (Fagus sylvatica L.): Evidence for a stomatal control of xylem embolism. Annals of Forest Science 59:19-27.

Lipp, C.C., and E.T. Nilsen. 1997. The impact of subcanopy light environment on the hydraulic vulnerability of Rhododendron maximum to freeze-thaw cycles and drought. Plant, Cell and Environment 20:1264-1272.

Sperry, J.S., and N.Z. Saliendra. 1994. Intra-and inter-plant variation in xylem cavitation in Betula occidentalis. Plant, Cell and Environment 17:1233-41.

Sperry, J.S., and M.T. Tyree. 1988. Mechanism of water stress-induced xylem embolism. Plant Physiology 88:581-7.

Sperry, J.S., and M.T. Tyree. 1990. Water-stress-induced xylem embolism in three species of conifers. Plant Cell and Environment 13:427-36.

Tyree, M.T., S.D. Davis, and H. Cochard. 1994. Biophysical perspectives on xylem evolution: Is there a tradeoff of hydraulic efficiency for vulnerability to dysfunction? IAWA Journal 106:1639-46.

Tyree, M.T., and J.S. Sperry. 1989. Vulnerability of xylem to cavitation and embolism. Annu. Rev. Plant Phys. Mol. Bio. 40:19-38.

van Bavel, C.H.M., R. Lascano, and D.F. Wilson. 1978. Water relations of frit-ted clay. Soil Science Society of America Journal 42:657-659.

Zimmermann, M.H. 1983. Xylem structure and the ascent of sap. Springer-Verlag, Berlin.


EFFECTS OF DIFFERENT FIRE INTENSITIES ON UNDERSTORY VEGETATION DIVERSITY IN THE

JASPER BURN AREA OF THE BLACK HILLS

Katie D. Derr and Steven L. Matzner Department of Biology


Lee A. VierlingInstitute of Atmospheric Sciences


ABSTRACT

Fire has become an increasingly important issue in western states because of the extent and severity of many recent wildfires. Years of fire suppression policy have resulted in higher fuel loads that cause fires to burn more intensely. Because these intense fires create conditions that may be uncommon in many forest eco-systems, it is important to monitor post-fire vegetation regrowth to understand the direction in which succession will occur. This paper presents a study on the effects of different fire intensities on the understory vegetation two years post-fire at the 84,000-acre Jasper Burn Area in the Black Hills of South Dakota. Our hypothesis was that high intensity fires would slow forest understory recovery, while low intensity fires might lead to increased diversity by stimulating the growth of more fire tolerant native species. The results, however, did not support our original hypothesis. Species richness and diversity were not significantly different between high and low intensity burn sites. In fact, the unburned sites had significantly higher species richness and diversity compared to either the low or high intensity burns. Analysis of individual species changes revealed that fire of any intensity resulted in the loss of fire intolerant species, but that fire did not appear to stimulate growth of fire tolerant native species. The absence of an increase in fire tolerant natives may be due to a lack of a seed source or individu-als to resprout. Recovery of natives in this system may occur more slowly than originally thought and may require additional burn events before recovery is complete.

Keywords

Fire intensity, understory, vegetation recovery, Pinus ponderosa


INTRODUCTION

Fire has been an important issue over the last few years as extreme wildfires have ignited each summer in western states. The increase in wildfires has also fueled the debate about forest management and the implications of burning or not burning. Research has shown that under natural conditions, ponderosa pine (Pinus ponderosa) forests need to burn every 10-25 years in order to maintain a “healthy” forest (Bock and Bock, 1984; Matthews, 2001; Weaver, 1967; Wright, 1978). It is thought that relatively frequent, low intensity fires (Fisher, 1985; Malakoff, 2002) burned pre-settlement ponderosa pine forests. Perhaps the most important issue with fire lies not in the question to burn or not, but in managing the intensity of the fire. A high intensity fire has more damaging effects than a low intensity fire. High intensity fires occur where there are higher fuel loads and these fires often reach the crowns of the trees (Bock and Bock, 1984). Fuel loads build up where fire is suppressed for many years, and forests often burn at such a high temperature in these areas that they scorch the soil and kill the older, larger trees that are normally resistant (Malakoff, 2002). Years of forest fire sup-pression in many Western states have created fuel conditions that, when finally burned, result in very intense fires that can severely damage forest areas (Fisher et al., 1985). Historically, ponderosa pine stand density was much lower than is typically seen today and as a result a much higher percentage of fire-tolerant grasses and forbs would have occurred in the understory (Fisher et al., 1985). These low intensity fires had the effect of reducing the fuel loads, decreasing ponderosa stand density, and increasing understory species diversity (Bock and Bock, 1984; Fisher et al., 1985). Species such as ponderosa pine and many of the native prairie species may depend on fire to remove competition and stimulate regeneration. However, when a high intensity fire destroys the seed source for ponderosa, regeneration of the trees may be very slow (Matthews, 2001). This study examines the effects of the 2000 Jasper Burn on understory vegetation recovery in a ponderosa pine forest within the Black Hills of South Dakota. As with many large fires, the Jasper Burn created a patchy landscape of fire severity; we therefore identified study sites that had experienced fire of high intensity, low intensity, or were unburned. The objective was to determine if the species in the low intensity burn sites had greater diversity compared to either the high intensity or unburned sites. Our hypothesis was that low intensity fires would result in higher diversity.

METHODS

The research took place in Black Hills National Forest near Custer, SD. Six sites were selected within the 84,000-acre Jasper Burn Area, that burned in August 2000. Three sites were classified as high intensity, three sites were low intensity and three unburned sites just north of the Jasper Burn were used as controls. All of the sites had a slope of five degrees or less. Approximately four sampling plots were established at each site. Plots were located at least 50 m from the nearest road, with a minimum of 25 m between plots. The plots were 25


X 25 m in size with six 1 m X 1 m subplots arbitrarily selected within the plots for understory vegetation surveys. Between July 7 and July 13 of 2002, photo-graphs of the understory vegetation were taken with a digital camera at each of the subplots within the plots at every site. The photographs were used to estimate plant species presence and abundance. The Shannon-Wiener species diversity index was calculated from the number of species and their relative abundance for each plot. Comparisons of species richness (number of species) and diversity (Shannon-Wiener index) between sites differing in fire intensity were made using ANOVA and two tailed t-tests.

RESULTS

Comparisons of species richness for different burn intensities indicated a significant effect of fire intensity on the number of species (P<0.005; Fig. 1). The unburned sites differed significantly from both the high intensity burn sites (P<0.01) and the low intensity burn sites (P<0.03). Both high intensity and low intensity fires appear to have reduced the number of species observed. The larger error bars observed from the data from the high intensity sites reflected a greater variation in the subplots from those sites (greater patchiness). Some subplots were completely bare while others had abundant plant growth (see Fig. 3 and Fig. 4).

Figure 1. Effect of burn intensity on the mean number of species (±SE) observed for each subplot (n=72) for high intensity, low intensity and unburned plots.


Figure 3. Vegetation recovery (or lack thereof) within a high intensity burn site two years post-fire within the Jasper Burn Area in the Black Hills of South Dakota.

Figure 2. Effect of burn intensity on average (±SE) diversity (Shannon-Wiener) observed for each subplot (n=72) for high intensity, low intensity and unburned plots.


The Shannon-Wiener index was calculated using both the number of species observed and the relative abundance of each species for each plot. The pattern observed for this measure of diversity matched the pattern observed for species richness (Fig. 2). There was a significant effect of burn intensity on diversity (P<0.02). Both the high intensity (P<0.03) and the low intensity (P<0.008) burns significantly reduced diversity compared to the unburned plots. In addition to changes in diversity, species composition also change between the different burn intensity plots and the unburned plots. In the unburned plots, Oregon grape (Berberis repens), bearberry (Arctostaphylos uva-ursi), bread-root scurfpea (Psoralea esculenta), ponderosa pine seedlings (Pinus ponderosa), and wild strawberry (Fragaria virginiana) were the most frequent species. The frequency of these species decreased dramatically with burning. In the low intensity burn plots the most common species was common yarrow (Achillea millefolium) with some bearberry (Arctostaphylos uva-ursi), and wild strawberry (Fragaria virginiana). Yarrow and wild strawberry tended to be in more shaded and mesic areas. Yarrow was not very prevalent in either the unburned or high intensity burn sites. A few of the more shaded subplots also had a small number of Oregon grape and bearberry. In the high intensity burn sites, only milkvetch (Astragalus spp.) were common. Standing milkvetch (Astragalus adsurgens) was consistently present in most of the subplots. Milkvetch dropped in numbers in the low intensity sites, but was slightly higher in the unburned areas.

Figure 4. Vegetation recovery at a high intensity burn site two years post-fire within the Jasper Burn Area in the Black Hills of South Dakota.


DISCUSSION

Our original hypothesis was that the low intensity burn sites would have higher diversity compared to either the high intensity or unburned sites. This hypothesis was based on the assumption that the low intensity fire would stimu-late growth and regeneration of more fire tolerant plants. In fact, both low and high intensity burning reduced species diversity, apparently by reducing species less tolerant of fire. These results are in contrast to a Western Australian study by Lewis (2003) which reported highest diversity 2-3 years post-fire. Other studies however show a variety of results, some of which are more consistent with our re-sults. Russell and Roberts (1996) did not see a difference between low intensity burns and unburned areas in a Eucalyptus community in southern Queensland. An interesting study by Elliot et al. (1999) saw a relationship between post-fire diversity and slope. Ridges exhibited an increase in diversity, mid-slope showed a decrease, and very shallow slopes did not change. Our study had a shallow slope and we observed a decrease in diversity. Although our study did not show an increase in more fire tolerant species, the sites did differ in species composition. A study by Bock and Bock (1984) also did not show an increase in diversity two years post-burn, although they were primarily looking at shrubs and trees. Most species they looked at either exhibited a slight decrease or no change between the pre-burn and the two-year post-burn census. The one exception was leadplant (Amorpha canescens) which exhibited a slight increase. Forest fire management will continue to be an area of intense debate in western states. In addition to research on fire management, it is important to monitor post-fire vegetation regrowth to understand the effects of fire intensity on recovery and the direction in which succession will occur. Although this study did not show differences in diversity between low and high intensity burns there were differences observed in species composition that may affect succes-sion at these sites. Although the high and low intensity burn sites did not differ in diversity, caution should be taken in concluding that fire intensity was not important due to the limited scope and time frame of this particular study. The lack of an increase in diversity in the low intensity burn may simply reflect a lack of individuals to resprout and/or seed source due to years of fire suppression.

ACKNOWLEDGEMENTS

This work was performed under NASA EPSCoR grant number NCC5-588, L. Vierling, PI. The authors would like to thank the many individuals within Dr. Lee Vierling’s lab and especially Rachel Smith for allowing Katie to work with them on their research, train her in field methods, and teach her the skills needed to help her put this study together. Thank you for your help and encourage-ment.


REFERENCES

Bock, J.H., and C.E. Bock. 1984. Effects of Fires on Woody Vegetation in the Pine-grassland Ecotone of the Southern Black Hills. The American Midland Naturalist 112:35-42.

Elliot, K.J., R.L. Hendrick, A.E. Major, J.M. Vose, and W.T. Swank. 1999. Vegetation dynamics after a prescribed fire in the southern Appalachians. Forest Ecology and Management 114:199-213.

Fisher, R.F.,M.J. Jenkins, and W.F. Fisher. 1985. Fire and the Prairie-forest Mosaic of Devil’s Tower National Monument. The American Midland Naturalist 117:250-257.

Lewis, J. 2003. Plant regeneration following fire in Bungendore Park, Bedfor-dale, Western Australia. Western-Australian Naturalist 24:37-72.

Malakoff, D. 2002. Arizona Ecologist puts Stamp on Forest Restoration Debate. Science 297:2194-2196.

Matthews, M. 2001. Bringing back the forest. American Forests 107:32-37.Russell, M.J., and B.R. Roberts. 1996. Effects of four low-intensity burns over

14 years on the floristics of a Blackbutt (Eucalyptus pilularis) forest in south-ern Queensland. Australian Journal of Botany 44:315-329.

Weaver, H. 1967. Fire and its relationship to ponderosa pine. Proc. Tall Tim-bers Fire Ecol. Conf., 7:127-149.

Wright, H. A. 1978. The effect of fire on vegetation in ponderosa pine forests. Texas Tech. Univ. Range and Wildl. Inform. Series No. 2, Coll. Agric. Sci. Publ. No. T-9 199. 21p.


FIELD EVALUATION OF PEDOTRANSFERFUNCTIONS TO ESTIMATE SATURATED

SOIL HYDRAULIC CONDUCTIVITY

Darrell W. DeBoerDepartment of Agricultural and Biosystems Engineering


R. G. TeKroneyGround Water and Drainage Group

U. S. Bureau of ReclamationDenver, CO 80225-0007

ABSTRACT

Obtaining accurate saturated hydraulic conductivity values for drainage system design purposes is time consuming and expensive. Bureau of Reclama-tion drainage engineers collected hundreds of soil characteristic data sets for the Lake Plain area of South Dakota. Pedotransfer functions based on a regression relationship of chemical and physical parameters were assessed for an ability to estimate saturated field conductivity values. Two functions explained just 18 and 28 % of the variation in measured conductivity values. One of the functions explained only 13 % of the variation in an independent data set.

Keywords

Pedotransfer function, hydraulic conductivity, field, measured, estimated

INTRODUCTION

Accurate measurements or estimations of saturated soil hydraulic conduc-tivities are critical for the optimum design of subsurface drainage systems. Drain spacing relationships are of marginal value without the availability of reliable conductivity data. Estimation of field hydraulic conductivity values for soils found in the Lake Plain area of the James River Valley (Figure 1) is the subject of this paper. An irrigation project, designated as the Oahe Unit, was proposed for the Lake Plain area (Bureau of Reclamation, 1973). One unique feature of the proj-ect was the simultaneous installation of a subsurface drainage system and a water distribution system before actual water delivery to the project area. Soils in the Lake Plain area were derived mainly from silty lacustrine (lakebed) sediments. Project lands were designated for areas where sediments varied from 3 to 12 m in thickness overlying a glacial till. Pre-delivery installation of the drainage system


was deemed necessary because drainage construction costs were anticipated to be prohibitive after the establishment of a water table due to unstable coarse silt materials and very fine sand located about 2 m below the soil surface. In response to the need for pre-water delivery installation of a subsurface drainage system, the Bureau of Reclamation assigned a team of drainage en-gineers to assess the drainage characteristics of Lake Plain soils. Since direct in-place measurements of saturated hydraulic conductivity is a laborious and expensive process, the Bureau initiated a special program to evaluate indirect methods for conductivity estimation based on physical and chemical soil pa-rameters (Bureau of Reclamation, 1974). The objective of this paper is to assess the feasibility of estimating field saturated hydraulic conductivity values from a pedotransfer function based on physical and chemical soil parameter values of Oahe Unit soils.

PREVIOUS WORK

Indirect methods of soil hydraulic conductivity estimation have been used for many years with initial approaches based on utilizing physical properties to estimate conductivity. Baver (1939) found a correlation between pore-size dis-tribution and hydraulic conductivity. Aronovici (1946) established a correlation

Figure 1. Field locations for Data Sets A and B within the boundary of the Lake Plain area.


between percent sand and hydraulic conductivity for silt loam to sandy soils in the Imperial Valley of California. Incorporation of detailed field description of soil structure along with soil pores and texture was used as the basis for placing soils into seven permeability classes ranging from very slow (less than 0.03 m/d) to very rapid (6.0 m/d or more) for soils at 182 locations in the USA (O’Neal, 1952). A more recent study by Suleiman and Ritchie (2001) dealing with the use of effective soil porosity (total porosity minus field capacity) to estimate saturated hydraulic conductivity shows a great deal of promise. Soil and water chemistry can also significantly impact the ability of water to move through soils. For example, Quirk and Schofield (1955) found that the proper ratio of electrolyte concentration of irrigation water to percent exchange-able sodium of soil was vital for stable soil conductivity. Hillel (1998) presented a summary of how solute concentrations can significantly affect soil hydraulic conductivity. The terminology “pedotransfer function” was presented by Bouma (1989) as “translating data we have into what we need”. Pedotransfer functions can be de-fined as predictive functions of certain soil properties from other easily, routinely, or cheaply measured properties (McBratney et al., 2002). Many pedotransfer functions have been developed to predict given soil properties for a geographi-cal region. Functions for predicting soil hydraulic properties have been given by Rawls et al. (1991) and Wosten et al. (2001). Saxton et al. (1986) used a multiple regression approach to develop an empirical relationship for the prediction of saturated hydraulic conductivity.

STUDY PROCEDURES

Bureau of Reclamation drainage engineers conducted the field drainage in-vestigations (Bureau of Reclamation, 1974). Drill crews initiated the field process by digging pilot bore holes that were logged and used to delineate areas where the subsoils appeared to be uniform for “in-place” hydraulic conductivity tests. One criterion used in the selection of a test site was that it must contain at least a 0.75 m horizon of uniform soil because the conductivity tests were conducted with a 0.6 m test zone and required a minimum of 0.15 m of the same uniform soil below the bottom of the test zone. Tests were conducted in soils below the crop root zone at depths from 1.5 to 5.0 m. Most of the field conductivity tests were conducted where water table or saturated soil conditions were not present. In this case, the shallow well pump-in test (constant head) was used to obtain saturated hydraulic conductivities from 10-cm diameter holes (Bureau of Reclamation, 1978). When a water table was present, single auger pump-out tests were used. Hence, the data sets used in this paper should be considered as pump-in data sets. Five centimeter diameter undisturbed soil cores were collected at each of the test sites. A portion of each core was used for physical and chemical analyses in the laboratory. All laboratory analyses were conducted by Bureau of Reclama-tion personnel in accord with standard analytical practices. A summary of the physical and chemical parameters used to describe 1963, 1971, 1972, 1973, and


1974 field soil data sets is presented in Table 1. These data sets were amalgamated into two unique data sets where Data Set A (sodium was expressed as sodium percent) contained the 1963, 1971, and 1972 data while the 1973 and 1974 data were placed into Data Set B (sodium expressed as exchangeable sodium percent). The two data sets were collected from two different areas within the Lake Plain boundaries (Figure 1). Table 1 soil parameters were used as independent variables in a step-wise multiple regression analysis to determine which, if any, of the pa-rameters can be used to estimate field conductivities. Soil textural classifications were in accord with USDA recommendations. Soil structure and consistency evaluations were made in the field.

Table 1. Physical and chemical parameters used to describe field soils.

Chemical parameters Electrical conductivity of saturation extract pH of saturation extract Sodium percent or exchangeable sodium percent of saturation extract

Physical parameters Sand, silt, and clay percent Soil structure Class (Ex: Coarse, medium, fine) Type (Ex: Platty, laminated, crumb) Grade (Ex: Weak, moderate, strong) Consistency (Ex: Loose, friable, very firm) Sample depth Soil texture (Ex: silt, silt loam, sandy loam)

FINDINGS

A step-wise multiple regression analysis of values in Data Set A was con-ducted using 372 measured hydraulic conductivities and associated physical and chemical parameters. Five independent parameters, in order of importance (percent sand, coarse structure class, structure consistency, percent sodium, and electrical conductivity) made significant (5 % level) contributions to the reduc-tion of sum of squares, which resulted in the following predictive relationship.

K=0.6096*(0.323+0.0112*S+0.250*CS–0.0927*C+0.159*Na–0.0102*EC) (1) where K = Saturated hydraulic conductivity (m/d) (range: 0.012 to 1.77 m/d) S = Sand (%) (range: 0.3 to 47.0 %) CS = Coarse structure (if absent = 1, if present =2) C = Structure consistency (range: 1 = loose to 3 = friable to 5 = very firm) Na = Sodium (%) (range: 0.0 to 24.0%) EC = Electrical conductivity (mmho/cm) (range: 0.16 to 15.0 mmho/cm)

The 0.6096 value was used to convert conductivity units from in/h to m/d. Hydraulic conductivity increased with an increase in percent sand and sodium


and when a coarse soil class was present. A decrease in hydraulic conductivity was caused by an increase in electrical conductivity and a change in soil consistency from loose to very firm. However, these five parameters could only explain 18 % of the variation in measured conductivity values. The standard error of esti-mate for equation (1) was 0.199 m/d. A graph showing the relationship between estimated conductivity values, based on equation (1), and measured conductivity values is presented in Figure 2. Data points tend to follow a horizontal trend line rather than the line of equality (1:1).

Results of step-wise multiple regression analysis of Data Set B, containing 539 measured hydraulic conductivities, were similar to the results summarized in equation (1). Again five independent parameters, in order of importance (per-cent sand, structure consistency, electrical conductivity, Sbk structure type, and exchangeable sodium percentage), made statistically significant contributions to the reduction of sum of squares as presented in the following relationship.

K=0.6096*(1.68+0.0218*S–0.168*C+0.0208*EC–0.970*Sbk–0.0120*ESP) (2) where K = Hydraulic conductivity (m/d) (range: 0.006 to 3.17m/d) S = Sand (%) (range: 1.0 to 71.0 %) C = Structure consistency (range: 1 = loose to 3 = friable to 5 = very firm) EC = Electrical conductivity (mmho/cm) (range: 0.4 to 19.5 mmho/cm) Sbk = Sub angular blocky structure type (if absent = 1, if present =2) ESP = Exchangeable sodium percent (%) (range: 0.0 to 24.5 %)

Figure 2. Estimated hydraulic conductivity values derived from regression equation (1) versus measured conductivity values using the combined 1963, 1971 and 1972 data sets (Data Set A).


Hydraulic conductivity increased with an increase in percent sand and elec-trical conductivity. It decreased with an increase in percent exchangeable sodium, when a Sbk structure type was present, and with a change in soil consistency from loose to very firm. These five parameters accounted for 28 % of the varia-tion in measured conductivity values and produced a standard error of estimate for equation (2) of 0.125 m/d. However, the arithmetic signs for the electrical conductivity and sodium related parameters are different in the two relation-ships. A graph showing the relationship between estimated conductivity values, based on equation (2) and measured conductivity values is presented in Figure 3. Data values tended to cluster around the line of equality (1:1) for measured values of less than 1.0 m/d but then followed a horizontal trend line for the few values greater than 1.0 m/d.

The ultimate test of a proposed estimation relationship is to evaluate its performance against an independent data set. Since the data sets are for two dif-ferent regions within the Lake Plain area and since two different sodium param-eters were used to assess the impact of sodium on conductivity in the developed predictive relationships, we do not have an unbiased avenue for the evaluation of a predictive relationship. However, if we consider that the sodium parameter in relationship (2) was responsible for reducing the total sum of squares by less than 1 of the 28 % (reduction becomes 27 %) associated with the relationship, removal of the sodium parameter from relationship (2) should provide some

Figure 3. Estimated hydraulic conductivity values derived from regression equation (2) versus measured conductivity values using the combined 1973 and 1974 data sets (Data Set B).


insight regarding the utility of using the relationship for estimation purposes. Re-lationship (2) without the exchangeable sodium percent parameter was then used to make conductivity predictions using Data Set A soil parameters. The results were far from satisfactory as the output of relationship (2) was able to describe only 13 % of the sum of squares of the measured Data Set A conductivity values. It appears that the chemical and physical parameters used in this study cannot be used to accurately predict saturated hydraulic conductivity of Lake Plain soils.

SUMMARY AND DISCUSSION

Two independent data sets composed of measured saturated soil hydraulic conductivities and physical and chemical soil parameters were used to evaluate an indirect method for estimating saturated hydraulic conductivity in the field. One data set was comprised of 372 data elements (collected in Spink County) and the second set contained 539 elements (collected in Brown County). Both sets were confined to the boundaries of the Lake Plain area of James River valley in South Dakota, and all measurements were made below the crop root zone. Percent sand and parameters related to soil structure were the most statistically significant soil parameters for the prediction of saturated hydraulic conductivity. However, no satisfactory predictive relationships were developed for the subsoils of the Lake Plain area. Recent literature indicates that inclusion of accurate descriptions of soil morphological characteristics is critical for the development of improved pedotransfer functions (McKenzie et al., 1991 and Lin et al. 1999). Techniques for proper quantification of soil structure information are still being developed.

ACKNOWLEDGEMENTS

This work was supported by the South Dakota Agricultural Experiment Sta-tion and approved as Journal Article No. 3439.

LITERATURE CITED

Aronovici, V.S. 1946. The mechanical analysis as an index of subsoil permeabil-ity, Soil Sci. Soc. Am. Proc. 11:137-141.

Baver, L. D. 1939. Soil permeability in relation to non-capillary porosity. Soil Sci. Soc. of Am. Proc. 3:52-56.

Bouma, J. 1989. Using soil survey data for quantitative land evaluation. Ad-vances in Soil Science 9:177-213.

Bureau of Reclamation. 1973. Final Environmental Impact Statement Initial Stage - Oahe Unit. US Dept. of The Interior, Denver, CO 80225.

Bureau of Reclamation. 1974. Unpublished Oahe Unit hydraulic conductivity data. US Dept. of The Interior, Denver, CO 80225.

Bureau of Reclamation. 1978. Drainage Manual. US Govt. Printing Office, Washington, DC.


Hillel, D. 1998. Environmental soil physics. Academic Press, San Diego. 771 pp.

Lin, H. S., McInnes, K. J., Wilding, L. P. and Hallmark, C. T. 1999. Effects of soil morphology on hydraulic properties: II. Hydraulic pedotransfer func-tions. Soil Sci. Soc. Am. J. 63:955-961.

McBratney, A. B., Minasny, B, Cattle, S. R. and Vervoort, R. W. 2002. From pedotransfer functions to soil interference systems. Geoderma 109:41-73.

McKenzie, N. J., Smettem, K. R. J. and Ringrose-Voase, A. J. 1991. Evaluation of methods for inferring air and water properties of soils from field morphol-ogy. Aust. J. Soil Res. 29:587-602.

O’Neal, A. M. 1952. A key for evaluating soil permeability by means of certain field clues. Soil Sci. Soc. Of Am. Proc. 16:312-315.

Quirk, J. P. and Schofield, R. K. 1955. The effect of electrolyte concentration on soil permeability. J. Soil Sci. 6:163-178.

Rawls, W. J., Gish, T. J. and Brakensiek, D. L. 1991. Estimating soil water retention from soil physical properties and characteristics. Adv. in Soil Sci. 16:213-234.

Saxton, K. E., Rawls, W. J., Romberger, J. S. and Papendick, R. I. 1986. Estimat-ing generalized soil water characteristics from soil texture. Soil Sci. Soc. Of Am. Jl. 50:1031-1036.

Suleiman, A.A. and Ritchie, J.T. 2001. Estimating saturated hydraulic conduc-tivity from soil porosity. Trans ASAE 44(2):235-239.

Wosten, J. H. M., Pachepsky, Y. A. and Rawls, W. J. 2001. Pedotransfer func-tions: bridging gap between available basic soil data and missing soil hydrau-lic characteristics. Jl. of Hydrology 251:123-150.

Abstracts of Senior Research Papers

presented at

The 89th Annual Meeting

of the



GROWTH OF NANO-CRYSTALLINEHYDROGENATED THIN FILMS WITH

EXTREMELY LOW HYDROGEN DILUTION

Yung M. HuhPhysics Department


Work Supported by NSF/EPSCoR Grant #EPS-0091948and the state of South Dakota

R. Shinar And V. L. DalalMicroelectronics Research Center

Iowa State UniversityAmes, IA 50011

ABSTRACT

A growth method of hydrogenated nano-crystalline silicon thin films was stud-ied using remote ECR-PECVD. Extremely low hydrogen dilution ratio to silane was employed for the high growth rate of crystalline hydrogenated silicon thin film. The dilution ratio was controlled from 7 down to zero. Nano-crystalline Si:H films were then investigated by means of Raman spectroscopy, x-ray dif-fraction, dark and photo conductivity, and activation energy measurements. The effect of systematic He addition to fixed dilution ratio was also discussed in terms of phase transition from crystalline to amorphous.


DEVELOPMENT OF A PCR PRIMER SETFOR THE HMW GLUTENIN GENES

EXPRESSED IN CHINESE SPRING WHEAT

Leslie M. Baehr and Michael K. WanousDepartment of Biology


ABSTRACT

The high-molecular-weight (HMW) glutenin genes, located on the group 1L chromosome arms, are a major determinant for baking quality in wheat (Triticum aestivum L.). We previously identified chromosome regions influenc-ing the expression of the HMW glutenins at the protein level (Wanous et al., 2003, Theor. Appl. Genet. 106: 213-220). Now we are extending this analysis to expression of these genes at the transcriptional level using real time reverse transcription polymerase chain reaction (RT-PCR). Primers and protocols for RT-PCR amplification of the HMW glutenin genes were designed. The ampli-fied products range in size from 175 to 409 bp and target transcribed segments of the genes, making them well suited to gene expression studies utilizing RT-PCR. For Glu-B1-1, Glu-B1-2, and Glu-D1-2, primers were designed from sequences in Genbank, available on the NCBI website, using Lasergene MegAlign software (DNASTAR) to find sequence differences between the HMW glutenin genes, and Lasergene PrimerSelect software to design the primers. The Glu-D1-1 prim-er sequences were obtained from D’Ovidio et al. (1995, Theor. Appl. Genet. 91: 189-194). Because the primers and PCR protocols are specific for each HMW glutenin gene, they can also be used for identification of the genes from genomic DNA samples.


LEVELS OF WATER-EXTRACTABLE N-NITROSO COMPOUND AND N-NITROSO COMPOUND

PRECURSOR IN 6 BRANDS OF SNUFF

Michael BellingDivision of Natural Science

Mount Marty CollegeYankton, SD 57078

Lin Zhou and Sidney S. MirvishEppley Institute for Research in CancerUniversity of Nebraska Medical Center

Omaha, NE 68198

ABSTRACT

Snuff conains substances known as water-extractable N-nitroso compounds (NOC) (Haorah et al., J. Agric. Food Chem. 49:6068, 2001). These NOC are derived from NOC precursors (NOCP). NOCP are amines or amides that are converted to NOC when they react with nitrite (are nitrosated). NOC obtained from nitrosation of NOCP in snuff were mutagenic in bacteria (unpublished). NOCP levels in snuff are much higher then NOC levels. We determined water-extractable NOC and NOCP levels in 4 commercial snuff brands (Grizzly, Timber Wolf, Skoal, and Copenhagen) and two non-com-mercial brands from Star Scientific (Ariva, and Stonewall) using a Thermal En-ergy Analyzer (TEA). In some tests, solvent-extractable tobacco-specific nitrosa-mines, e.g. NNN and NNK, and their amine precursors, e.g. nornicotine, were removed by dichloromethane extraction from aqueous solutions brought to pH 12, and water extracts were re-analyzed. Grizzly, Timber Wolf, Skoal, and Cop-penhagen snuff had total NOC levels of 320, 720, 230, and 110 nmol/g (mea-sured after adding sulfamic acid). After extraction from alkali, three brands had 20-32% losses in NOC, but Coppenhagen had a 14% increase in NOC level. NOCP levels were determined by treating water extracts with 110 mm nitrite and then with sulfamic acid. The four commercial brands had NOCP values of 13-16 µmol/g snuff. After extraction from alkali, the 4 brands showed a 27-51% loss of NOCP. Ariva and Stonewall snuff showed 0.36 and 0.46 nmol/g NOC and 14-19 µmol/g NOCP. Results suggest that water-extractable NOC and, perhaps, NOCP may help cause snuff-induced oral cancer, in addition to nitrosamines NNN and NNK, which should be extractable by dichloromethane. The Star Scientific brands, known to have very low levels of NNN and NNK, showed very low water-extractable NOC levels indicating their safety if these NOC are hazardous, but their high levels of NOCP could be a concern if they were nitrosated in the mouth.


SYNTHESIS OF DEUTERATED TRIS(2,2,6,-TETRAMETHYL-3,5-HEPTANEDIONATO)

EUROPIUM (III) (EUROPIUM THD)

Kathryn A. HenningDepartment of Chemistry


Annie Thompson, Allison Caster, and Mary T. BerryDepartment of Chemistry

University of South DakotaVermillion, SD 57069

Krisma D. DeWittDivision of Natural Science

Mount Marty CollegeYankton, SD 57078

ABSTRACT

Deuterated Tris (2,2,6,-tetramethyl-3,5-heptanedionato) europium (III) (europium THD) was synthesized in this study. The study also found ways to improve the efficiency of the synthesis process and increase the yield. It was theorized that the compound could be produced with only deuterated acetone and ethanol, with the rest of the reagents non-deuterated. This would greatly reduce the costs of producing this non-commercially available complex. This synthesis proved successful. The reaction required to make europium THD is a multi-step process. Overall, twelve acetone molecules go into making one europium THD molecule. There are seven steps involved in the synthesis scheme. The following sequence outlines the steps in the synthesis.

Acetone → Pinacol Hydrate → Pinacolone → Pivalic Acid → Pivaloyl Chloride → THD → Copper THD → Europium THD

A mass spectrometer was used to confirm deuterated product. In the future, the team will study the locations and decay dynamics of the excited electronic states of the deuterated europium THD compared to the non-deuterated com-pound in the gas phase. They will also study the nonradiative contributions to decay, which include charge transfer and electronic to vibrational energy transfer. Electronic energy from Eu+3 may be transferred to carbon-hydrogen and carbon-oxygen stretches. The C-H contribution can be eliminated by replacing the hydrogen atoms with deuterium. The deuterated europium THD is needed to study nonradiative energy contributions separately and analyze their significance. A tunable wavelength dye laser will be used to study the electronic states of the


two europium compounds. This material is based upon work supported by the National Science Foun-dation/EPSCoR Grant #EPS-0091948 and by the State of South Dakota.


DIRECT OBSERVATION AND TENSILE STRENGTH OF SOME POLYCARBONATE NANOCOMPOSITES

CONTAINING CARBON NANOTUBULES

Josiah Reams, Tsvetanka Filipova, Guy Longbrake, and David A. BoylesSouth Dakota School of Mines and Technology


ABSTRACT

Bisphenol A (BPA) polycarbonate is used in compact disks, automobile headlight lens covers, and jet fighter canopies owing to its high optical clarity, high impact resistance, and high tensile strength. Recently, our laboratory has synthesized a family of high aspect monomers similar in structure to BPA that have been polymerized to yield new polycarbonates. These monomers incor-porate additional phenyl rings in their molecular backbone providing tri- and tetraaryl monomers. Among these monomers is tetraaryl bisphenol A (TABPA). Polycarbonates with TABPA monomer units are expected to have improved mechanical properties, such as higher tensile strength since the molecular cross sectional area of higher aspect materials allows for the possibility of strengthen-ing mechanisms in the bulk polymer. Incorporation of nanoparticles, such as multiwalled carbon nonotubes (MWCNTs), has been demonstrated to modify the bulk mechanical properties of many polymers. Preparation of TABPA poly-carbonate with MWCNTs dispersed in the polymer will be described, and in-vestigations of MWCNT incorporation with scanning electron microscopy will be presented, as will methods and results for the determination of the tensile strength of TABPA polycarbonate nanocomposite.


THE SIMULTANEOUS EVOLUTION OF DEFENSE AND COMPETITIVENESS: NATURAL SELECTION

Riston Haugen and David SiemensBlack Hills State University

Spearfish, SD

ABSTRACT

Plants in nature that are attacked by herbivores and pathogens often occur in communities with other plants that represent potential competitors. Therefore, one would expect the simultaneous evolution of defense and competitiveness in many cases. However, optimality theory for plant defense evolution predicts an evolutionary tradeoff between competitiveness and defense. To test for the trad-eoff, we recently took a more direct approach than previous studies by examining the effects of genetic variation in competitiveness on defense expression. Our results were in apparent contrast to optimality theory. We found that competi-tiveness and defense were sometimes positively correlated, possibly because some plants deploy defenses that also function in competition. More interesting was our result on how plants may deploy defenses with putative dual functions. The plants switched from a toxin-based strategy to a growth-based compensatory (tolerance) strategy. Recent theoretical studies on the simultaneous evolution of resistance and competitive ability found that plants may switch to a toler-ance strategy when natural selection by competitors is greater than selection by herbivores. These patterns of selection may also have shaped the plastic switch that we observed. To estimate strength of natural selection we conducted path analysis and microsatellite genotype mean correlations. Results of path analysis were contradictory to recent theoretical studies, whereas results from microsatel-lite correlations depended on intensity of herbivory; at low intensities there was no difference, while at high intensities selection by competitors was greater than selection by herbivores, in accordance with theory.


SIMULTANEOUS EVOLUTION OFCOMPETITIVENESS AND DEFENSE:

INDUCED SWITCHING IN ARABIS DRUMMONDII

Tessa Jones, Shannon Kulseth, Karl Mechtenberg,Charles Jorgenson and David H. Siemens

BiologyBlack Hills State University

Spearfish, SD

Michael ZehfusChemistry

Black Hills State UniversitySpearfish, SD

Paul BrownEnvironmental Studies

Trinity Western UniversityLangley, B.C. Canada

ABSTRACT

Optimality theory for plant defense against herbivores predicts an evolution-ary tradeoff between the abilities to compete and defend. We tested this hypoth-esis by studying the effects of genetic variation in competitiveness on defense expression. Two closely related and differentially competitive congeners were compared for levels of resistance, tolerance, and secondary metabolite produc-tion. In a growth room experiment, plants of Arabis drummondii and A. holboel-lii were grown in the presence and absence of the common bunch grass Boutel-loua gracilis, the specialist herbivore Plutella xylostella and generalist herbivore Trichoplusia ni. Measures of defense varied inconsistently between the Arabis species, depending on type of herbivore, competition, and type of defense. The better competitor A. drummondii was more resistant to specialist herbivores, as in the field, and exhibited greater herbivore- and competition-induced changes in glucosinolate profiles. Further, when plants of A. drummondii were fed upon in competitive environments, the induced glucosinolate response was reduced while tolerance levels increased. We suggest that competitiveness and defense responses are sometimes positively correlated because some defensive traits also function as competitive traits. A competitive function for defenses may be why defenses were affected by competition. Alternatively, since the induced response did not increase total glucosinolate content significantly, minimal defense costs might also allow the simultaneous evolution of competitiveness and defense. Finally, when faced with both herbivory and competition, some competitive spe-cies, such as A. drummondii, may switch to growth-based rather than toxin-based strategies as recent theoretical models predict.


IDENTIFYING MICROSATELLITE MARKERS IN THE GENOME OF THE ENDEMIC ANTIGUAN

GROUND LIZARD, AMEIVA GRISWOLDI

Nathan T. Stephens, Brian E. Smith and Cynthia AndersonDepartment of Biology

Black Hills State UniversitySpearfish, SD

Paul ColbertDepartment of Ecology, Evolution, and Organismal Biology

Iowa State UniversityAmes, IA

ABSTRACT

The Antiguan ground lizard, Ameiva griswoldi, is found only on the islands of Antigua and Barbuda as well as many associated offshore islands. Populations of A. griswoldi have been fragmented on Antigua and occur at extremely high densities on some islands around Antigua. It is currently unknown how isola-tion and population density may affect the population genetics of A. Griswoldi. Our work focused on the development of a small fragment genomic library enhanced for microsatellites, a valuable tool for the study of population genetics. The purpose of this research was to locate several polymorphic base repeats or microsatellites in the genome of A. griswoldi. Genomic DNA was isolated from the tissues of A. griswoldi using the PCR-based isolation of microsatellite arrays (PIMA) method to screen samples for base repeats. We found that the PIMA method was effective in locating several repeat areas that may be useful as mic-rosatellite markers. DNA samples that showed positive results for base repeats were then sequenced to identify microsatellite arrays.


GENERATION OF gE-EGFP ANDgI-EGFP CONSTRUCTS TO STUDY

THE ROLE OF BHV-1 GLYCOPROTEINSgE AND gI IN BHV-1 PATHOGENESIS

Ihab Halaweish, Ehab Hassan, C. C. L. Chase,and Lyle BraunAnimal Disease Research and Diagnostic Laboratory

South Dakota State UniversityBrookings SD 57007

ABSTRACT

Bovine herpesvirus-1 (BHV-1) is one of the major ruminant herpesviruses. BHV-1 is an economically important pathogen of cattle and is responsible for a variety of clinical symptoms including rhinotracheitis, conjunctivitis, encephali-tis, and occasionally abortions. This translates to a loss of revenue due to dimin-ished milk and meat production. The key to understanding the complex patho-genesis of BHV-1 is a detailed understanding of BHV-1 glycoproteins. Research strongly indicates that BHV-1 glycoproteins gE and gI play an important role in intracellular transport and egress of BHV-1. Here we describe cloning of gE and gI as a fusion to Enhanced Green Fluorescent Protein (EGFP). The generated gE-EGFP and gI-EGFP constructs served as a tool for marking and visualizing the localization of BHV-1 glycoproteins gE and gI in bovine cells. BVH-1 gly-coproteins gE and gI were both found to localize in the peri-nuclear region of Madin-Darby Bovine Kidney cells (MBDK) and especially in the endoplasmic reticulum of the cells. Also, cytoskeleton staining determined that the gE and gI proteins associate very closely with microtubules in the cytoskeleton. In addition, Stable MBDK cell lines were generated that express gE-EGFP and gI-EGFP. These stable cell lines will serve as an essential tool for a more detailed understanding of the role of BHV-1 glycoproteins gE and gI in the intracellular transport and egress of BHV-1.


NON-CHLORAL-BASED SYNTHETIC ROUTESFOR THE PRODUCTION OF BPC MONOMERS

Rachel Waltner, Josiah Reams, Guy Longbrake,Tsvetanka Filipova and David Boyles


ABSTRACT

Bisphenol C polycarbonate (BPC-PC) is a polymeric material that possesses the unique properties of being both extremely fire resistant and self-extinguish-ing. Consequently, it has been considered for many fireproof applications, such as the interior panels of commercial aircrafts. In our interest to synthesize new, more ductile polycarbonates, we have undertaken to examine the effect of ad-ditional aryl functionality on bisphenol monomers by producing new triaryl and tetraaryl monomers for polycarbonate synthesis. Novel BPC-PC monomers have been synthesized through the introduction of additional aryl rings into the fundamental structure of the bisphenolic monomer. Previously, we reported synthetic routes for these monomers that relied on the use of chloral as one of the critical starting materials. This starting material, used in the original synthesis of tetraaryl BPC, has become nearly impossible to obtain, forcing us to find a new synthetic route utilizing a different starting material. In this presentation we report alternative, non-chloral- based synthetic routes that may enable us to ob-tain the desired BPC monomers. Additionally, the synthesis of a new monomer will be described, namely, the triaryl, asymmetric BPC. Synthetic routes for the asymmetric monomer along with results of the polymerization and characteriza-tion studies will be discussed.


NANOBIO-PLASTICS ANDCOMPOSITES FROM LINSEED OIL

AND SACCHARIDIC SOURCE MATERIALS

MichelIe R. While, Annie M. Thompson,David A. Bovles, Jon J. Kellar and William M. Cross

Materials Engineering and ScienceSouth Dakota School of Mines and Technology


ABSTRACT

Commercial resins are nearly entirely synthetic, poorly biodegradable and petroleum derived. Owing to these limitations, an investigation to determine the viability of novel bio-based resins both as stand-alone materials for poly-mer matrix composites as well as additives for commercial synthetic resins has been conducted. The objective was to ultimately evaluate the mechanical and thermal properties and compare the new materials and their additive counter-parts with the known epoxy vinyl resin, Derakane Momentum 470-300 (Dow Chemical). Syntheses have relied on commercially available flax seed fatty ac-ids and saccharidic source materials of which the latter have included sucrose octaacetate and β-cyclodextrin. Initial studies were conducted with methyl stearate as a model compound for interesterification reactions prior to the use of methyl linolenate. Syntheses of sucrose octastearate, sucrose octalinolenate and heptokis(2,3-0-linolenyl)β-cyclodextrin were performed and monitored by 1H NMR, 13C NMR and FTIR. Methyl stearate, sucrose stearate and methyl linolenate were compounded with Derakane, and polymerization was performed using dimethylaniline accelerator and benzoyl peroxide catalyst. Young’s Modu-lus has been determined to be lower relative to that of Derakane coupon stan-dards. Glass transition temperatures have also been determined by differential scanning calorimetry.


SYNTHESIS OF NOVEL TETRAARYLCOPOLYCARBONATES AND DETERMINATION

OF MARK-HOUWINK CONSTANTS

Guy Longbrake, Tsvetanka Filipova and David A. BoylesDepartment of Chemistry/Chemical EngineeringSouth Dakota School of Mines and Technology


ABSTRACT

Polycarbonates are a class of high performance engineering resins that dis-play high impact strength, thermal stability, and optical clarity. Research efforts throughout the world are dedicated to expanding the already superior properties inherent in this material, which finds a host of niche markets owing to its un-usual properties. The goal of this research was the synthesis of novel polycarbon-ates having improved properties over those of conventional polycarbonate (such as General Electric’s “Lexan”) for potential transparent, bulletproof applications for the military. The synthesis and preliminary structure-property relationships of different isomeric analogues of our novel materials will be presented. Three novel tetraaryl polycarbonates—the meta, ortho, and para analogues—will be discussed. The synthesis of each relied on the Suzuki reaction, and subsequent interfacial phosgenation was utilized to produce high molecular weight poly-carbonates. Problems encountered in determination of the Mark-Houwink constants by both manual and instrumental dilute solution viscosity experiments will be presented.


SOME EFFECTS OF LAND-USE MANAGEMENT PRACTICES ON SEED SET

OF CYPRIPEDIUM CANDIDUM

Carol WakeDepartment of Biology/Microbiology


ABSTRACT

During the summer of 2003, three different Cypripedium candidum (small white ladyslipper orchid) populations were examined during seed-set and ovary development stages of the life cycle for sexual reproduction success in relation to surrounding vegetation environments. Populations at Cottonwood WPA (Roberts Co), private land near Ortley (Roberts Co), and Lake Cochrane WPA (Deuel Co) grow in diverse micro-environments, particularly with respect to height and density of surrounding flora. By midsummer, flowers were senesc-ing and ovary development in pollinated individuals was evident. The Ortley and Lake Cochrane sites have not been cut or grazed for many years resulting in dense shade and competition for the intermingled orchids. The Cottonwood site, by contrast, was mowed late the previous fall (2002) so the orchids were growing in a much more exposed environment. To compare the pollination success between these different growing conditions, ovary development (success-ful pollination) vs. ovary abortion (no pollination) was recorded for every flower located. The Ortley and Lake Cochrane sites demonstrated only 22% (n=74) and 18% (n=615) seed set, respectively, while at the more open Cottonwood location, successful seed set was 44% (n=258). While orchids also reproduce vegetatively by underground rhizomes, sexual reproduction provides the critical genetic diversity to insure survival. Further study of flower pollination and cul-tural needs of this native orchid, in relation to land-use management practices, will be useful.


EMERGENCE DATE AFFECTS GROWTHAND FECUNDITY OF AMARANTHUS SPP

E. Uscanga-Mortera and S.A. ClayPlant Science Department


F. ForcellaUSDA-ARS

North Central Soil Conservation Research LabMorris, MN 56267

J. GunsolusAgronomy DepartmentUniversity of Minnesota

St. Paul, MN 55108

ABSTRACT

New cropping systems, such as minimum tillage coupled with the use of glyphosate-tolerant cultivars, could change the field biodiversity. Many annual species are escaping control because they are either 1) tolerant to glyphosate or 2) emerge late in the season after the last chemical application. Weeds that escape control may produce viable seed that continues the infestation. The amount of seed produced by a plant generally is a function of time of emergence and the amount of interference or competition exerted by surrounding plants. Seed production from escaped plants that germinate after the last herbicide ap-plication could be very low because the plants in the glyphosate systems are 1) late emerging and 2) competing with established and vigorously growing crop plants. Several Amaranthus species escape control under the glyphosate-tolerant system, but there is little or no information about the amount of seed produced by these plants. The objective of this 2-year study were to determine (a) the effect of simulated emergence date (transplanting date) on seed production of three species of Amaranthus (A. retroflexus, A. powellii, and A. rudis) that potentially could escape control under the glyphosate-tolerant system, and (b) the effect of soybean or corn competition on the growth and seed production of these three species “emerging” at different times. Corn and soybean were planted in a western Minnesota field on June 1. Greenhouse grown Amaranthus spp seedlings that were at the first-true-leaf stage of growth were planted into the crop areas on 4 and 18 June; and 2 and 16 July in 2001 and 11 and 25 June; and 9 and 23 July in 2002. Seedlings were spaced 25 cm apart along a 4-m transect and planted midway between two crop rows (76 cm wide). Amaranthus plants were measured periodically for height and plant diameter. Weed biomass and seed production were determined at the end of the growing season.


Estimated maximum seed production for A. retroflexus, A. powellii, and A. rudis were 8,200, 6,000, and 41,500, respectively when planted with corn and 23,700, 32,500, and 196,000 when planted with soybean. There were negative correlations between planting date and plant height, biomass, and seed produc-tion. Correlation values for planting date and seed production were: r = -0.92 (P<0.0001) for corn; R= -0.97 (P<0.0001) for soybean. Two to 18 times more A. retroflexus seed and one to 29 times more A. powellii seed were produced when plants were present in soybean than corn from zero to 30 days after crop emer-gence (DAE). When these species were planted later than 30 DAE, three to 46 more times A. retroflexus seed and three to 51 more times A. powellii seed were produced in corn than soybean. Three to seven times more A. rudis seed were produced when plants were present in soybean than corn from to 0 to 20 DAE. When this species was planted later than 20 DAE, two to 133 times more seed were produced in corn than soybean. These changes in the crop/weed interaction could be due to one or more factors including differential canopy cover between in corn and soybean late in the season, differences in N level or differences in surface available water. Late escapes of Amaranthus spp in corn will increase the weed seed bank more than late escapes in soybean.


DEVELOPMENT OF A METHOD FOREVALUATING THE YIELD GOAL APPROACH

K. Kim, D. E. Clay, C. G. Carlson, and S. A. ClayPlant Science Department


ABSTRACT

Yield goals have been used to determine N recommendations in South Dakota, North Dakota, and western Minnesota. However, some states, such as Wisconsin and Iowa have eliminated yield goals from N recommendations because of poor correlation between yield and economically optimum N rates. The objective of this study was to determine the feasibility of switching from a yield goal approach to a non-yield goal approach in South Dakota. Field ex-periments were conducted in Aurora, South Dakota between 2002 and 2003. Treatments were natural rainfall and natural rainfall + irrigation and four N rates (0, 60, 120, 180 kg N /ha). Plant samples were analyzed for 13C discrimination (Δ) and total N. Research results showed that; (i) adding N rates increased yield and Δ; (ii) applying supplemental irrigation increased yield and decreased Δ; (iii) yields were not influenced by an interaction between water and nitrogen; and (iv) δ15N values increased with irrigation and decreased with increasing N. These results suggest that nitrogen and water stress had independent impact on yield, and irrigation increased N mineralization. These findings partially support the hypothesis that fertilizer rates should be independent of yield goal. Research needs to be conducted to determine the long term impact of changing the rec-ommendation approach on soil N levels.


THE PRESENCE OF FLUORAPATITE IN PRISMATIC CARTILAGE FROM THE PERMIAN OF TEXAS

Tylor B. Sampson and Gary D. JohnsonDepartment of Earth SciencesUniversity of South Dakota

Vermillion, SD 57069

ABSTRACT

Prismatic cartilage in the Craddock bone bed from the lower Clear Fork Group (Early Permian) in Baylor County, Texas, belongs to a xenacanth shark (Orthacanthus platypternus), the only shark present in this fauna. The composi-tion of this fossilized cartilage is fluorapatite and quartz, based on X-ray diffrac-tion (XRD) and thin–section analyses. Under a polarized-light microscope, in-dividual prisms can be seen. Each prism consists of concentric layers dominated by quartz and fluorapatite. The prisms are surrounded by a matrix composed of calcite and dolomite. Prismatic cartilage from an earlier xenacanth shark from the Lower Permian Archer City bone bed has a similar chemical composition based on XRD. A third sample of prismatic cartilage from a ctenacanth shark (Ctencanthus amblyxiphias) from the Winfield Limestone in Morris County, Kansas, which is about the same age as the Archer City specimen, yielded a similar XRD result. This ctenacanth cartilage also has concentric layers of flu-orapatite and quartz, based on thin-section analysis. The fluorapatite may have been partially replaced by silica (quartz) during the fossilization process. This suggests that the composition of prismatic cartilage in the Craddock bone bed is not unique because of age or taxonomic affinity.


MACROFUNGI COLLECTED IN THEBLACK HILLS OF SOUTH DAKOTA

AND BEAR LODGE MOUNTAINSOF WYOMING FROM 1998-2003

A. C. Gabel , E. Ebbert, K. Lovett, S. Herrin, S. Mullen and D. WoolwineBiology Department

Black Hills State UniversitySpearfish, South Dakota 57799

ABSTRACT

The survey has documented 290 species of macrofungi collected from 110 sites in the Black Hills of South Dakota and Bear Lodge Mountains of Wyoming from 1998-2003. Two-hundred and sixty-nine are new reports for the region. Two-hundred and forty-four species were identified from seven selected perma-nent sites, each of which was visited four times each year it was in the survey. Five of these seven sites were in the survey six years, one site five years and one site four years. Six of the sites were selected on the basis of favorable predicted fungal diversity and one drier site was selected for comparison purposes. Areas for permanent sites ranged from 0.5 to 10.8 hectares and canopy cover was es-timated from 30-82%. Botany Bay, a narrow, moist canyon with dense vegeta-tion dominated by Picea glauca, Ostrya virginiana, Betula papyrifera and Pinus ponderosa had the highest fungal species diversity/ha/yr (30.67) and Alabaugh Canyon, a dry, open woodland dominated by Pinus ponderosa and Juniperus scopulorum had the lowest fungal species diversity (0.54). Statistical analysis showed a significant correlation between species diversity and canopy cover (r = 0.80; p = 0.03). Twenty-one percent of genera collected from the seven perma-nent sites are primarily mycorrhizal, 36% are soil/litter/dung saprobes and 52% wood saprobes. Numbers of newly collected species from five of the seven sites that were in the study for six years decreased each year, but 29 of the total 196 species were collected in 2003.


ANALYSIS OF MICROSATELLITE VARIATIONWITHIN AND AMONG POPULATIONS

OF THE TOPEKA SHINER, NOTROPIS TOPEKA

Shane Sarver and Cynthia AndersonBlack Hills State University

Spearfish, SD 57799

ABSTRACT

Populations of the Topeka shiner, Notropis topeka, have declined throughout the historic Midwestern range of this species and it is currently listed as endan-gered by the U.S. Fish and Wildlife Service. Significant populations of Topeka shiners are now restricted to the Flint Hills in Kansas, tributaries of the Missouri River in Missouri, and the Vermillion, Big Sioux and James rivers in South Dakota. Eight microsatellite primers were developed and used to genotype indi-viduals from each of 9 populations throughout South Dakota, Iowa, Kansas and Minnesota. Allele frequencies at each locus for each population were calculated and the data used to estimate the degree of genetic divergence among sampling locations by calculating Fst values. Multilocus Fst values indicated that moderate to high levels of genetic divergence exist between the South Dakota population and the Kansas populations. With relatively low levels of genetic differentiation between South Dakota populations and the populations from Iowa and Min-nesota.


3-D MODELING AND COMPUTER-GENERATED ARTICULATION OF

A TYRANNOSAURUS REX FORELIMB

Ellen Naito StarckDepartment of Geology and Geological Engineering


ABSTRACT

Casts of Tyrannosaurus rex specimen FMNH PR 2081 (Field Museum of Natural History), including the right scapula, humerus, radius, ulna, metacarpals I and II, phalanges II-1 and II-2, and unguals I-2 and II-3, were reproduced at one-sixth scale by means of Computer Tomography imaging and fuse deposi-tion modeling. CT scanning offered the most accurate representation of the originals, with little room for error resulting from data distortion. CT data files were imported into the Stratasys Insight 3.3 program, and the images were scaled, sliced, and oriented for subsequent tooling. A rapid prototype machine, the Stratasys FDM 3000, constructed the models and their respective support layers. The model material used was a clear, medical-grade ABS plastic, and the sup-port material was a substance of water-soluble composition. The materials were heated and extruded through separate steel tips, each layer one-ten thousandth of an inch (0.00254 mm). A minimum of five support layers per model was created to ensure horizontality. Due to the disproportionate size and curvature of the scapula, total building time was fourteen hours. Following construction, the models were placed in an ultrasonic bath to allow dissolution of the support material. Articulation of the individual pieces was accomplished using the engi-neering program, Solidworks. Benefits of non-contact reproduction techniques include no degradation of an original specimen due to molding. Shared CT data files allow more research-ers to replicate and study specimens without being compelled to travel to where specimens are housed, and computer-generated articulation allows for easy rota-tion and orientation of cumbersome specimens.


THE EFFECT OF CALCIUM-CHANNELBLOCKERS ON THE CLOSURE OF TRAPS OF

VENUS FLYTRAP PLANTS (DIONAEA MUSCIPULA)

Brenda Simon, J.C. Neilson, Erin Talsma and James SorensonDepartment of BiologyMount Marty College

Yankton, SD

ABSTRACT

Currently, the most plausible mode of conduction of impulses in plant cells involves the influx of Ca++ into the vacuole allowing the outflux of H+ ions into the cytoplasm (Wayne 1993). All of the ions involved are believed to move through membrane proteins which may be specific for certain ions or may allow the passage of more than one type of ion, but most of these membrane channel proteins have been little studied and are poorly characterized. The purpose of this research was to examine this model of ion exchange and membrane depolarization involving Ca++ ions in plant cells. Since the trap-closure response is easily induced and observable, Venus Flytraps were used as the study organism. To test the Ca++ model, two commercially available calcium-channel blockers (Verapamil and Diltiazem) were applied to the traps and the closure response was timed in comparison to untreated controls. Traps treated with the calcium-channel blockers were significantly delayed in closing with mean trap-closure time nearly double that for controls when the Ca-chan-nel blocker was applied at the base of the trigger hairs inside the trap, and nearly triple when the blocker was applied on the trap hinge. The results suggest that calcium-channels are involved in the conduction of the impulse in Venus Fly-traps, but that the response might involve other alternative ions as the exogenous application of calcium-channel blockers delayed, but never completely sup-pressed, the closure response.


QTL ASSOCIATED WITH MAIZE KERNELTRAITS AMONG ILLINOIS HIGH OIL X B73

BACKCROSS-DERIVED LINES

James Wassom46989 W. 57th St.

Sioux Falls, SD 57106

J. WongCalifornia Polytechnic State University

Horticulture and Crop Science

T. RochefordDepartment of Crop Science

University of Illinois

ABSTRACT

Illinois long-term selection strains of maize (Zea mays L.) have extreme levels for kernel oil, protein, and starch concentration (hereafter simply called oil, protein, and starch) and have been useful for identifying genomic regions controlling kernel composition, but they do not represent practical breeding lines. To identify kernel trait QTL in a more relevant genetic background, 150 BC1-derived S1 lines (BC1S1s) were produced from Illinois High Oil and recur-rent parent B73. Oil, protein, and starch were measured in the BC1S1s and in Mo17 testcross lines (TCs). Kernel weight of BC1S1s and grain yield of TCs were also determined. There were positive phenotypic correlations of starch with kernel weight in the BC1S1s (rp = 0.67**, α ≤ 0.01) and starch with grain yield in the TCs (rp = 0.59**). There were negative correlations of oil with kernel weight in the BC1S1s (rp = - 0.29**) and oil with grain yield in the TCs (rp = - 0.30**). Oil was negatively correlated with starch in the BC1S1s (rp = - 0.0.75**) and TCs (rp = - 0.66). A genetic map was created with 110 molecular markers. Multiple regression models with QTL detected by composite interval mapping explained 47, 45, 44, and 18% of phenotypic variance for oil, protein, starch, and kernel weight, respectively, in BC1S1s and 17, 23, 40, and 24% for oil, protein, starch, and grain yield, respectively in TCs. Fewer QTL were detected with TCs, but most mapped near QTL found in BC1S1s. QTL for multiple traits were often mapped within common intervals. For example, QTL for oil, protein, and starch in BC1S1s were mapped to a 22 cM interval on chromosome 6. This interval includes a QTL explaining 37% of the phenotypic variation for oil in BC1S1s. No grain yield QTL were detected in this region, suggesting that introgression of this oil QTL into an elite line might increase oil without an unacceptable loss in grain yield potential.


Keywords

Plant breeding, QTL, maize kernel, kernel oil, kernel protein, kernel starch


POLYMORPHISMS IN THE AGOUTI-RELATEDPROTEIN (AGRP) GENE IN PIGS

Juanita Perera and Nels H. GranholmDepartment of Biology/Microbiology


ABSTRACT

The agouti-related protein (AGRP) plays an important role in appetite regu-lation and energy homeostasis in mammals. Variations in the agrp gene (alleles) could be important determinants of appetite, basal metabolic rate (BMR), body composition, and other metabolic activities in livestock. Identification and selec-tion of optimal agrp alleles could be of benefit to livestock producers. In a previous study, we isolated and characterized the agrp gene in 16 do-mestic pig breeds and breed combinations. Our results showed that the coding regions of agrp are highly conserved among domestic pigs. Considering the important role agrp plays in appetite regulation, mutations in agrp’s coding regions maybe lethal. Also, due to intense selective breeding of pigs during the domestication process, agrp alleles may have been lost or become very rare. We hypothesized that wild or undomesticated pigs and pig relatives may carry agrp alleles and exhibit polymorphisms in the coding regions. We sequenced the agrp gene in several members of the pig family. A 978 bp region of the agrp gene was amplified using the polymerase chain reaction (PCR) and sequenced. The sequence results showed that polymorphisms exist within the coding regions of agrp in wild pigs. In the European wild boar (Sus scrofa), we found a C829T in exon 3. In the Sulawesi babirusa (Babyrousa baby-russa) we found G88C and G95C within exon 1, and G957T in exon 3. In the Southern bush pig (Potamocheorus larvatus) we report changes in exon 1 (T38C and C117T) and exon 3 (G820T, C933T, and C954T). In the Visayan warty pig (Sus cebifrons) we found one variation in exon 2 (G329A). Some of these nucleotide substitutions cause a definite change in the respective amino acids. Such variations could alter the functionality of AGRP causing major physiologi-cal changes in appetite, BMR, and body composition. Characterization of the agrp gene in pigs may help identify agrp alleles that produce animals with greater production efficiency and optimal carcass traits.


CONTROLS ON SEDIMENTATION IN THELATE CAMBRIAN DEADWOOD FORMATION

NEAR LEAD, SOUTH DAKOTA

Melissa M. Campbell and Christopher J. PellowskiDepartment of Geology and Geological Engineering


ABSTRACT

An outcrop of the basal Deadwood Formation was measured and described south of Lead, South Dakota along Whitewood Creek. At the base of the Dead-wood is a fluvial and tidally reworked, matrix-supported conglomerate unit with sub-rounded to angular pebbles that are imbricated in a northwest orientation. This, along with their composition, suggests a southeast provenance of quartzite, vein quartz, and phyllite. Overlying the conglomerate is compositionally and texturally mature quartz sandstone with bedding typical of beach deposition. Regional studies have shown that steeply dipping Precambrian quartzite and phyllite units formed topographic ridges and valleys that were flooded by a transgressing Late Cambrian sea. These topographic highs and lows controlled the distribution and thickness of the basal congolmerate unit.


ACTIVITY PERIOD OF FOUR SPECIESOF CARRION BEETLE ON THE

PINE RIDGE INDIAN RESERVATIONOF SOUTH WESTERN SOUTH DAKOTA

Louden Whirlwind HorseLittle Wound High School

Kyle, SDUniversity of Nebraska-Kearney

Kearney, NE

Daniel G. SnethenLittle Wound High School

Kyle, SDUniversity of Nebraska-Lincoln

Lincoln, NE

William Wyatt HobackUniversity of Nebraska-Kearney

Kearney, NE

ABSTRACT

Ecologically, species segregate overlapping niches through a number of mech-anisms including period of activity. One group which exhibits apparent activity pattern niche separation is the carrion beetles in the family Silphidae. Among this group Nicrophorus marginatus have been well-studied but little is known about the activity periods of N. obscurus and N. guttula. These species co-exist on the Pine Ridge Indian Reservation in Shannon County. We used 4 baited pitfall traps which were checked approximately every other hour for 73 hours between 19 August 2003 and 22 August 2003 to determine the activity of these species at our test site. A total of 292 carrion beetles were collected. Including 250 N. mariginatus, 20 N. guttula, 16 N. obscurus, and 6 Thanatophilus lapponicus. All of these species appear to be diurnal as no carrion beetles were collected during evening hours. All 4 species however were collected during the early hours of daylight and during late afternoon periods, showing a decrease in activity dur-ing the hottest part of the day. Our study site appears to be devoid of nocturnal carrion beetles. This observation requires further investigation as more eastern carrion beetle species assemblages have approximately 30% nocturnal species.


POPULATION DENSITY OF NICROPHORUSAMERICANUS THE FEDERALLY ENDANGERED

AMERICAN BEETLE IN THE NEBRASKA COUNTIES OF BLAINE, BROWN, LOUP, AND ROCK



Lincoln, NE


Kearney, NE

ABSTRACT

The decline of Nicrophorus americanus (ABB) across the eastern half of the United States is well documented. Listed as endangered in July of 1989, ABB cannot be reclassified until 3 populations have been discovered or established within each of four geographically assigned areas. We developed two ten day mark recapture surveys for ABB in northern Nebraska to quantify a newly dis-covered population. Four counties were sampled in June/July and August. We collected a total of 378 individual ABB. We used mark recapture model which assumes a closed population with equal probability of capture for all individu-als to the ABB population. Over the approximately 24.00 km2 area covered by the traps, we estimated a mean (±90% confidence interval) of 449 ± 86 ABB in Blaine County, 138 ± 33.4 ABB in Brown County, and 53 ± 21.5 ABB in Rock County. Although several ABB were captured from Loup County, our recap-ture data was not sufficient to establish a population density for ABB in Loup County, Nebraska. Together these data suggest a population of over 500 adult ABB from these four Nebraska counties, a criterion for the recovery plan.


THE EFFECT OF THREE TYPES OF LIGHT:HALOGEN, MECURY VAPOR AND ULTRA

VIOLET ON NOCTURNAL SPECIES OF CARRION BEETLE INCLUDING NICROPHORUS AMERICANUS THE FEDERALLY ENDANGERED

AMERICAN BURYING BEETLE



Lincoln, NE


Kearney, NE

ABSTRACT

The decline of Nicrophorus americanus (ABB) across the eastern half of the United States is well documented. ABB regularly appeared in entomological collections from the first half of the twentieth century and very sparsely since then. There are many hypotheses for this decline. One possibility is that ABB and other nocturnal insect species have been adversely affected by the electri-fication of North America. To test the potential impact of lights on carrion beetles including ABB a study site was established, in Southern Tripp County South Dakota, and three kinds of light (halogen, mercury vapor, ultraviolet), were tested. Arrays consisted of light + bait, light only, and bait only. A total of 112 (37 Necrodes surinamensis, 32 Nicrophorus americanus, 30 Nicrophorus orbicollis, 13 Nicrophorus pustulatus) nocturnal carrion beetles were trapped over 24 nights. Five of these carrion beetles were collected from light only traps in-cluding four ABB and one N. pustulatus. In all cases, where nocturnal carrion beetles were trapped in light only traps, the light being used was ultraviolet. This research shows the attractiveness of light even when carrion is present nearby and suggest that electric lights have played a role in the decline of nocturnally active carrion beetles.


DUST FRACTION IN CATCHMENT SEDIMENTS, PRAIRIE POTHOLE REGION, SOUTH DAKOTA

Richard Faflak and Jodie RamsayNorthern State University

Aberdeen, SD

ABSTRACT

Sediment core samples were collected from 17 wetland and marsh areas from portions of the Prairie Pothole Region of northeastern South Dakota during the fall of 2003. The marshes were located in Brown, Day, Clark, Codington, Hamlin, and Brookings counties. The catchments acted as sediment traps for wind blown sediments. Thus, a vertical profile of sediment was established. Ground penetrating radar was employed to gather data on depth to glacial till. The majority of profiles indicated that approximately one meter or less of sedi-ment overlay glacial till. The sediment size parameters were analyzed by moment statistics to derive mean grain size, kurtosis, and sorting coefficients. Results indi-cated that wind blown (eolian) sediment could be distinguished from fluvial and other sediment. This study suggested that considerable historic windblown sedi-ment may be related to road dust and recent agricultural practices. Additional implications of this study suggested that the pH of catchment deposits may have been raised by the addition of cations derived from dust.


THE EFFECTS OF pH CHANGES ON AEROBICDENITRIFICATION CONDUCTED BY

PSEUDOMONAS AERUGINOSA IN HYPERBARIC AND NORMOBARIC CONDITIONS

Ryan M. Klenner, Joshua J. Bathke, Sheena M. Benson, Jena R. Christianson, Ashley J. Hughes, Jessica A. Maschino, Amy M. Schmidt, Adam J. Smith,

Angela L. Steams, Paul G. Van Heukelom and William J. SoeffingNatural Science Area

University of Sioux FallsSioux Falls, SD 57105

Keywords

Aerobic denitrification, ATCC 27853, Hyperbaric, Normobaric, pH, Pseu-domonas aeruginosa

ABSTRACT

Subsidiary to continuing comparative investigations regarding the character-ization of Pseudomonas aeruginosa (ATCC 27853) cultured in hyperbaric condi-tions, this study examined the changes induced in aerobic denitrification when P. aeruginosa was subcultured in pH-adjusted nitrate broth media in normobaric and hyperbaric conditions. Nitrate broth media were adjusted to six pH levels, 6.6, 6.7, 6.8, 6.9, 7.0, and 7.1, using a 0.02 N NaOH solution. Standard low-evaporation 96 well cell culture cluster plates were used in bacterial cultivation and subsequent nitrite detection procedures. Cultures were incubated at 37 °C in humidified normobaric (101 kPa) and hyperbaric (161 kPa) air microenviron-ments, where the atmospheric partial pressure of carbon dioxide varied from 30 Pa to 500 Pa and the partial pressure of oxygen varied from 21 kPa to 35 kPa. The first step of denitrification, the conversion of nitrate to nitrite by nitrate reductase, was evaluated hourly during the fourteen-hour incubation period. A commercial colorimetric nitrite testing kit was used to detect the nitrite con-centrations at 0, 0.25, 0.5 1, 2, and 5 ppm. Positive, negative and contamina-tion controls were conducted in parallel with each sample. Experimental results suggest that variation of culture media pH within the 90% activity range of P. aeruginosa (pH 6.5-7.1), did not significantly influence the organisms’ abilities to reduce nitrate to nitrite via aerobic denitrification in normobaric (p=0.935) and hyperbaric (p= 0.588) conditions. Comparisons of normobaric and hyper-baric nitrate reduction models, developed consequential to this study, distinguish differing rates of reduction and residual accumulations of nitrite. Hyperbaric conditions delayed reduction by approximately two hours during the first eight hours of incubation and subsequently surpassed normobaric rates thereafter. Normobaric models depict an accumulation and continuation of 1 ppm nitrite


concentration after seven hours of incubation, suggesting that P. aeurginosa may select alternative electron transport systems under normoxic and hyperoxic con-ditions.


ARE THE NORTHERN GREAT PLAINS ANINLAND “COAST”? – FALL AGE STRUCTURE

OF NEOTROPICAL MIGRANT BIRDSIN SOUTHEASTERN SOUTH DAKOTA

Kurt L. Dean, Heather A. Carlisle and David L. SwansonDepartment of Biology


ABSTRACT

Age ratios of migrant birds at coastal stopover sites are often skewed toward juveniles relative to age ratios at inland stopover sites, a phenomenon called the “coastal effect.” The northern Great Plains form the western boundary of the migratory ranges for many Neotropical woodland migrants. Moreover, wood-land habitat is sparse in this region, so it might be expected that adults would avoid this region during fall migration, thus producing age ratios skewed toward juveniles, similar to those at coastal sites. We tested this “inland coast” hypoth-esis for Neotropical migrants captured during fall migration at natural riparian woodland sites and a farmstead woodlot site in southeastern South Dakota. For this analysis, we divided birds into four species groups, those having the west-ern boundary of their migratory range in South Dakota (SD), those having the western boundary in Montana/Wyoming (MW), and those for which South Dakota is central to the migratory range (Not Peripheral, NP), and compared age ratios among these groups. Neotropical migrants breeding in South Dakota (BR) were considered separately from those that occur only as migrants. The pooled age ratio for all Neotropical migrant species (including breeders) was 79.3 % juveniles (hatching-year, HY) in corridors and 86.5% HY in woodlots. For both corridors and the woodlot, SD (corridors = 81.7% HY, woodlot = 88.0% HY) and MW (corridors = 78.3% HY, woodlot = 90.3% HY) migrants showed higher percentages of juveniles than NP migrants (corridors = 71.9% HY, wood-lot = 84.8% HY), but the differences were not significant, although the MW-NP difference approached significance (P = 0.09) in the woodlot. The age structure of Neotropical migrants breeding in South Dakota was 84.2% juveniles in cor-ridors and 85.9% juveniles in the woodlot. The juvenile percentage was signifi-cantly greater than that for non-peripheral migrants in corridors, but not in the woodlot. The overall juvenile percentage for Neotropical migrants at our study sites was at the upper end of those recorded for inland sites in North America, but were lower than those for coastal sites. These data are only partially consistent with the inland coast hypothesis. Also of interest, the woodlot had a significantly higher percentage of juveniles than corridors for the overall Neotropical migrant population and for NP and MW migrants. This suggests that adults may either competitively displace juveniles from riparian habitats or that adults are better at finding natural riparian habitats than juveniles.


FALL STOPOVER DURATION ANDENERGETIC CONDITION OF THREE

SANDPIPER SPECIES IN WESTERNMINNESOTA AND EASTERN SOUTH DAKOTA

Nathan E. Thomas and David L. SwansonDepartment of Biology

University of South DakotaVermillion, SD

ABSTRACT

Differences in body mass and plasma metabolites of shorebirds are associated with fattening during stopover at large and small coastal mudflats, suggesting that differences affecting stopover ecology may occur between large and small wetland sites. In this study, we investigated stopover duration and plasma me-tabolite levels in shorebirds at several small natural wetlands and a large managed wetland (Big Stone NWR) in eastern South Dakota and western Minnesota. Stopover duration for Least Sandpipers, as measured by radio-tracking, averaged 6.2 ± 0.8 (SE) days at natural wetlands and 9.7 ± 1.6 days at Big Stone NWR. For Pectoral Sandpipers, stopover duration averaged 5.3 ± 1.1 days at natural sites and 6.6 ± 1.1 days at Big Stone NWR. Stopover durations in natural and managed stopover sites did not vary significantly. Mass did not differ signifi-cantly for shorebirds in natural and managed sites for any species or age class, suggesting that both natural and managed sites provide similar stopover habitat quality. Analyses of blood metabolites in Least Sandpipers showed a significant positive relationship between plasma triglycerides and energetic condition (mass/wing chord) (n = 38, P < 0.001, R2 = 0.48) and a significant negative relationship between plasma β-hydroxybutyrate and energetic condition (n = 28, P < 0.001, R2 = 0.33). In addition, plasma creatine kinase activity approached significance with lower levels at higher energetic condition (n = 23, P = 0.20, R2 = 0.08). These data suggest that Big Stone NWR and the surrounding natural wetlands serve equally well as stopover sites for sandpipers.


OAK LAKE FIELD STATION AS A MODEL FORETHNOBOTANICAL RESEARCH IN THE PRAIRIE

POTHOLE REGION OF SOUTH DAKOTA

Gretchen McClintock-Ames and R. Neil ReeseDepartment of Biology & Microbiology

South Dakota State UniversityBrookings, SD, 57007

ABSTRACT

Prairie pothole lakes are invaluable refuges for plants that were important to the survival of the region’s indigenous peoples. Oak Lake provides a unique meeting of biomes; wooded slopes, open prairie expanses, and lush littoral marshy zones. Ethnobotanical accounts were examined to see how plants were traditionally utilized by Native Americans and pioneers. Plants were identified using scientific names and the names given to the plants by the two groups of indigenous peoples most recently associated with this region, the Dakota and Omaha-Ponca. This was done because the indigenous languages describe char-acteristics of the plants much the same way Latin does, imparting insights into their perceived values. We experimented with traditional recipes for the prairie turnip; milkweed buds, forest greens, cattail roots, and various fruits. The Plants were collected, herbarium mounts made and field data recorded. Research was done on the archeology profile of the area, and on the different tribes who had used the hills, plains and beaches along the lake. The archeological record and history at Oak Lake spans prehistoric times and includes vintage pioneer fruit orchards typical of the early farms in the area. There is value in preserving prairie pot holes because of the wealth of plants and forbs they preserve. These native plants may again prove to have value as food and medicines for future genera-tions.


METHOD DEVELOPMENT FORTHE DETERMINATION OF

NEOMYCIN IN ANIMAL FEED

Amanda Dupay, Nancy Thiex, Dave Ferris and Douglas E. RaynieDepartment of Chemistry and Biochemistry


ABSTRACT

Aminoglycoside antibiotics are routinely added to animal feeds to prevent and combat illness. We developed a method to determine the amount of apra-mycin, neomycin, and gentamycin, as representative aminoglycoside antibiotics, in animal feed. The current standard method for determination is based on a microbiological assay. There are many disadvantages of the standard method, in-cluding time, precision, and lack of adequate selectivity (i.e., the method did not distinguish between different aminoglycosides, but looked at all aminoglycosides in general). Our method consisted of two steps. The first was a solvent extraction of neomycin from feed. Two different extraction solvents were tested, as well as the effect of heating the sample to denature the protein. Results showed that extracting with NaCl-CaCl2 solution and heating the sample provided the best recovery. Next, the extracted antibiotics were separated using ion-exchange liquid chromatography and identified with post-column derivatization and fluo-rescence detection. Our results showed recoveries of 90-100% of antibiotics in powdered milk replacer at 350-1000 g/ton level with improved precision. Cali-bration curves of neomycin and gentamycin were linear in the tested range with correlation coefficients of 0.9983 and 0.9995 respectively. Other advantages of our method included replacement of the 16-18 hour microbiological assay with a 30-60 minute liquid chromatograph (LC) run, with minimal sample handling, and the opportunity for speciation of the antibiotic.


PRIMARY GUSTATORY CENTERSIN THE ZEBRAFISH BRAIN

Judy Locati, Matt Nehl and Charles LambBlack Hills State University

Spearfish, SD

ABSTRACT

The Cyprinidae are a group of fishes with highly developed gustatory sys-tems. Most notable among these are the goldfish and carps, which possess elabo-rate neural structures in the hindbrain that receive gustatory input from numer-ous taste buds in the oropharyngeal cavity. This sensory input travels through the facial nerve from the rostral oral cavity and through the glossopharyngeal and vagal nerves from the pharynx. The gustatory afferents terminate in highly organized enlargements of the dorsal medulla called the facial, glossopharyngeal, and vagal lobes, respectively. We are studying the gustatory system of a less taste-specialized cyprinid, the zebrafish, in order to understand the morphologi-cal, functional, and evolutionary bases for the development of complex neural structures. The gustatory lobes of the zebrafish medulla are similar to those in the more specialized carps and goldfish. Facial afferents terminate in a single, midline fa-cial lobe, while glossopharyngeal and vagal fibers terminate in paired glossopha-ryngeal and vagal lobes. The zebrafish vagal lobe possesses a superficial sensory zone for afferent input, and a deeper motor zone containing motor neurons which project to muscles in the pharynx. The sensory zone appears to lack the degree of laminar organization seen in other cyprinids, but is still highly orga-nized. The gustatory lobes possess ascending connections similar to the carp and goldfish, with overlapping input to the secondary gustatory nucleus and distinct projections to the diencephalon.


EFFECT OF A SPRING-TIMEDEWORMING PROGRAM ON

STRONGYLE EGG OUTPUT IN WEANEDCALVES FROM EASTERN SOUTH DAKOTA

A.F. Harmon, W.B. Epperson and M.B. HildrethDepartments of Biology/Microbiology & Veterinary Science


ABSTRACT

Many South Dakota cattle producers are adopting spring-time deworming programs in attempts to limit economic losses caused by strongyle nematodes in their grazing cattle. These programs are designed to limit contamination of pastures with strongyles juveniles for the next grazing season. There have been no studies in South Dakota that have measured the effectiveness of these programs at limiting pasture contamination. Therefore, a study was conducted during the fall of 2003 to compare the egg-output of spring-born calves from herds whose cows had been treated with Dectomax (or Ivermectin) at the begin-ning of the summer grazing season. Two veterinary clinics from eastern South Dakota selected cattle producers whose herds were either using a spring-time deworming program or not using this type of program, but only samples from 5 treated herds and 4 untreated herds were received. For each herd, 20 fecal samples were collected from randomly-selected calves at the end of the grazing season (September-October). The samples were stored at 4ºC until analyzed for the presence of strongyle nematode eggs using a standard sugar floatation technique. The mean number of eggs for the untreated herds ranged from 5.5 to 41.6 eggs/gram (EPG). The mean egg output for these 4 untreated herds was 27.15 (S.D. 16.36). The mean number of eggs for the treated herds ranged from 0.0 to 18.6 EPG, and the mean egg output for the 5 treated herds was 6.50 (S.D. 7.53) EPG. These values were not quite significantly different at the p<0.05 level (p=0.064) based upon a Mann-Whitney statistical T-Test.


SEASONAL FLUCTUATIONS IN ADULTMOSQUITO POPULATIONS IN EASTERN

SOUTH DAKOTA DURING THE SUMMER OF 2003

J.R. Bradley, K. Dahmash, R. Beyer and M.B. HildrethDepartments of Biology/Microbiology & Veterinary Science


ABSTRACT

An adult mosquito survey was conducted during the summer of 2003 fo-cusing on collection sites in 6 different locations in eastern and central South Dakota. Mosquitoes were collected using battery- operated CDC Miniture Light Traps (John W. Hock Company, Gainesville, Florida 32604; Model 512). Each trap was equipped with a photoswitch and air-actuated gate system, and was baited with CO2 regulated through a photocell-controller. Trapping was conducted most aggressively at Brookings (May 18 to October 2, 2003; 118 trapping days [TD]). Starting and ending dates varied for the remaining sites, and collections were made less frequently (Huron = 42TD, Oak Lake Field Sta-tion =61TD, Pierre A =21TD, Pierre B =23TD, Pierre C =19TD, Sioux Falls =77TD, and Watertown =91TD). A total of 63,643 mosquitoes were collected from the 6 sites, and 18 different mosquito species from 7 different genera were identified. The vast majority of mosquitoes were Aedes vexans (85.1%) and Culex tarsalis (10.14%). The only other species accounting for more that 1 percent of the population was Ochlerotatus dorsalis (1.1%). Watertown accounted for 36.4% of the mosquitoes and averaged 254.6 mosquitoes each trapping day. The Oak Lake Field Station only averaged 81.3 mosquitoes/trapping day. The highest daily average of Culex tarsalis, the primary vector of West Nile in the western U.S.A., was collected in Huron (36.5 mosquitoes/day). The Oak Lake Field Station site only averaged 1.3 Culex tarsalis per day. The greatest increase in mosquito collections occurred during the second and third week of July. The populations begin decreasing during the second week in August.


ATTEMPTS TO LOCATE CULEX TARSALISLARVAE IN TRADITIONAL MOSQUITO-

BREEDING HABITATS FROMBROOKINGS COUNTY, SOUTH DAKOTA

M.L. Hart, D.J. Thorpe and M.B. HildrethDepartments of Biology/Microbiology & Veterinary Science


ABSTRACT

A total of 33 potential mosquito breeding sites were identified and selected during the early spring of 2003. These mosquito habitats were selected from both natural and urban settings within Brookings County, Sd, and represented the various classifications of habitats potentially used by various mosquito species for larval development including: flowing streams, ponded streams, lake edges, swamps and marshes, shallow-permanent ponds, shallow-temporary pools, inter-mittent-ephemeral puddles, natural containers and artificial containers. These sites were monitored weekly during the summer for the condition of the water and sampled for the presence of mosquito larvae. A standard collection of 10 “dips” using a 13 cm mosquito dipper was made from each location. Collected larvae were preserved (70% ethanol), dehydrated in an ethanol series, mounted on glass slides, identified and enumerated. Focus was given to Culex larvae because Culex tarsalis is a primary vector for West Nile Virus in South Dakota. Moderate numbers of larvae were recovered from many of the sites during the summer, but by August 6 only 21 of the sites contained water, and therefore, still able to support larval development. Culex tarsalis larvae were not found in any natural site; the only sites they where they could be found were 6 artificial containers possessing these larvae only after August 5. Culex tarsalis larvae were found in 2 locations that were not part of the original study sites. These included larvae found during early summer in an artificial barrel mesocosm within Oak Lake, and larvae found in a water-filled, shallow tire-imprint located in a hay-field in Minnehaha County, South Dakota.


CHARACTERIZATION AND COMPARISONOF CELLS BY ANTIGEN PRESENTATION,

MICROFILAMENT ORGANIZATION,AND PHAGOCYTIC PROPERTIES

N. Harms, B. Tigabu, L.J. Braun,E.A.D. Hassan, G. Elmowalid and C.C.L. Chase

South Dakota State UniversityDepartment of Veterinary Science

Brookings, SD 57007

ABSTRACT

Bovine viral diarrhea virus (BVDV) is a significant problem in the cattle industry. The disease affects both beef and dairy herds. Infection in unprotected animals leads to gastrointestinal (GI) lymphoid tissue depletion, GI mucosal lin-ing destruction, and animal death. Persistently infected (PI) animals result when a cow is infected during gestation; its calf never reaches its production potential and is a lifelong carrier of the virus. Commercial vaccines are available but yield only moderate results. Understanding the mechanisms of the acquired immune response is crucial for further advancements to be made. In the immune response, professional antigen presenting cells (APC’s) en-gulf, process, and present antigen on their cell surface to white blood cells, name-ly the T lymphocytes and B lymphocytes. In doing so, they initiate the cascade of events necessary for developing acquired immunity. Given that APC’s serve as a cornerstone to the immune response, much research focuses on their role in immunity. Isolating and preparing APC’s for experiments is rather laborious. Some immortalized, bovine macrophage (BOMAC and BoMac) cell lines had become available that had potential to be used in place of the monocyte derived macrophages (MDM’s) which have traditionally been used with success. Analy-sis of the BOMAC’s and BoMac’s was done by comparing them with MDM’s. Microfilament organization, phagocytic characteristics, and cell surface recep-tor presence was investigated. Our results indicated that these immortalized cell lines would be poor in vitro models for MDM’s. Characteristic differences between the MDM’s and both the BOMAC and BoMac cell lines leads one to believe the cell lines cannot be considered to be representative models of APC’s in vivo.


THYROID HORMONE DISTRIBUTIONASSOCIATED WITH THE

DEVELOPMENT OF HYPERTENSIONAND CONGESTIVE HEART FAILURE

IN DILATED CARDIOMYOPATHIC HAMSTERS

Laurie Keogh and Patricia M. TilleMount Marty CollegeYankton, SD 57078

A. Martin GerdesCardiovascular Research Institute

Sioux Falls, SD 57104

ABSTRACT

The effect of thyroid function related to cardiovascular disease in recent stud-ies has increased our understanding of the molecular and cellular basis for the cardiovascular changes occurring as a result of thyroid dysfunction. This study examined the effects of thyroid hormone treatment on Dilated Cardiomyopathic hamsters (DCM) from 4 to 6 months of age vs. age-matched controls. Thyroid hormone exerts significant effects on the heart and cardiovascular system as well as on most cells in the body. Through stimulation of enzymes concerned with glucose oxidation, thyroid hormone increases basal metabolic rate, oxygen consumption, and body heat production. It also plays an important role in maintaining blood pressure, and regulating tissue growth and development. In advanced congestive heart failure, T3 levels are low and rT3 levels are high, while T4 and TSH levels tend to remain normal. Blood samples were collected and measured for circulating triiodothyronine (T3), thyroxine (T4), thyrotropin (TSH), and reverse triiodothyronine (rT3) levels. The total T3 and T4 are solid phase radioimmunoassay (RIA), where T3 or T4 I125 labeled hormone competes for the thyroid hormone in the serum sample for antibody binding. The higher the I125 counts, the lower the thyroid hormone content is in the serum sample. The T3 is increased in almost all cases of hyperthyroidism and becomes elevated before T4. T3 is therefore a more sen-sitive indicator of hyperthyroidism than the Total T4. Results indicated that thyroid hormone treatment of DCM hamsters provid-ed significant benefits by normalizing the T3 levels. This normalization prevents a decline in left ventricle function and any further loss in myocytes. Supported by National Science Foundation (NSF) Experimental Program to Stimulate Competitive Research.


POLYARYLATE CO-POLYCARBONATESCONTAINING THE MONOMER

2,2-BIS[4-(4-HYDROXYPHENYL)-PHENYL]PROPANE

Marci Medalen, Guy Longbrake, Tsvetanka Filipova and David BoylesDepartment of Chemistry and Chemical Engineering

South Dakota School of Mines & TechnologyRapid City, SD 57701

ABSTRACT Polyarylate co-polycarbonates (PAr-PCs) are a known type of polyestercar-bonates that combine the high impact strength and optical clarity of polycarbon-ates with the high heat resistance of polyarylates. Several polymers containing the monomers 2,2-bis[4-(4-hydroxyphenyl)-phenyl]propane, bisphenol A, tri-phosgene, terephthaloyl chloride, and isophthaloyl chlorides were synthesized by an interfacial polymerization method as new materials for investigation of solubility and glass transition characteristics. The monomers were used in vari-ous ratios. The details of the syntheses of these polymers will be reported as will NMR and IR characterization of the molecular structures. Solubility, molecular weight, and glass transition characteristics will also be discussed.


COMPARATIVE PROTEOMIC ANALYSIS OFPLANKTONIC AND BIOFILM PSEUDOMONAS

AERUGINOSA CELLS: IDENTIFICATION OFDIFFERENTIALLY EXPRESSED PROTEINS

Sébastien Vilain, Pascal Cosette, Guy-Alain Junter and Thierry JouenneUMR 6522 CNRS, IFRMP 23Faculté des Sciences de Rouen

76821 Mont-Saint-Aignan cedexFrance

Volker S. BrozëlDepartment of Biology and Microbiology


ABSTRACT

To identify physiological alterations following bacterial adhesion, protein extracts from planktonic and sessile P. aeruginosa cells were subjected to two-di-mensional gel electrophoresis. Principal Component Analysis (PCA) was used to interpret spot quantity variations. As the ability of bacteria to attach to surfaces depend in part on the nature of the surface material and process conditions, we used two biofilm formation processes, i.e., by using clay beads and glass wool as supports. Moreover, as the proteomes of biofilm bacteria are strongly dependent on biofilm age, we used two biofilm incubation times, i.e., 18 h and 48 h. Thus, eight different incubation conditions were tested: bacteria attached to glass wool or clay beads after incubation for 18 h or 48 h, and the corresponding free cell counterparts. A total of 914 protein spots were detected on 2-DE electrophero-grams and quantified by computing scanning densitometry. PCA extracted three components from standardized values, accounting together for 78.4% of the variability in the data. Component 1 opposed free-cell cultures to biofilm ones. Component 2 was related essentially to free-cell cultures, discriminating between the two incubation times. Component 3 opposed the two modes of biofilm growth. The dependence of the protein patterns on the biofilm formation process was further illustrated by the identification of more than 80 polypeptides whose amount remained unchanged or was altered in adherent bacteria.The results reinforce the topical assertion that biofilm display a specific physi-ological behaviour but also question the existence of a unique biofilm pheno-type.


CONVERSION OF LEWIS AND CLARK LAKE AND LAKE FRANCIS CASE TO SUSTAINABLE SYSTEMS

Howard CokerDepartment of Chemistry


ABSTRACT

Sedimentation will eventually fill the mainstem reservoirs of the Missouri River. The two lower reservoirs are the most endangered because they receive 25% of the sediment but constitute only 8% of the total storage volume. Both have lost about 20% of their original capacity. Proposed is the removal of the sediments from these reservoirs at their annual rate of inflow and transporting them by pipeline below Gavins Point Dam for return to the Missouri River channel. From there the sediments can wend their way to the Gulf as they did historically. As the reservoirs fill, several benefits are lost. The relationship between the loss of these benefits and sedimentation may be described by a loss function whose general economic relationship to project cost has been analyzed. The sum of the annualized capital and annual operating costs of the pipeline has been minimized to determine expressions for the optimal slurry velocity and pipe diameter. Two special purpose pipeline loaders, one for sand and the other for silt and clay, are proposed as being much cheaper to operate than dredges. The overall analysis is still in process, but there appears to be sufficient basis justify a comprehensive economic, engineering, and environmental investigation of the proposal.


FECAL COLIFORM STRAIN IDENTIFICATIONTO FACILITATE WATER RESOURCEMANAGEMENT IN SOUTH DAKOTA

Erick Jorgenson, Nels Troelstrup Jr. and Bruce BleaklevSouth Dakota State University

Brookings, SD 57007

ABSTRACT

A well-established biological indicator of fecal pollution is the bacterium Escherichia coli, an intestinal inhabitant of many animal species and human that is not native to soils or waters. Traditional methods have not been able to discern whether an E. coli isolate is from an animal or human source. For the study de-scribed here, antibiotic resistance analysis (ARA) of E. coli strains from human and animal sources was done to see if the ARA patterns of E. coli strains from different hosts and ecoregions in the state differed significantly. This required construct-ing a reference database of bacterial isolate so the likely origin of E. coli found in the surface waters could be identified. A database of approximately 2800 E. coli isolates from seven different host sources obtained from four different ecoregions in South Dakota was constructed. E. coli isolates from fecal and water samples were sent to another laboratory for ARA, and the results and database were sent back to us for statistical analysis. Once comparisons between ARA patterns of E. coli isolates from different ecoregions and animals are made, the testing of the database will begin. This will be done by analyzing water samples with a known source and testing them against the database. Finally, blind water samples are taken from various lakes and streams to implement the bacterial source tracking method. The database will be turned over to South Dakota state agencies for use in monitoring South Dakota water supplies for fecal contamination.


THE HYDROCLIMATOLOGY OF THECENTRAL TERRITORY OF CONTERMINOUS

U.S. AND STREAM FLOW REGIMEIN UPPER MISSOURI RIVER BASIN

Boris A. Shmagin and Carol JohnstonSouth Dakota Water Resources Institute and

Center for Biocomplexity StudiesSouth Dakota State University

Brookings, SD 57007

ABSTRACT

Long-term stream flow records (1940-1970) from 103 U.S. Geological Survey gauging stations with drainage area in range 300-3500 sq mi within the mid-continental United States were analyzed using multivariate statistics. Factor analysis of average annual flow revealed five patterns of river runoff within five distinct regions: I- Middle Missouri-Mississippi basin, II- West Bank Middle Mississippi basin, III - Gulf of Mexico basin, IV - Upper Great Lakes area and V - West-East Mountains Uplands. This factor model reflected 53% of the variance of the initial matrix. Every group of watersheds obtained as a factor was present-ed by one gauging station with time series of annual discharges (I- 06810000, II- 07018500, III- 08095000, IV- 04079000 and V- 06119500). Within this time interval streams represented by patterns I, III, IV and V have increasing values (i.e. increasing discharge) and those represented by II have a decrease. The positive trend for pattern I and V is statistically significant. For the five typical flow records, monthly average values were obtained from three to four seasons composed of different ensembles of months. The trend for seasonal components was analyzed for IV and V typical watersheds and a significant increase was obtained for fall-winter season for type V. For each of the five stream patterns, regression equations were obtained from four indices of global atmospheric circulation (AO, NAO, AAO, PNAO). The equations contain from one to five variables (predictors) and have coefficients of correlation from 39% to 85%. Also teleconnections equations were obtained for three seasons components for IV and V typical watersheds with from three to six predictors and coefficients of correlation from 60% to 81%. The same methodology was applied to 46 time series from gauging station with drainage areas from 113 to 398 sq mi in the Upper Missouri basin with mutual period of observation 1963-1991. Factor analysis of average annual flow present revealed five patterns of river runoff within four distinct subregions of the territory (east, two carbonate karsts areas, uplands). This factor model reflected 62% variance of initial matrix. Each of four groups of watersheds obtained as a factor was presented by one gauging station with time series of annual and monthly discharges (I- 06218500, II- 06478690, III- 06412500,


and IV- 06323000). Streams represented by patterns I, II and IV have increase of values and those represented by III have a decrease. The positive trend for pat-tern II is statistically significant. For four typical flow records, monthly average values were obtained from three to four seasons composed of different ensembles of months. The trend for seasonal components were analyzed for four typical watersheds and a significant increase was obtained for fall-winter season for type IV. Stream runoff is the most appropriate regional indicator for hydroclimato-logical processes. With multidimensional statistics this process may considered as spatiotemporal function of different scale of landscape properties and dynamics. The hydrological regionalization of the central conterminous U.S. presented in this paper puts the hydrologic regime of the Upper Missouri River Basin in a more general picture. This material is based upon work supported by the National Science Foun-dation/EPSCoR Grant No. 0091948 and by the State of South Dakota.


AEROBIC ENZYME ACTIVITIES ANDSEASONAL ACCLIMATIZATION IN

RESIDENT PASSERINE BIRDS

Eric T. LiknesDepartment of Biology

Northern State UniversityAberdeen, SD 57401

Department of BiologyUniversity of South Dakota


David L. Swanson Department of Biology


ABSTRACT

Improved winter cold-tolerance is widespread among small birds from tem-perate climates and is usually associated with improved shivering endurance and elevated peak metabolic rate (Msum), although precise mechanisms of acclimatiza-tion are incompletely understood. Elevated Msum and improved cold-tolerance may be associated with greater metabolic intensity of thermogenic tissues, due to mass-specific increases in oxidative enzyme capacity, or to increases in the masses of metabolically active tissues. To examine mechanisms underlying win-ter increases in Msum, we captured wild, free-living individuals of three resident passerine species, White-breasted Nuthatch (Sitta carolinensis), Black-capped Chickadee (Poecile atricapillus), and House Sparrow (Passer domesticus). Each of these three species shows large seasonal variations in Msum, in excess of 35%. We investigated seasonal changes in the pectoralis and supracoracoideus muscles of mass-specific and total activities of key enzymes in catabolic pathways; Phospho-fructokinase (PFK), Citrate synthase (CS), and β-hydroxyacyl Co-A Dehydroge-nase (HOAD). Preliminary analyses of these data suggest that neither total nor mass-specific activities of PFK, CS, and HOAD varied seasonally in any species. If correct, this suggests that the modulation of cellular aerobic capacity is not an important mechanism of seasonal acclimatization. Such a result is generally consistent with other studies of seasonal acclimatization in small birds.


SPECIAL SYMPOSIA

FOLLOWING IN THE FOOTSTEPS OF LEWISAND CLARK: THE GEOLOGY AND

ALEONTOLOGY OF THE MISSOURI RIVER

James E. Martin and David C. Parris, Moderators

Introduction. Jim Martin

The Geology And Paleontology Recorded By The Corps Of Discovery, In-cluding The First Fossil Reptile From The American West. David C. Parris, New Jersey State Museum and Sally Shelton, Smithsonian Institution

Revised Stratigraphy Of The Lower Pierre Shale Of South Dakota, And The Description Of New Lithostratigraphic Members. James E. Martin, SD School of Mines and Technology and Janet L. Bertog, Northern Kentucky University

Bentonite Correlation Of The Pierre Shale Of South Dakota. Janet Bertog, Northern Kentucky University

Stratigraphy And Paleoecology Of The Middle Pierre Shale Along The Mis-souri River (Central South Dakota). Paul A. Hanczaryk, New Jersey Depart-ment of Transportation, and William B. Gallagher, New Jersey State Museum

Rare Earth Element Analysis (Ree) Of Fossil Vertebrates In The Pierre Shale: Paleoenvironmental Conditions. Doreena Patrick, University of Pennsylvania and David Grandstaff, Temple University

Geological Structures And Timing Constraints Along The Missouri River, Central South Dakota. J. Foster Sawyer, SD Geological Survey and James E. Martin, SD School of Mines and Technology

The Origin Of The James Ridge, Yankton County, South Dakota. Gary D. Johnson, University of South Dakota and Kelli A. McCormick, South Dakota Geo-logical Survey

Nannofossils And Environment Of The Upper Unit Of The Crow Creek Member Of The Pierre Shale (Upper Cretaceous), Crow Creek Sioux Indian Reservation, Central South Dakota. Andrea Concheyro, University of Buenos Aires, Argentina and James E. Martin, SD School of Mines and Technology

Fossil Fish From The Cretaceous Of The Western Interior: Clarifying The Taxonomic Record. David C. Parris, Barbara S. Grandstaff, and William B. Gal-lagher, New Jersey State Museum

The First Named Marine Reptile From The American West, A Mosasaur, Mosasaurus Missouriensis Harlan 1834; Its History, Source, And Osteology. James E. Martin, SD School of Mines and Technology

Rare Earth Element (Ree) Analyses Of Fossil Vertebrates In The Pierre Shale: Fossil Provenience. Doreena Patrick, University of Pennsylvania, James E. Mar-tin, SD School of Mines and Technology, David C. Parris, New Jersey State Museum, and David E. Grandstaff, Temple University

The Largest Mosasaur (Squamata; Reptilia) From The Missouri River Area (Late Cretaceous; Pierre Shale) Of South Dakota. Robert Meredith, University of California, Riverside, James E. Martin, SD School of Mines and Technology and Paul Wegleitner, SD Department of Environment and Natural Resources

Unusual Preservation Of The Pelvic Region Of Platecarpus (Mosasauridae; Reptilia) From The Sharon Springs Member Of The Pierre Shale. Bevin R. O’Grady and John A. Pappas, Rutgers University and Randolph J. Moses, SD School of Mines and Technology

Molluscs In The Stomach Contents Of Globidens, A Clam-eating Mosasaur, From The Late Cretaceous Pierre Shale, Big Bend Area Of The Missouri River, Central South Dakota. James E. Martin and James E. Fox, SD School of Mines and Technology

Size Variation In Cranial Morphology Of Late Cretaceous Toxochelys (Tes-tudines; Cheloniidae) Of South Dakota. Margaret Hart, SD School of Mines and Technology

Comparison Of Gastroliths Within Plesiosaurs From The Late Cretaceous Marine Deposits Of Vega Island, Antarctic Peninsula, And The Missouri River Area, South Dakota. Wayne Thompson and James E. Martin, SD School of Mines and Technology and Marcelo Reguero, Museo de La Plata, Argentina

Distribution And Taphonomy Of The Flying Reptile, Pteranodon, From The Campanian Lower Pierre Shale Of South Dakota. Jennifer Roberts, SD School of Mines and Technology

A New Species Of The Diving Bird, Baptornis (Aves) From The Sharon Springs Member Of South Dakota. Amanda Cordes, University of Oklahoma and James E. Martin, SD School of Mines and Technology

Paleoecological Implications Of The Fox Hills Formation (Maastrichtian) Reptilian Fauna From South-central North Dakota. John W. Hoganson, ND Geological Survey and J. Mark Erickson, St. Lawrence University

Meek And Hayden’s Nonmarine Paleontology Of The Upper Missouri River Section. Joseph H. Hartman, University of North Dakota


Fossil Mammals Of The Sentinel Butte Formation (Late Paleocene) Of North Dakota. Allen J. Kihm, Minot State University, David W. Krause, Stony Brook University, and Joseph H. Hartman, University of North Dakota



FOSSIL FISH FROM THE CRETACEOUSOF THE WESTERN INTERIOR:

CLARIFYING THE TAXONOMIC RECORD

David C. Parris and William B. GallagherBureau of Natural HistoryNew Jersey State Museum

Trenton, NJ 08625

Barbara Smith GrandstaffSchool of Veterinary Medicine

University of PennsylvaniaPhiladelphia, PA 19104

ABSTRACT

Current investigations in the formations on the Missouri River have ad-vanced understanding of Cretaceous fish faunas. Taxa originally described with type specimens from the Eastern Seaboard and Gulf Coasts can now be more readily compared with those described from the Western Interior. Increased knowledge of their anatomy and stratigraphic ranges allows clarification of their classification and nomenclature. The genus Cylindracanthus, as recently reviewed, essentially was advanced from a form taxon to a biological one, based on the dentition, seen for the first time in the first known Western Interior specimen. Possible acipenseriform affinities are hypothesized. The teleost species Xiphactinus audax, originally described from the Niobrara Group of the Western Interior, is now better known due to additional specimens collected from the Sharon Springs Member of the Pierre Formation (Early Cam-panian). Specimens from Eastern Seaboard formations of later Campanian age have been previously referred to the species X. vetus. Some of these records are corrected herein. The enigmatic species Protosphyraena gladius is now known from specimens ranging into the late Campanian and from the Atlantic and Gulf Coasts as well as the Western Interior. While much of the skeletal anatomy is as yet unknown, some taxonomic and nomenclatural details may be clarified by analysis of re-cently collected specimens. The new information on the widespread and long-ranging genus Enchodus, of which a thorough review is now in progress, is most valuable for biostrati-graphic purposes. By reinvestigation of type specimens as well as the discovery of new material, the taxonomy, stratigraphic ranges, and evolutionary trends of Enchodus species are now better established. Several lineages should prove useful as stratigraphic markers.


GEOLOGY AND PALEONTOLOGY RECORDEDBY THE CORPS OF DISCOVERY,INCLUDING THE FIRST FOSSIL

REPTILE FROM THE AMERICAN WEST

David C. ParrisBureau of Natural HistoryNew Jersey State Museum

Trenton, NJ 08625

Sally Y. SheltonNational Museum of Natural History

Smithsonian InstitutionWashington, DC 20560

ABSTRACT

Geological observations of the Corps of Discovery have received less sub-sequent attention than other aspects of natural history. While this may seem surprising in view of President Jefferson’s interest in agriculture and mineral resources, conventional interpretation has generally held that the expedition was so overwhelmed by the new discoveries and observations of plants and animals that geologic information was given less attention. Perhaps a more accurate view is that geologic observations are more subtle within the accounts. While subse-quent analyses have been more focused on botany and zoology, it is also true that geological observations of great interest were recorded early in the expedition. Fortuitously, the geology of the route along the Missouri River afforded an aston-ishingly good introduction to a general section of Upper Cretaceous and Early Tertiary strata in sequence lower to upper, a fact now recognized, but obscured from the expedition itself. Fossil discoveries are so common there now, that it is difficult to understand why the expedition did not record more of them. The Soldier River discovery of a fish fossil, Saurocephalus lanciformis, pre-sumably moved from original context, has long defied complete interpretation, but the specimen still exists and is subject to further analysis. The more perplex-ing journal entries regarding a so-called petrified fish (September 10, 1804) almost certainly refer to a Cretaceous reptile, likely Tylosaurus sp., noting recent discoveries in the area. Unfortunately, no specimen has been found that would document the discovery. As a footnote to history, the expedition’s naming of Corvus Creek, based on the observation of magpies there, ultimately bestowed the name Crow Creek on the Siouxan peoples whose reservation is one of the most scientifically productive two hundred years later.


BENTONITE CORRELATION OF THEPIERRE SHALE OF SOUTH DAKOTA

Janet L. BertogDepartment of Physics and Geology

Northern Kentucky UniversityHighland Heights, KY 41099

ABSTRACT

Bentonites are common in the lower Pierre Shale of the Cretaceous Western Interior. These bentonites provide a distal record of volcanic activity during this time frame while also providing event-stratigraphic horizons useful for regional correlations. Stratigraphic correlation of the bentonites provides a high-resolu-tion framework for regional tectono-stratigraphic interpretations in an active retro-arc foreland basin. In combination with the volcanic interpretation, this framework can provide a more comprehensive interpretation of basin activities. Based on whole rock geochemistry, phenocryst composition and biotite geochemistry, individual bentonite horizons can be recognized and are useful for regional correlation. Whole rock geochemical analysis indicates that volcanism was active in a forearc, island arc and backarc setting during the deposition of the Lower Pierre Shale. Bentonites can be further distinguished based on the mineral compositions and biotite geochemistry, when available. Stratigraphic correlation of the bentonites provides a high-resolution framework for regional tectono-stratigraphic interpretations in an active retro-arc foreland basin.


STRATIGRAPHY AND PALEOECOLOGY OF THE MIDDLE PIERRE SHALE ALONG THE MISSOURI

RIVER (CENTRAL SOUTH DAKOTA)

Paul A. HanczarykNew Jersey Department of Transportation

Bureau of MaterialsTrenton, NJ 08625

William B. GallagherBureau of Natural HistoryNew Jersey State Museum

Trenton, NJ 08625

ABSTRACT

A study of the stratigraphy of the middle Pierre Shale was undertaken, con-centrating on two sections located along the eastern side of the Missouri River, in the vicinity of Fort Thompson. The members of the Pierre Shale represented within these sections are, from oldest to youngest, the Gregory, Crow Creek, DeGrey and Verendrye. Micropaleontologic samples were analyzed with factor analysis performed to classify the assemblages. Four foraminiferal biofacies are recognized: two predominantly agglutinated, Boreal in origin and indicative of cooler, less oxygenated marine conditions, one shallower and the other deeper; one composed of calcareous benthic foraminifera, Tethyan in origin, indicative of warmer, more oxygenated and shallower waters; and a fourth, that is inter-preted as a mixed assemblage, transitional between the others. 87Sr/86Sr age analysis yielded twelve useable results. Samples from near the Baculites compres-sus biozone gave an age of approximately 72.43 +/- 2 Ma, a number relatively close to Obradovich’s 1993 time scale 40Ar/39Ar value for this horizon of 73.35 +/- 0.39 Ma. An average 87Sr/86Sr date of 72.41 Ma +/- 2 Ma was determined from fossil shell at a significant paleontologic find – the globular-toothed mosa-saur, Globidens sp. More equivocal dates of 74.58, 74.63 and 75.09 +/- 2 Ma were determined for the basal Crow Creek Member, a calcareous sandy siltstone within the upper Gregory Member and the B. gregoryensis biozone located at the base of the section, respectively. Based on the data from this study, the Crow Creek Member is interpreted as the result of the Bearpaw transgression (TST) and the DeGrey/Verendrye Member’s boundary is a maximum flooding surface. The Verendrye Member is interpreted as an HST; the basal Gregory Member is interpreted as an HST but part of the regressive pulse of the Claggett deposi-tional cycle. A preliminary correlation with the eustatic record from the late-K coastal plain of New Jersey places the basal Crow Creek unconformity coincident with an unconformable bounded surface at the base of the Marshalltown Forma-tion linked to Haq’s UZA-4.4 onlap cycle. A distinct, calcareous sandy siltstone


unit in the upper Gregory Member is suggested to represent a storm-dominated deposit or, perhaps, a tsunamtite resulting from the approximately coeval Man-son Impact.


RARE EARTH ELEMENT ANALYSIS (REE) OFFOSSIL VERTEBRATES IN THE PIERRE SHALE:

PALEOENVIRONMENTAL CONDITIONS

Doreena PatrickEarth and Environmental Science


David E. GrandstaffDepartment of Geology

Temple UniversityPhiladelphia, PA 19122

ABSTRACT

Concentrations of rare earth elements (REE) were measured in vertebrate fossils from five members (Sharon Springs, Gregory, Crow Creek, DeGrey, and Verendrye) of the upper Cretaceous Pierre Shale, South Dakota. REE signa-tures of fossil vertebrates are dependent only on the composition of the original diagenetic waters in which fossilization occurred. REE in modern waters show systematic variations as a function of redox, pH, sorption, and alkalinity, which affect the concentration and availability of REE in the water. REE signatures and their stratigraphic variations in fossil materials can be used to interpret variations in paleoenvironmental conditions by comparisons of REE signatures in fossils with modern waters. Such variations may be best visual-ized by plotting normalized concentrations of representative light (Nd), middle (Gd) and heavy (Yb) REE in ternary diagrams. This allows the basic shape of the REE signature to be represented. REE signatures in Pierre Shale fossils can be related to differences in mixing of oxic, shallow seawater and anoxic deep waters. If mixing differences are interpreted as depth variations, then the lower Sharon Springs was deposited in anoxic deep waters, the middle and upper Sharon Springs and Gregory in progressively shallower, and the overlying Crow Creek, DeGrey and Verendrye Members in progressively deeper marine waters. These interpretations are generally consistent with those based on faunal diversity, lithologic interpretation of the members, and previous sea level curves.


UNUSUAL PRESERVATION OF THE PELVICREGION OF PLATECARPUS, (MOSASAURIDAE;

REPTILIA) FROM THE SHARON SPRINGSMEMBER OF THE PIERRE SHALE

Bevin Rose O’Grady and John A. PappasRutgers University

New Brunswick, NJ 08901

Randolph J. MosesDepartment of Geology and Geological Engineering

SD School of Mines and TechnologyRapid City, SD 57701

ABSTRACT

A specimen of Platecarpus sp. was collected from the Sharon Springs Mem-ber of the Pierre Shale in Lyman, County, South Dakota, during the summers of 2001-2002. Upon removal, the specimen was discovered to have an articu-lated vertebral column beginning with the posterior dorsal vertebrae and ending with terminal caudal vertebrae, encompassing hind limb and pelvic elements. Detailed investigation has focused on the well-preserved rear paddles and pelvic structures. The right hind limb is well articulated, while the left is slightly disar-ticulated but equally well preserved. Similar to the hind limbs the right portion of the pelvic region is well articulated while the left portion is slightly disarticu-lated. This specimen provides valuable information concerning the structure and function of the hind limbs and pelvic region of Platecarpus, providing additional insight into the musculature of the posterior region of Platecarpus.


SIZE VARIATION IN CRANIAL MORPHOLOGY OF LATE CRETACEOUS TOXOCHELYS (TESTUDINES;

CHELONIIDAE) OF SOUTH DAKOTA

Margaret HartGeology/Paleontology


ABSTRACT

In 1988, five species belonging to the genus Toxochelys were reexamined. As a result, only two species were considered valid, T. latiremis Cope and T. moorevil-lensis Zangerl. T. browni (Hay), T. weeksi Collins, and T. barberi (Schmidt) were designated as junior synonyms of Toxochelys latiremis (Nicholls 1988). Prior to the reinvestigation, T. browni was a species defined by its broader size limits and distinct premaxillary region. Nicholls dismissed these characters as natural varia-tions within a species, attributing them to temporal factors. T. browni was also temporally restricted, only being found in the lower Pierre Shale, a unit that had not been known to yield T. latiremis. Eleven undescribed toxochelid skulls from the Pierre Shale of South Dakota have been studied and all exhibit sizes that are less than or greater than the size range specified for T. latiremis. Furthermore, all of the toxochelids retaining an anterior cranial region display the characteristic sigmoidally-curved lateral margins of the premaxilla and maxilla. This character is not observed in T. latiremis. It is a possibility that T. browni may have been incorrectly assigned junior synonym status. In order to examine this uncer-tainty, a preliminary study has been conducted to determine if the systematics of Toxochelys needs to be readdressed. 46 characters of the skull and lower jaw were measured on the group of eleven Pierre Shale toxochelids and on a separate group of turtles belonging to T. latiremis. Measurements were first compared within each group to establish generalities. This and the original data were then used to compare proportionality of the various measurements and to compare the average growth rate between the two groups. Differing proportions between the two groups in both morphology and growth rate indicate the possibility of an error in the synonymization. Further research is recommended to clarify this uncertainty.


THE LARGEST MOSASAUR (SQUAMATA; REPTILIA) FROM THE MISSOURI RIVER AREA (LATE

CRETACEOUS; PIERRE SHALE) OF SOUTH DAKOTA

Robert W. MeredithBiology Department

University of CaliforniaRiverside, CA 92521

James E. MartinMuseum of Geology


Paul N. WegleitnerDepartment of Environment and Natural Resources

Pierre, SD 57501

ABSTRACT

The Cretaceous Pierre Shale along the Missouri River is extremely fossilifer-ous and has produced numerous mosasaur specimens since the western fossil dis-coveries of Lewis and Clark. Many of these marine reptile specimens represent the largest of mosasaurs, the tylosaurines, and in 1990, the Jim Wilkens family, discovered the largest heretofore recorded along the Missouri River near Nicholas Creek, Lyman County, central South Dakota. Unfortunately, high water pre-vented collection of the specimen; finally, in 2000, water levels dropped, and the specimen was collected and found to consist of vertebrae, ribs, paddle elements, and a partial skull. Skull elements consist of anterior cranial and jaw elements, including a nearly complete right lower jaw measuring 1.53 meters, making it the largest mosasaur ever collected from the Missouri River. The partial skeleton is referable to the subfamily Tylosaurinae based on large size, tooth structure, and long predental rostrum. Further identification must await resolution of the taxonomy of the Tylosaurinae, a project currently underway. The specimen was recovered from a lag deposit representing an unconformity at the middle portion of the Campanian Sharon Springs Member of the Pierre Shale along the Mis-souri River.


MEEK AND HAYDEN’S NONMARINEPALEONTOLOGY OF THE UPPER

MISSOURI RIVER SECTION

Joseph H. HartmanDepartment of Geology and Geological Engineering

University of North DakotaGrand Forks, ND 58202

ABSTRACT

The details of F.B. Meek and F.V. Hayden’s earliest studies of molluscan fossils and marine and nonmarine strata along the Missouri River are known to specialists and an eclectic few. In the latter part of the 19th and first half of the 20th centuries, their new species and stratigraphic nomenclature were used by geologists to map large tracts of the Western Interior plains and intermontane basins and interpret the age relations of the “Judith River” and “Fort Union” time without substantial revision until the advent of mammalian biochronol-ogy. As the years have passed, their work continues to serve as the foundation for ongoing Paleogene molluscan evolutionary and biostratigraphic studies. We now know that both stasis and rapid faunal turnover are present in the freshwater molluscan record. The terrestrial record, which was not as well known to Meek and Hayden, is less well sampled, but appears to have left a more biostratigraphi-cally and paleobiogeographically complicated record. The early Paleocene begins with opportunistic unionid bivalve (Unionidae) species largely unadorned by the sculpture of the Cretaceous. Only a few freshwater gastropod species can be dem-onstrated to cross the K/T boundary into the Fort Union Formation (Group), including two taxa named by Meek and Hayden. Molluscan diversity appears to remain relatively low until the final regression of the Paleocene Cannonball Sea. The “classic Fort Union” molluscan fauna described by Meek and Hayden is largely confined to strata overlying the marine Cannonball Member of the Fort Union Formation, including the Tongue River (Bullion Creek) and lower part of the Sentinel Butte Members. This stratigraphic interval represents the late Paleo-cene North American Land Mammal Ages Tiffanian-3 (Ti) and Ti4, about 59 to 56 Ma. Thus the “Fort Union molluscan fauna time” used to correlate strata in dozens of U.S. Geological Survey coal-mapping programs was in fact correlating a much more restricted biostratigraphic interval of the Paleocene, based origi-nally on fossils collected by Hayden from outcrops along the Missouri River.


DISTRIBUTION AND TAPHONOMY OF THEFLYING REPTILE, PTERANODON, FROM THE

CAMPANIAN LOWER PIERRE SHALE OFWESTERN SOUTH DAKOTA AND EASTERN WYOMING

Jennifer RobertsDepartment of Geology and Geological Engineering


ABSTRACT

Sediments of the Campanian Lower Pierre Shale in South Dakota and Wyo-ming contain numerous fossils, including the pterosaur Pteranodon, nineteen of which are now housed at the Museum of Geology at the South Dakota School of Mines and Technology in Rapid City, South Dakota. The basal member of the Pierre Shale, the Gammon Ferruginous Member, is composed mostly of grey mudstone and shale and contains concretions. Four of the nineteen pterosaur specimens were found in this member. A 1 to 1.5 meter thick bentonite suc-cession, the Ardmore Bentonite, marks the lower contact of the Sharon Springs Member with the underlying Gammon Ferruginous Member. The Sharon Springs is a grey to black shale and has produced fifteen of the specimens. Over-lying the Sharon Springs Member is the Mitten Black Shale Member, which consists of blue-black fissile shale, containing calcareous and siderite concretions, as well as cone-in-cone structures. Sixty-six percent of the South Dakota School of Mines and Technology Mu-seum of Geology’s pterosaur collection consists of wing elements. The remaining percentages of the pterosaur fossils include associated skull fragments, articulated phalanges, scapulae, coracoids, carpals, femora, tibiae, and a part of a basicra-nium. Two specimens also contain fish vertebrae, which appear to be stomach contents. Two hypotheses explain the abundance of wing elements. The first hypothesis is the existence of a predatory preference. The wing membrane, com-pared to fleshier body parts, may not offer much to hungry mosasaurs or other predators. The second hypothesis is that the wing membrane would secure the wing elements in place while other elements were free to decompose and fall away, contributing to possible scavenging and breakage. In either case, the wing membrane may have served as a protective layer over the bones until burial.


PALEOECOLOGICAL IMPLICATIONS OF THEFOX HILLS FORMATION (MAASTRICHTIAN)

REPTILIAN AND AMPHIBIAN FAUNAFROM SOUTH-CENTRAL NORTH DAKOTA

John W. HogansonNorth Dakota Geological Survey

Bismarck, ND 58505

Mark J. EricksonGeology Department

St. Lawrence UniversityCanton, NY 13617

ABSTRACT

Fossil vertebrates of the Fox Hills Formation in the Missouri River valley of North Dakota play a singular role for interpretations of both paleoecology and paleogeography at the margin of the Western Interior Seaway (WIS). They link interpretations of terrestrial, estuarine, and marine ecosystems in a manner that invertebrates alone generally cannot. The Fox Hills Formation, which underlies and interfingers with the Hell Creek Formation, and associated marine tongues in the Hell Creek Formation indicate that the WIS persisted into the latest Cretaceous in North Dakota. Invertebrate fossils are common in the nearshore Fox Hills deposits consisting primarily of sandstone and siltstone. Deposition occurred during the Late Maastrichtian Jeletzkytes nebrascensis Western Interior ammonite Zone. Remains of reptiles and amphibians in the Fox Hills Formation in North Dakota are not common and consist mostly of teeth and other isolated skeletal parts. Only partial skeletons of any reptiles have been found in the Fox Hills Formation. The meager reptilian and amphibian fauna, recovered from nearshore sandstone facies, consists of marine and terrestrial taxa. The marine taxa present include the mosasaurs Mosasaurus dekayi and Plioplatecarpus sp. Terrestrial taxa represented are dinosaurs, Tyrannosauridae indet. and Theropoda indet.; turtles, the trionychid Aspideretoides sp. and the nanhsuingchelyid Basilemys sp.; the crocodile Leidyosuchus? sp.; the Choristodera Champsosaurus sp. and the salaman-ders Opisthotriton kayi and an undetermined species. This mixed fauna indicates that marine mosasaurs frequented shallow water areas of the WIS. The occurrence of Champsosaurus with the horseshoe crab, Casterolimulus kletti, exemplifies faunal mixing in an estuarine habitat. Dino-saurs, crocodiles, turtles, and salamanders inhabited Late Maastrichtian shoreline areas.


COMPARISON OF GASTROLITHS WITHINPLESIOSAURS FROM THE LATE CRETACEOUS

MARINE DEPOSITS OF VEGA ISLAND,ANTARCTIC PENINSULA, AND THE

MISSOURI RIVER AREA, SOUTH DAKOTA

Wayne A. Thompson and James E. MartinDepartment of Geology and Geological Engineering


Marcelo RegueroDepartamento Paleontologia Vertebrados

Museo de La PlataLa Plata, Argentina

ABSTRACT

Plesiosaur remains containing gastroliths have been recovered from the Late Cretaceous Cape Lamb Member of the Lopez de Bertodano Formation, Antarctic Peninsula, as well as the Sharon Springs Member of the Pierre Shale, South Dakota. The significance of the Antarctic specimen lies in the large number and relatively small size of the stones. All specimens compared are of relatively large elasmosaurids, and rib diameters indicate similar overall size of the three specimens. One testable hypothesis was whether or not the weight of the gastroliths might be similar among similarly sized individuals. However, comparisons between the gastroliths recovered from the Antarctic specimen and those recovered from the Pierre Shale show a large degree of difference in both the size and number of gastroliths. The total mass of the stones collected in the specimen from Antarctica was 3.02 kg., whereas those recovered from the South Dakota specimens totaled .92 kg and 9.3kg, respectively. Perhaps not all stones originally within each plesiosaur were found, but efforts were designed to recover every stone. The number of stones recovered from the Antarctic plesiosaur was exceptionally large (2,626), and may represent the most gastroliths ever recovered from a single plesiosaur specimen. However, those recovered from South Dakota totaled only 42 and 253, respectively. Therefore, neither weight nor number of stones corresponds among these large individuals. Lack of correspondence among individuals furthers questions concerning the utilization of gastroliths for neutral buoyancy, ballast, or as an aid in digestion. Many parameters remain unexplored, and questions arise not only as to the utilization of the gastroliths but also as to whether physiology of the Antarctic plesiosaurs differs from those at lower latitudes. The Antarctic research was funded through a grant from the Office of Polar Programs, National Science Foundation (OPP #0087972), and the US Corps of


FOSSIL MAMMALS OF THE SENTINELBUTTE FORMATION (LATE PALEOCENE)

OF NORTH DAKOTA

Allen J. KihmDepartment of Geosciences

Minot State UniversityMinot, ND 58703

David W. KrauseDepartment of Anatomical Sciences

Stony Brook UniversityStony Brook, NY 11794

Joseph H. HartmanDepartment of Geology and Geological Engineering

University of North DakotaGrand Forks, ND 58202

ABSTRACT

The fossil mammals of the Tongue River Formation in North Dakota have been described or discussed in a number of papers, whereas those of the overly-ing Sentinel Butte Formation have received little attention. Recent work has located several new localities and reinterpreted previously known localities as occurring in this formation. The Grassy Butte Locality is in a channel sandstone at the Tongue River-Sentinel Butte formation contact and has produced four taxa (one multituberculate, three eutherians). The River Basin 3 Locality (2-19 m above the base) and the Cross Locality (10-19 m above the base) have each produced only Titanoides primaevus. The Riverdale Locality is approximately 23 m above the base of the formation and contains five taxa (one multituberculate, four eutherians). L5500b is approximately 26 m above the base and contains 10 taxa (two multituberculates, one marsupial, seven eutherians). The Red Spring Locality contains nine taxa (four multituberculates, five eutherians) and is ap-proximately 36 m above the base of the formation. The highest occurrence of mammals is at the type locality of Titanoides primaevus (Witter Locality), the first Paleocene fossil mammal reported from North Dakota, 48 m above the base of the formation. All of the localities probably represent the Tiffanian-4 biochron, although not all localities contain time diagnostic taxa. Titanoides primaevus has the greatest stratigraphic distribution, occurring in the lowest and highest locali-ties, as well as in all other localities except L5500b and Red Spring. Ptilodus and Plesiadapis occur from the base of the formation through the Red Spring Locality. Other taxa are more restricted, occurring in only one or two locali-ties. The Sentinel Butte Formation is between 115 and 189 m thick in North Dakota. The overlying Bear Den Member of the Golden Valley Formation has


been determined to be Clarkforkian in age (based upon fossil plants). Therefore, the upper Sentinel Butte Formation may represent a significant portion of the late Tiffanian and/or the Clarkforkian.


GEOLOGICAL STRUCTURES AND TIMINGCONSTRAINTS ALONG THE MISSOURI

RIVER, CENTRAL SOUTH DAKOTA

J. Foster SawyerSouth Dakota Geological Survey




ABSTRACT

The newly compiled geological map of South Dakota presents one of the most comprehensive views of the geology and structural features of central South Dakota. Information included in this compilation resulted from structures that were newly mapped by the authors, as well as previously documented structures gleaned from the literature, resulting from over one hundred years of sporadic research in this area. Structures include folds and faults, both normal and reverse faults, and most displacements are within the range of one to ten meters, but some are greater. Compilations of drill data from eastern South Dakota have also illuminated numerous potential structural features beneath the glacial drift, several of which extend across the Missouri River into western South Dakota. Central South Dakota is on the margin of the Williston Basin, and many of these structures are also potential traps for hydrocarbon accumulations, particularly natural gas. Although geological structures are common along the Missouri River, the timing of their formation has been difficult to ascertain. Nearly all faults, ex-cluding landslides, occur within the Late Cretaceous Niobrara and Pierre Shale units, and most faults have only late Quaternary deposits lying undisturbed above. Previous authors have suggested Laramide, post-Miocene, and even Quaternary times of formation, but noted that faults could not be unequivocally dated as younger than Late Cretaceous because direct evidence was wanting. Now, a normal fault discovered in Brule County, central South Dakota, appears to offset glacial deposits and glacially derived boulders are cemented within the fault gouge. This occurrence constrains the date of faulting to after the initia-tion of glaciation. Therefore, at least some faulting within central South Dakota occurred relatively recently in the Quaternary, and this fault represents the first direct evidence of the youngest known structural event.


REVISED STRATIGRAPHY OF THE LOWERPIERRE SHALE (UPPER CRETACEOUS)

OF CENTRAL SOUTH DAKOTA



Janet L. BertogDepartment of Physics and Geology

Northern Kentucky UniversityHighland Heights, KY 41099


Trenton, NJ 08625

ABSTRACT

The Pierre Shale is extensively exposed throughout the Northern Great Plains and is well exposed along the Missouri River Trench in central South Dakota. Currently, the Pierre Shale is of formational rank, but should be el-evated to group status and all members should be elevated to formational rank as they are of distinctive lithology and are mappable throughout the Missouri River area. Extensive geological and paleontological investigations of the lower Missouri River Trench indicate a number of previously described units should be subdivided. In particular, the lowest described unit of the Pierre Shale, the Sharon Springs, exhibits three separate disconformity-bounded lithostratigraphic units and should be newly designated. The lowermost is unique, whereas the upper two units may be observed in the type area of the Sharon Springs in western Kansas. The lowermost unit is characterized by numerous bentonite beds, is normally disconformably superjacent to the Niobrara Formation, and may be absent where degraded. This unique unit requires separate designation. The two units within the Sharon Springs consist of a lower siliceous shale that weathers vertically, whereas the upper unit is more bentonitic and characterized by phosphatic concretions. These two units should be considered as members of a hierarchically elevated Sharon Springs Formation.


NANNOFOSSILS AND ENVIRONMENT OFTHE UPPER UNIT OF THE CROW CREEK

MEMBER, PIERRE SHALE (UPPER CRETACEOUS), CROW CREEK SIOUX INDIAN RESERVATION,

CENTRAL SOUTH DAKOTA

Andrea ConcheyroDepartment of Geological Sciences

University of Buenos AiresBuenos Aires, Argentina



ABSTRACT

The Crow Creek Member of central South Dakota has received notoriety as having perhaps been caused by a tsunami associated with a meteor impact in the adjacent state of Iowa. This event, known as the Manson Impact, has been cited as the source of this light-colored calcareous unit interbedded within oth-erwise thick members of black shale. Interestingly, however, two light-colored units, informally termed the lower and upper Crow Creek units, respectively, occur in the Big Bend area of Crow Creek Indian Reservation near the type area. The lowermost unit is that normally considered the Crow Creek Member by most authors elsewhere where only a single unit occurs and is characterized by a coarse, ferruginous basal unit overlain by calcareous yellow-brown marl. The upper non-calcareous marl is separated from the lower Crow Creek by a thick interval of bentonitic shale identical to that of the DeGrey Member. Samples were taken at 20 cm superposed intervals through both Crow Creek units, and nannofossils were found in both units but are particularly abundant and defini-tive in the upper Crow Creek unit. Stratified nannofossils that elsewhere have been found superposed over great intervals of time suggest that the upper unit was not deposited during a single depositional event. The nannofossils indicate a Campanian age of the upper Crow Creek, and the occurrence of Braarudosphaera bigelowi from the upper part of the unit would normally suggest shallow water, probably associated with a marginal marine setting or restricted marine environ-ment. Overall, two yellow-brown marl units exist in the Big Bend area of central South Dakota, rather than the single unit found elsewhere. The upper unit is not calcareous, was deposited in shallow marine waters, and superposed assemblages argue against a tsunami origin.


MOLLUSCS IN THE STOMACH CONTENTSOF GLOBIDENS, A SHELL-CRUSHING

MOSASAUR, FROM THE LATE CRETACEOUS PIERRE SHALE, BIG BEND AREA OF THE

MISSOURI RIVER, CENTRAL SOUTH DAKOTA

James E. Martin and James E. FoxDepartment of Geology and Geological Engineering


ABSTRACT

Globidens is one of the most rare of marine reptiles and is characterized by a massive, bulbous dentition rather than sectorial as are most other mosasaurs. Rarity of the taxon coupled with the unusual dentition has prompted a number of theories about life style. Most theories suggest the dentition was utilized for crushing resistant elements such as turtles or pelecypods or perhaps for scaveng-ing. Finally, a partial skeleton of Globidens, secured from the upper DeGrey Member of the Pierre Shale along the shores of the Missouri River in the Big Bend area of central South Dakota, provides direct evidence. During excavation, numerous pelecypod fragments were found associated with the skeleton. Such concentrations were not found laterally nor above or below the skeleton. Once in the laboratory, the shell fragments were found associated within the ribcage area. Included within the stomach area are small (4 cm) bivalves with lamellar shells, probably of the genus Anomia. The most common pelecypods exhibit a prismatic shell microstructure typical of inoceramids. Of these, two very differ-ent morphotypes are included: a coarse-ribbed type and a large, flat, thin-shelled taxon. Because of their position in the mosasaur, their fragmented condition, limited taxonomic diversity, and absence from surrounding sediments, the pe-lecypods are considered stomach contents. Some smaller, complete shells of Anomia escaped breakage whereas larger inoceramids were always crushed. This specimen of Globidens appears to have had a preference for the large, flat, thin-shelled inoceramids that probably had a large, fleshy visceral mass.


RARE EARTH ELEMENT (REE) ANALYSESOF FOSSIL VERTEBRATES IN THE PIERRE

SHALE: FOSSIL PROVENIENCE

Doreena PatrickEarth and Environmental Science





Trenton, NJ 08625

David E. GrandstaffDepartment of Geology

Temple UniversityPhiladelphia, PA 19122

ABSTRACT

Rare earth element concentrations were measured in mosasaur bones col-lected from five members (Sharon Springs, Gregory, Crow Creek, DeGrey, and Verendrye) of the upper Cretaceous Pierre Shale at localities near the Missouri River in Brule, Buffalo, Hughes and Hyde counties. Fossils from each member of the Pierre Shale have REE signatures similar to one another, but statistically different from those of other members. Fossils collected from the Sharon Springs Member have distinctive REE signatures that may be further subdivided statisti-cally into three superposed groups that correspond with the upper, middle, and lower Sharon Springs Member. Because REE signatures differ between members, fossil bones removed from stratigraphic context can be assigned to a member based on REE signature comparisons. A sample of a mosasaur originally collected in the 1800’s that was known to have come from the Pierre Shale in central South Dakota but which could not be as-signed to a particular member on the basis of collection information was tested and compared with the already analyzed in-situ fossils. Discriminant analysis grouped this fossil statistically with those of the DeGrey Member; therefore the provenience of this fossil could be assigned using REE analyses.


THE FIRST NAMED MARINE REPTILE FROMTHE AMERICAN WEST, A MOSASAUR:

MOSASAURUS MISSOURIENSIS (HARLAN) 1834;ITS HISTORY, SOURCE, AND OSTEOLOGY



ABSTRACT

Although the first marine reptile from the American West may have been the “45-foot long fish” that Lewis and Clark noted, the first described and surviving marine reptile is that of a mosasaur taken by Prince Joseph Maximilian, later to be Maximilian II, King of Bavaria, from what is now South Dakota to Germany in the early 1830’s. The specimen was later named as Mosasaurus maximiliani in 1845. In the meantime, the snout of a marine reptile was named as Ichthyosaurus missouriensis in 1834. Evidently, the snout of the specimen taken by Maximilian had been sent earlier to the East Coast where it was thought to be an ichthyo-saur. However, the snout matches perfectly with that of the skull in Germany and are therefore the same individual. Therefore, priority dictates that the speci-men be considered as Mosasaurus missouriensis. The geographic source of the specimen was documented as the Big Bend of the Missouri River. However, the stratigraphic source of the specimen has long been in doubt and was considered to have come from a concretionary level within the Verendrye Member of the Pierre Shale. Invertebrate fossils associated with the skeleton commonly occur in the DeGrey Member, and geochemical analyses of concretionary fragments in which the original skull is encased indicate a DeGrey Member source. More-over, extensive investigation of the Big Bend area has resulted in many additional specimens of the species, and most have been found in the DeGrey Member. Some specimens are exquisitely preserved and reveal additional characteristics of the osteology of the species. Overall, the first mosasaur from the American West came from the Big Bend of the Missouri River and was derived from the upper DeGrey Member of the Late Cretaceous Pierre Shale.


A NEW SPECIES OF THE DIVING BIRD,BAPTORNIS, FROM THE LOWER PIERRE

SHALE (UPPER CRETACEOUS) OFSOUTHWESTERN SOUTH DAKOTA

Amanda Cordes-PersonDepartment of Ornithology

Sam Noble Museum of Natural HistoryUniversity of Oklahoma

Norman, OK 73068



ABSTRACT

Fossil birds are relatively rare in Cretaceous deposits of the Northern Great Plains. From marine deposits of the Niobrara Formation in Kansas, a small diversity of birds is known, but until now, the large diving bird, Hesperornis, was the major bird taxon known from the Pierre Shale of South Dakota. Now, a par-tial skeleton of another, smaller diving bird, Baptornis, has been secured from the Sharon Springs Member of the Pierre Shale in Fall River County, South Dakota. The specimen is represented by pelvic fragments and lower leg elements that are similar to but much more robust than Baptornis advenus from the subjacent Niobrara Formation. The taxon is nearly twice the size of the Niobrara species, principally in robustness rather than length of elements. Overall, the specimen represents the first occurrence of Baptornis from the Pierre Shale, represents a new species, and indicates a greater diversity of birds from the Pierre Shale than previously thought.


THE ORIGIN OF JAMES RIDGE,YANKTON COUNTY, SOUTH DAKOTA

Gary D. JohnsonDepartment of Earth SciencesUniversity of South Dakota


Kelli McCormickSouth Dakota Geological Survey

Science CenterUniversity of South Dakota


ABSTRACT

James Ridge is adjacent to and sub-parallel to the James River near its mouth at the Missouri River in southeastern South Dakota. The Missouri River paral-lels the margin of the last Pleistocene glacial advance. Unlike Turkey Ridge and Yankton Ridge, also in Yankton County, James Ridge does not have a bedrock core. Instead, it is cored by a thick (60-120m) deposit of sand and gravel resting on a glacially-eroded surface of Carlile Shale (Upper Cretaceous), and covered in places by late Wisconsin oxidized till, but mostly by collapsed debris. The probable age of the core sediments is pre-Wisconsin. There is substantial sub-surface evidence for three pre-Wisconsin till sheets in Yankton County. The core of James Ridge was probably emplaced during the third pre-Wisconsin glacia-tion by thrusting of frozen blocks of sand and gravel from at least the last two pre-Wisconsin outwash units. Also present in the core are at least two blocks of Niobrara chalk, one of which is >40m thick, suspended in the sand and gravel. The present-day elongate shape of James Ridge parallels the late Wisconsin ice flow direction, likely reflecting considerable modification in geometry by gla-cial erosion. The surface topography of James Ridge shows evidence of meltwater drainage and collapsing of debris from the ice during glacial retreat. However, before the ice retreated, its axis was breached by Beaver Creek, then a meltwater channel, but now a sometimes intermittent stream.


PLENARY SESSION

RECOVERING LEWIS AND CLARK’SRIVER: TODAY’S PERSPECTIVES ONTHE HISTORY, MANAGEMENT ANDECOLOGY OF THE MISSOURI RIVER

Steve Chipps and Robert Tatina, Moderators

Recovering Lewis And Clark’s River. Carter Johnson, South Dakota State Uni-versity

A Missouri River Perspective: Fisheries And Recreation. Wayne Nelson-Stastny, South Dakota Department of Game, Fish & Parks

Operation Of The Missouri River Dams For Multiple Project Purposes. Jody Farhat, U.S. Army Corp of Engineers

Botanical Discoveries Of The Lewis And Clark Expedition In South Dakota. Dave Ode, South Dakota Department of Game, Fish & Parks

A Case Study Of Changing Land Use Practices In The Northern Great Plains: An Uncertain Future For Watershed Conservation. Ken Higgins, USGS Coop Unit-South Dakota State University, David Naugle, University of Montana, Kurt Forman, U.S. Fish & Wildlife Service

Culture Of Pallid Sturgeon At Gavins Point National Fish Hatchery. Herb Bollig, U.S. Fish & Wildlife Service

Fluvial Processing And Recreational Opportunities Of The Lower Missouri River. Perry Rahn, South Dakota Schools of Mines and Technology

Invasive Species In The Missouri River. Wayne Stancill, U.S. Fish & Wildlife Service

Gathering Of Natural Resource Information: Comparison Between The Corps Of Discovery (1804-1806) And The National Ecological Oberserva-tory Network. Charles Berry, USGS South Dakota Coop Unit, South Dakota State University


FLUVIAL PROCESSES ANDRECREATIONAL OPPORTUNITIES OF

THE LOWER MISSOURI RIVER

Perry H. RahnDepartment of Geology & Geological Engineering

South Dakota School of Mines & TechnologyRapid City, SD 57701

ABSTRACT

During 1997, a motorboat trip down the Missouri River from Yankton to St. Louis revealed changes in the river morphology since the time of the Lewis and Clark expedition. These changes are primarily caused by channelization and the sustained discharge from the upstream reservoirs. Hazards for pleasure craft include man-made dikes and revetments and the presence of commercial barge traffic. The lower Missouri River is an interstate recreational asset that is virtually unused.

Keywords

Missouri River, barges, fluvial processes, channelization

INTRODUCTION

On May 14, 1804, Meriwether Lewis and William Clark began their two-year expedition. They and 27 men of the permanent expedition started at St. Louis. They used 3 boats to navigate up the Missouri River: a 55-ft keelboat and two pirogues (flat-bottomed dugouts). They eventually reached the Pacific Ocean November 8, 1805. During 1806 they returned to St. Louis. The Missouri River was a natural meandering river at that time. In the 20th Century a large number of engineering features were constructed, and now the Missouri River is considered to be the most “engineered” river in the nation. Six large dams were constructed upstream (Fig. 1) in order to provide hydroelec-tric power, reduce downstream floods, and maintain a near-constant discharge downstream for the navigation of commercial barge traffic. The uppermost dam, Fort Peck Dam, MT, was constructed in 1937, and Big Bend Dam, SD, was completed in 1963. Below the lowermost dam, Gavins Point Dam at Yankton, SD, the river has been extensively channelized in order to facilitate commercial navigation. The U.S. Army Corps of Engineers (COE) is charged with maintain-ing a 9 ft deep channel from Sioux City, IA, to St. Louis, MO, The discharge of the lower Missouri River is controlled by the COE releases from Gavins Point Dam. Figure 2 shows the average monthly discharge at Yank-ton for the years 1932 to 2000. Ft. Randall dam began collecting water in 1952. Before the dam closures, the river typically had high spring discharge and had


Figure 2. Discharge vs. time of the Missouri River, 1937 to 2000. USGS gage #06467500 at Yankton.

Figure 1. Map of the Missouri River Basin (from US Army COE).


low discharge during the fall. Discharge releases are approximately 40,000 cfs during the navigation season (April through October) and 20,000 cfs the rest of the year. The discharge has been much more constant since the main-stem dams were constructed. This manipulation helps reduce flooding, and it facilitates commercial navigation (barge traffic). Throughout July and August 1993, devastating floods hit the lower Missouri and Mississippi Rivers (Rahn, 1996). The U.S Geological Survey (USGS) gage at Hermann, MO, recorded 750,000 cfs instantaneous peak discharge, well above 620,000 cfs estimated for the 100-year flood. During the 1993 flood, numerous levees failed and damage was estimated at $20 billion (Dohrenwend and Stone, 1995). The Missouri River’s elevation at Yankton is approximately 1155 ft, and its elevation at St. Louis is approximately 405 ft. It drops 750 ft over this 805-mile reach; therefore the gradient is 0.932 ft per mile. Despite this low gradient, there is an intensely powerful current. The river in this reach has a fairly smooth sur-face; it is quite wide but is not very deep. On my 1997 motorboat trip I narrowly avoided a giant dead tree which was rolling on the bottom and suddenly sprang up out of the river just in front of me. At places the streambed contains dunes that cause the water surface to swirl as a boil, making my motorboat lurch to the right or left. Dunes are the forms taken by the moving sand on the streambed; they appear to be transient, i.e., they wash out and form again. Thus the boils are unpredictable. The Missouri River has relatively good water quality as attested by the pipe-line projects being built in South Dakota that will carry river water to nearby communities. The river is relatively sediment-clear just below Gavins Point dam, but quickly acquires a suspended load and bedload as it erodes its banks down-stream (Rahn, 1977). The suspended sediment of the Missouri River averages approximately 50 million tons/year at Kansas City (National Research Council, 1993). In the reach below Omaha the “Big Muddy” still lives up to its name.

1997 MOTORBOAT TRIP

In 1997 I took a 17 ft motorboat from Yankton to St. Louis. This paper describes my observations, primarily relating to the changes in the river mor-phology since the days of Lewis and Clark. On September 28, 1997, I set out at Yankton, just below Gavins Point dam. The water impounded by this dam is called the Lewis and Clark Reservoir. The water is free flowing from this point all the way to the Gulf of Mexico. Sioux City, IA, is at Mile 734 according to the U.S. Army Corps of Engineers (COE) designation, using mileage above the confluence with the Mississippi River at St. Louis. I started at the Yankton boat access that I estimate to be Mile 805. According to COE records, on September 23, 1997, the discharge at the USGS gage #06467500 at Yankton was 64,000 cfs, well above the average 26,670 cfs. My motorboat (Fig. 3) averaged about 20 mph when functioning well. The river was at a fairly high stage and had a velocity of approximate 5 mph; thus my speed totaled 25 mph. At this rate, the trip would not take too long, but there


are problems navigating this river. The trip took me 7 full days. It is instructive to compare this travel time to that of the 1806 return trip of the Lewis and Clark expedition. They canoed past the Sgt. Floyd grave at Sioux City, IA, on Septem-ber 4 and reached St. Louis on September 23. [Below Sioux City the Missouri River today is 734 miles long, but before channelization it was 127 miles longer (Boue, 1998).] I made approximately 105 miles per day whereas Lewis and Clark made approximately 45 miles per day. I had the benefit of a motorboat on a swiftly flowing channelized river whereas they were paddling during low flow conditions through channels, chutes, and islands. The 51-mile reach from Yankton, SD, to Ponca, NE, contains the sub-merged remnants of stone walls (dikes) that were originally built to enable commercial navigation all the way up to Yankton. But here the dikes have been abandoned, and the quartzite boulders used to construct the walls now jut out from the bank, mostly hidden under water. In just a few hours after I departed from Yankton the propeller of my boat hit a large rock and broke. Thus I was left to drift down the meandering river. Drifting in the current may sound easy, but the wind takes over and can pin a drifting boat against giant log jams in an eddy behind a dike (Fig. 4). It was only with considerable effort that I managed to continue. Fortunately on Day 2 another motorboat saw me and towed me to the marina at Sioux City for a new propeller. I saw very few pleasure craft on the water or people along the shore. On this entire 805-mile reach I saw no one swimming and I saw only three people

Figure 3. My motorboat anchored in an eddy at campsite at Mile 755. This was one of the few open vistas encountered; the banks are typically covered with riparian trees.


along the shore. These were fishermen at Mile 256, near Glasgow, MO. I saw two other motorboats just below Yankton, SD, two just above Sioux City, IA, one above Omaha, NE, and five near New Haven, MO. The paucity of pleasure craft was surprising since the weather on my whole trip was sunny and warm. On the entire 805 mile reach I found only five operating marinas: Sioux City, IA; 15 miles above Decatur, NE; Parkville, MO (Mile 375); Cooper’s Landing (Mile 168), and New Haven, MO (Mile 81). Near Omaha five marinas are shown on the COE map, but all were closed (presumably damaged by the 1993 floods).

Figure 4. Log jam behind a submerged dike. October 4, 1997. Mile 160.


The paucity of marinas was an unpleasant surprise since my motorboat averaged only 5 miles per gallon, and initially I had only three 5-gallon gas cans. Thus I had to anchor, climb the bank, and hitch-hike to a nearby town to get additional gas (and gas cans). Quickly repaired on Day 2, my motorboat behaved well the rest of the trip. But there are still numerous risks involved in any recreational venture like this. The COE maintains a “navigation channel” from Mile 734 at Sioux City to the mouth of the Missouri River just north of St. Louis. I used the COE “Missouri River Navigation Charts” which provide excellent maps and information. I also found Quimbey’s Cruising Guide (1997) helpful; they mention the possibility of submerged dikes during high water, and that channel buoys are not gener-ously provided. They also warn of the scarcity of fuel docks: “There is at present one fuel dock at Mile 81.5, the next fuel dock up the Missouri, however, is 370 miles.”

CHANNELIZATION

The main engineering modification of the lower Missouri has been the con-struction of dikes (also known as spur dikes, dykes, groins, groynes, or jetties). Dikes typically are rock walls or a row of wooden piles. They are generally built on the inner (“slip-off slope”) of a meander, and help force the thalweg against the outer edge (“undercut bank”) of a meander. This is done to maintain a cur-rent swift enough to move the sediment and thus keep the channel deep enough for commercial traffic. [On the Rhine River, probably the most engineered river in the world, dikes are systematically constructed every 200 m to facilitate the large volume of barge traffic.] The COE is charged with maintaining a 9 ft deep channel between Sioux City and St. Louis. The “navigation channel” is 300 ft wide and the “minimum service flow” channel is 200 ft wide. The navigational channel is well marked with mile posts and marker buoys (green on the right side and red on left side going downstream). Figure 5 is an example of the COE navigation guide map near Mile 325, below Kansas City, showing the location of the main navigation channel (the thalweg), and the revetments and dikes. Figure 6 is an air photo showing the dikes exposed at a typical stage. [Note: On my trip, most of these dikes were submerged due to the high discharge. In fact, the marker buoys, tied by a steel cable at the end of the dike, were often submerged. Buoys that were partially submerged would suddenly bob up in front of me.] Revetments, another form of channelization, are typically constructed along the outside of many meanders to prevent the erosion and migration of a mean-der (Keller, 1976). Bank stabilization structures such as rock or pile levees are used locally on many rivers. Private landowners sometimes tie old car bodies together with a cable to mitigate erosion along meandering stream in the prairie states. Along the Missouri River the COE revetments are typically walls of rock (rip-rap) or wooden piles reinforced with lumber and rock. In some places the revetments are steel.


Figure 5. Map showing revetments and dikes along the meandering Missouri River at Mile 325, near Wellington, MO (from US Army COE).

Figure 6. Oblique air photo showing dikes and revetments (COE). The view is looking upstream into Upper Monona Bend, 32 miles below Sioux City, IA.


Maintaining a navigable channel requires dredging at places. For example, downstream at Mile 557 on the Mississippi River (near Arkansas City, AR), dredging is extensively utilized where a bar builds out and impairs commercial navigation. Near Kansas City on my trip I saw one stationary dredge loading sand onto a barge. Steel cables extended upstream from the dredge to keep it in place. Channelization concentrates the Missouri River into a single channel so that the actual river area is narrower than a natural river. Sloughs, chutes, and wetlands that gave the river its diverse fish and wildlife habitat are cut off from the river by channelization. The lack of variable discharge (Figure 3) causes changes in river morphology. At low flow, the natural Missouri River would have contained a morphologic diversity that approaches a braided stream. Numerous chutes would have carried smaller flows in shortcuts, forming islands and sand bars. Some old channels no longer connected to the river during low flow would form lakes and marshes. The COE releases have reduced the high discharge events. The annual “spring rise” has been eliminated. High discharge events tend to rejuvenate a river. Floodwa-ters scour new holes and shallows, build sand bars, and create oxbow lakes. The entire ecosystem has evolved under these variable discharge conditions, resulting in plant succession and habitat diversity. Now this has all changed.

BARGES

Despite the danger of running into submerged dikes, I found that the great-est impediment to recreation use of the Missouri River is commercial navigation (barge traffic). A barge is the vessel used for transporting material such as grain, fertilizer, and rock. It is typically a box-shaped steel container, 35 ft wide and 195 ft long. A barge is roughly 12 ft high, and the leading edge (the “rake”) of the front barge typically juts out at a sharp angle over the water. When empty, and hence high in the water (Fig. 7), the front of a barge could ride completely over a motorboat even twice the size of mine. In fact, an empty barge, anchored along the shore, still represents a danger because, drifting down into it, the current would pull a small boat (or swimmer) completely under it. The draft on a barge can be adjusted so that it doesn’t scrape the bottom or operate in the “high friction suspended sediment zone” just above the streambed. It may be only inches less than the authorized channel depth of 9 feet. Occa-sionally during very high flows some barges are loaded to drafts of 9 ft to 10 ft. During a drought, when there is only “minimum-service navigation discharge” the channel may be maintained at an 8 ft depth, and the barge drafts are then only 7.5 ft. During these low flows some of the heavier towboats stay off the river. A large (54 ft x 297 ft) asphalt barge would carry 2410 tons with a 7.5 ft draft, but fully loaded to 2945 tons would have a draft of 8.5 ft. A powerful towboat is used to propel the barges. [“Towboat” may seem like a misnomer since it pushes the barges instead of towing them.] Towboats are typically 160 ft long and 50 ft wide. Seven towboats presently operate on the


Missouri River. The towboat usually pushes more than one barge. Often two or more barges are pushed (Fig. 7), and there may even be two lines of barges, three in a row, totaling six barges. Together the barges and towboat are referred to as a “tow”. Tows on the Mississippi River may have more than 40 barges; the towboats there average 23,000 hp with three 10 ft diameter five-bladed stainless steel propellers housed in Kort nozzles (Julien, 2002). A Missouri River towboat travels about 3.5 to 5 mph upstream. The current in the main channel is about 4 - 5 mph. Towboats have powerful diesel engines. Most are turbocharged; some are supercharged (blown). The engines turn at a maximum of 1000 to 2000 rpm. Most towboats are “twin screw”, which means two wheels (propellers). Some are single screw and a handful are triple screw. Horsepower can range from 800 hp to 10,000 hp total for all engines. Some towboats may have three engines on two wheels, but generally all twin screw towboats have two engines. They have reduction gears with a ratio typically of 7:1 or 4:1 or 2.5:1. There are towboats with Kort nozzle type wheels that can swivel, but towboats used on the Missouri River have stationary wheels that use large rudders to control direction. Relative to the characteristics of tows on the Missouri River, John LaRan-deau of the COE (pers. comm., 9/30/03) offered these comments:

“You generally want a towboat that is shallower than the draft of your barges. If you get grounded you can still move the towboat to get the barges off ground. The MEMCO Barge Lines has two towboats: the HARRY WADDINGTON and the JACK FLAHAUT. These towboats have drafts of 8’6”. All the towboats can light load with fuel to reduce their drafts a few inches if the river has shoaling issues…The wheels (propellers) are in the rear tucked above the hull with the deepest part even or slightly above the bottom of the hull. Most hulls are flat bottom without keels…The draft of the towboats varies from 6.5 feet to 8.5 feet on the Missouri River. The wheels (propellers) vary in diameter but many are 6 feet in diameter. Where the bed of the river is shallow the thrust of the wheels does indeed move the

Figure 7. Tow encountered at Mile 59. The towboat is pushing three barges. The barge in front is empty and riding high in the water.


bed sediment around. During very low water conditions the tows can make their own channel sometimes. For very small boats, it is best to wait and let the tows go by.”

A moving towboat or barge with an 8.5 ft draft may actually strike the bot-tom of a 9 ft channel due to “squat”. Squat is the tendency for a ship to increase its draft because of its own forward motion. It has a simple explanation due to the Bernoulli effect, caused by the increase in velocity and hence the decrease in pressure of water moving past the middle of the ship. A rule of thumb (Tuck, 1978) is: “a foot for every five knots”. The line of sight on a meandering river is limited. Near Sioux City the me-ander wavelength is approximately 3 miles, but near St. Louis the line of sight is better where the meander wavelength is approximately 6 miles. The longer the wavelength, the longer time a pleasure craft has to pull over out of the way of an oncoming tow. A recreation boater, rounding a meander bend and facing an oncoming tow with all its momentum, has only minutes to get out of the way. An ominous warning is provided by the American Waterways web site: “A tow can travel one mile in seven minutes…and it generally takes ¾ to 1½ miles to stop. For example, if a water skier falls a thousand feet in front of a tow, the skier has less than a minute to get out of the way.” Even being missed by a tow is dangerous because the powerful engine creates a current toward the propellers from alongside the towboat.

WAVES AND SEDIMENT TRANSPORT

A moving boat produces a V-shaped train of waves from the bow of the vessel. Large bow waves are produced by a tow, particularly when it is going up-river. They can swamp a small boat. These waves may reflect off a shoreline or where shallow water (shoaling conditions) exists. Following the passage of these waves, the surface of the river may have drifting logs and debris washed out from the banks or from behind the dikes; this debris constitutes a hazard to a motor-boat. After the bow waves (and all their descendants reflected from the shore) sub-sided, I noticed another peculiar waveform. The river surface had the appearance of local rapids that seemed to occur along the route that the tow had just taken. The turbulence seemed out of place, and it persisted in place for perhaps 10 minutes after the tow went by. This turbulence is not simply prop wash (“wheel wash”), a current caused by the direct blast of water from the towboat propel-ler, because the tow could be a mile upstream by then. It seems very unlikely that the propeller could be pushing a column of water from that far away. Fluid mechanics theory shows that a jet of water discharging from a submerged nozzle slows down as it enters water at rest. Streeter and Wylie (1986) show that the jet width (b) spreads out linearly with distance (w). The expansion rate is: b = w/8. Formulas showing the velocity distribution of a submerged jet entering fluid at rest have been given by Schlichting and Gersten (2000). As the turbulence spreads out into the ambient fluid, the velocity of the central core is reduced.


Rajaratnam (1983) found that the central core of the jet flow vanishes after the jet travels an axial distance of approximately six times the jet width. I think the peculiar turbulence I saw were actually “standing waves” that de-veloped along the route taken by the tow. I suspect these standing waves formed over a ditches and dunes created by the scour of the propeller. The formation of a standing wave over a dune can be mathematically analyzed by considering that a hydraulic jump occurs at “critical flow”, where the Froude Number abruptly drops from greater than unity to less than unity. Using average estimated param-eters for the Missouri River, the Froude Number (Nf) can be calculated for a velocity (V) of 5 mph (equivalent to 7.33 ft/sec) and a depth (d) of 9 ft:

Nf = V[gd]-0.5 = 7.33 ft/sec [(32.2 ft/sec2)(9 ft)] –0.5 = 0.43

Let us assume the current is flowing somewhat oblique to this (hypothetical) dune so that the river depth is halved to 4.5 ft. Hence the velocity will be locally doubled to 10 mph (14.67 ft/sec). Therefore:

Nf = 14.67 ft/sec[(32.2 ft/sec2)(4.5 ft)]-0.5 = 1.22 Under these assumptions, the water goes from subcritical to supercritical flow over the dune. When the water continues downstream to the original 9 ft depth, it wouldwould return to subcritical flow. Thus a hydraulic jump would form just below the dune. Within a few minutes it seemed that the standing wave vanished, presumably because the dunes wash out and the streambed returned to its original morphology. Leopold (1994) describes similar transient features on the Colorado River where standing waves form above antidunes that go through buildup/breakdown cycles. Implicit in the above calculation of a standing wave is that the propeller wake is sufficient to alter the streambed morphology. The velocity of the water expelled by the propeller (also called screw or wheel) depends on its size, shape, and rpm. An example of a towboat on the Missouri River is the OMAHA, oper-ated by Blaske Marine. It is a twin screw with two diesel engines with a total of 2200 hp. Consider OMAHA’s propellers, six feet in diameter, and engine going at 800 rpm with a reduction gear of 2.5:1. The propeller would be rotating at 320 rpm. Assume there are three blades designed so that when the propeller goes around one revolution one blade would push 2.5 feet of water out the back. [Note: this assumption ignores what mariners call “slip”, the difference between theoretical and actual advance.] Therefore 3 blades would push water 6 ft out the back in one revolution. At 320 rpm, the velocity of this water = 320 revolu-tions/minute X 6 ft/revolution = 1920 ft per minute. This equals 32 ft/sec or 22 mph. Thus the current passing over the streambed would increase from 5 mph to 27 mph when a towboat passes by. The discharge of this prop wash would be: Q = VA =22 ft/sec [3.1416 (3 ft)2] =905 cfs. Prop wash greatly increases the shear stress on the streambed. In a natural channel the velocity in a 9 ft deep channel may average 5 mph, but 6 inches above the bottom it’s much less than 5 mph because the vertical velocity profile


has a logarithmic distribution (Leopold, 1994). However, when a towboat goes over, the prop wash increases the velocity just 6 inches above the bottom to 27 mph. This would cause a lot of scour. For example, a field study of erosion of sandy channels near El Paso, Texas (Jepsen et al., 2003), showed that an approxi-mate ten-fold increase in shear stress (from 0.2 to 1.6 N/m2) caused more than a hundred-field increase in the erosion rate (from 0.000,5 to 0.07 cm/sec). The equations governing sediment erosion are complex. The “critical shear velocity” needed for bedload movement is related to grain size. Bed forms chang-es from ripples to dunes to antidunes as velocity increases (Leopold, 1994). An analysis of the depth of scour caused by a towboat propeller is beyond the scope of this paper. However, recourse can be made to some empirical relationships around man-made structures (Julien, 2002). For example, below a 2.25-meter drop structure (weir) in noncohesive sand (d50 = 2 mm), the water velocity reach-es 2.3 m/sec and a scour depth of 2.7 m would occur. Water flowing at 5.9 m/sec under a sluice gate in fine gravel (d50 = 5 mm) would scour to a depth of 4.8 m. These two examples give some indication of the depth of scour by rapidly flow-ing water, and an indication of the scour that may occur as a towboat passes by. The velocities in these two examples are close to those for the propeller wash and hypothesized streambed erosion. As a tow goes by, the sediment unquestionably gets moved around. Haydel and McAnally (2004) found that sediment eroded by towboat prop wash impairs marina access on the Tennessee-Tombigbee Wa-terway. Prop wash direction is not always consistent along the navigation channel as tows head up and down the river. Towboats on the Missouri River have large rudders to control direction; hence the prop wash can be deflected laterally.

PARADISE LOST

The lower Missouri River has been changed since the days of Lewis and Clark. In a natural river, low discharge would be expected in September, exposing sand bars, channels, chutes, and other fluvial forms that approach a braided stream. In this reach Lewis and Clark, paddling their canoes downstream, described some of the morphological diversity. According to DeVoto (1953) they noted: “…the Sand bars which choked up the Missouri and confined the river to a nar-row snagey chanel…encamped on a sand bar…proceeded on to an island near the middle of the river (and) encamped.” Later in the 19th century steamboats would traverse the natural meanders of the Missouri River and experience that same morphological diversity. In low flow these steamboats would wind around the chutes, trying to find the thalweg. But today there is little morphological diversity. The thalweg is confined into a monotonous channel, defined by COE structures such as dikes and revetments. There are no islands or chutes, and few irregularities in the shoreline offering protection from waves and current where a pleasure craft could anchor. The navigation channel is kept at 200 ft wide and 9 ft deep. The discharge is maintained throughout the navigation season, and is nearly constant the whole year. There is debate today concerning the release of water below the main stem dams. The lower states (MO, NE, IA) support steady discharge to maintain


barge traffic and supply adequate water for cooling electric power plants and for dilution of municipal wastewater. Steady discharge necessitates the fluctuation of the upstream reservoirs as they store the high spring runoff. The upper states (SD, ND, MT) want less fluctuation of the reservoirs for improved fishing, ma-rina access, and recreational opportunities in the reservoirs. On the other hand, marina owners in the lower Missouri River prefer steady discharge so they have consistent river access. Clearly, the schedule of releases from the reservoirs cannot satisfy everyone. Environmentalists advocate a discharge that would mimic the natural high “spring rise” and a low discharge by the end of the summer. They are concerned about the loss of fish habitat on the lower reach because of the absence of the normal discharge fluctuation. According to Brokaw (2002), who grew up in Yankton in the 1950’s: “Missouri River fishing in those days before the massive dams altered the water flow was an ichthyologist’s smorgasbord.” Today commer-cial fishing is virtually nonexistent. Lower Missouri River backwater fish habitats have been lost due to channelization, leading to a reduction of fish diversity and biomass (Sandheinrich and Atchinson, 1986; Powell and Chipps, 2002). The thalweg of a natural meandering stream crosses the stream centerline on each meander. This gives rise to the meander morphology known as the undercut bank and slip-off slope. The morphology of the Missouri River channel has been changed into a trapezoidal cross-section; this loss of diversity limits fish and wildlife habitat (Hesse and Sheets, 1993). Streambed scour caused by towboat propellers as described in this paper would have a severely deleterious effect on the river’s ecosystem. On my trip I talked to natives who used to set out gill nets for catfish and carp; they have abandoned this because of the lack of fish. My own experience at fishing was dismal; I set out bait lines every night and never caught a thing. Despite these shortcomings, the natural beauty of this riverine setting is in-spiring. In most reaches, I suspect that the riparian trees and adjacent bluffs look nearly the same as when Lewis and Clark floated by. The banks are lined with big deciduous trees. In Gasconade County, MO, the river enters a picturesque gorge-like physiographic setting. I saw monarch butterflies fluttering across the water as they headed to Texas. Despite the large urban populations nearby, in many respects this entire 805-mile reach of the Missouri River is similar to a wilderness because of the absence of people, cars, and houses. In order to travel the Missouri River by small vessel it is necessary to plan ahead. It is recommended to obtain good maps such as the COE “Navigation Charts”, become familiar with the engineered structures, and understand Coast Guard Rules. A marine band radio (VHF channel 14 and 16) is not essential but is recommended so that if you have trouble you can contact an oncoming towboat. A towboat can usually see you during the day, but if you are drifting at night it is all over. Another important bit of advice is to have plenty of gasoline since fuel docks are far and few between. The Missouri River is a national recreational asset that is virtually unused. People don’t go on the river because the barges intimidate them. As long as one of these behemoths might appear around the next bend there is the danger of getting run over. A greater appreciation of the recreational opportunities would


be recognized if barge traffic were restricted. For example, they could be ex-cluded certain times such as summer weekends. The river was originally made into a navigation channel for an estimated 12 million tons/year of freight. But according to the “American Waterways” newsletter (August, 2003), during the past decade it has only averaged 8 million tons/year. Grunwald (2003) called the engineering accomplishments “a wasted effort”, and said the economic benefits of recreation would exceed the barge traffic. In addition to the barge traffic, the sustained discharge and the channeliza-tion structures present an adverse impact to recreation. The current moves swiftly all summer. There are no islands or sand bars for a recreationalist to stop along. Chutes, channels, and oxbows have vanished. A natural river would provide recreational opportunities. Now they are virtually lost.

ACKNOWLEDGEMENTS

John R. LaRandeau from the Omaha Office of the COE kindly supplied data concerning towboats, barges, and the Missouri River navigation channel.

REFERENCES CITED

Boue, K. 1998. Missouri River restoration. Nebraskaland, March, 1998, p. 16-23.

Brokaw, T. 2002. A long way from home. Random House Trade Paperbacks, New York, 233 p.

Daniels, W. W., et al. 1976. Environmental impact of stream channelization. Water Resources Bulletin 122: 799-812.

DeVoto, B., ed. 1953. The journals of Lewis and Clark. Houghton Mifflin Com-pany, Boston, 504 p.

Dohrewend, J. C., and B. D. Stone. 1995. Impact of the 1993 floods in the Up-per Mississippi River Basin. U.S. Geological Survey, Yearbook, Fiscal year 1994, p. 12-17.

Grunwald, M. 2003. Washed away: Bush vs. the Missouri River. New Republic, October 27, 2003.

Haydel, J. F., and W. H. McAnally. 2004 (Abs). Transportation Research Board. 83rd Annual Meeting. January 11-15, 2004.

Hesse, L.W., and W. Sheets. 1993. The Missouri River hydrosystem. Fisheries, v. 18, n. 5, p. 5-14.

Jepsen, R., R. Langford, J. Roberts, and J. Gailani. 2003. Effects of arroyo sedi-ment influxes on the Rio Grande River channel near El Paso, Texas. Envi-ronmental and Engineering Geology, vol. IX, no. 4, p. 305-312.

Julien, P. Y. 2002. River mechanics. Cambridge U. Press, UK, 434 p.Leopold, L. B. 1994. A view of the river. Harvard University Press, Cambridge,

MA, 198 p.National Research Council. 1993. Solid-earth: sciences and society. National

Academy Press, Washington, DC.


Powell, K. A., and S. R. Chipps. 2002. Longitudinal patterns in fish community composition of upper Missouri backwaters. South Dakota Academy of Sci-ence 81: 211-217.

Rahn, P. H. 1977. Erosion below main stem dams on the Missouri River. As-sociation of Engineering Geologists Bulletin 14:157-191.

Rahn, P. H. 1996. Engineering geology, an environmental approach. Prentice Hall, Upper Saddle River, NJ, 657 p.

Rajaratnam, N. 1983. Theory of turbulent jets. In Cheremisinoff, N.P., and R.Gupta. Handbook of fluids in motion. Butterworths. Boston. Chapter 10.

Sandheinrich, M. B., and G. J. Atchinson. 1986. Fish associated with dikes, revetments, and abandoned channels in the Middle Missouri River. Proceed-ings, Iowa Acad. Sci.. 93 (4): 188-191.

Schlichting, H., and K. Gerten. 2000. Boundary layer theory. Springer, Berlin.Streeter, V. L., and E. B. Wylie. 1986. Fluid mechanics. McGraw-Hill, Inc., New

York, 586 p.Tuck, E. O. 1978. Ships in restricted waters. In Van Dyke, M., J. V. Wehausen,

and J. L. Lumley, eds., Annual review of fluid mechanics. Annual Reviews, Inc., Palo Alto, CA, p. 33-46.

Date post:	07-Jul-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

Proceedings of the South Dakota Academy of …South Dakota Academy of Science Volume 83 2004...

Documents