The Middle Rio Grande:Its ecology and management

Jeffery C. Whitney’

Abstract.-The Middle Rio Grande (MRG) riparian forest, or “bosque”, repre-sents the largest cottonwood gallery riparian forest in the southwestern UnitedStates. This reach of the Rio Grande extends from Cochiti Dam downstream260 Km to San Marcial, New Mexico. It constitutes 8% of the river’s totallength and 34% of if its length in New Mexico. The valley traverses threemajor biotic communities, as defined by Brown and Lowe (1980). The MRGreach can be subdivided into 4 reaches which coincide roughly with the 4geologic basins or “grabens” along this portion of the Rio Grande Rift. Thissystem has been affected by man’s activities throughout prehistoric andmodern eras. The Rio Grande is regulated for water supply (primarily irriga-tion) and flood control. The effects of this interaction have contributed to thecharacter of the riparian ecosystem in its current expressron. Over 40% ofNew Mexico’s population lives within the MRG reach. This paper will discussthe climate, geology, hydrology, subsequent river mOrphOlOgy, and anthropo-genie factors which contribute to the past and current expressions of theriparian habitat associated with the Middle Rio Grande.

INTRODUCTION

The Rio Grande is one of the longest rivers inNorth America (1900 miles). The Rio originates inthe southern Rocky Mountains of Colorado, flowsthe whole length of New Mexico and forms theentire border between the state of Texas and theRepublic of Mexico (fig. 1). The Rio is the greatestsource of permanent water in the desert southwestother than the Colorado River. It is home to thelargest cottonwood forest in North America,locally referred to as the “Bosque”.Human populations have increased dramaticallyalong the Rio Grande since European settlement.Human use of water for irrigation and consump-tion, and human use of land for agriculture, urbancenters, livestock grazing and recreation have

l Middle Rio Grande Coordinator, U.S. Fish and Wildlife Service,Albuquerque, NM.

changed Rio Grande ecosystems by altering floodcycles, channel geomorphology, upslope processes,and water quality and quantity. Such abioticchanges have influenced the biological diversityand ecological functions of the MRG, altering thedistribution, structure, and composition of riparianplant and animal communities.

The Rio Grande basin above El Paso, Texas, isone of the oldest regions of agriculture in theUnited States. Agricultural activity extends backcenturies to prehistoric inhabitants of the RioGrande valley (fig. 2.) and includes the seven-teenth and eighteenth century Pueblo Indians andSpanish colonists, and European-Americans in thelatter part of the nineteenth century (Wozniak1987). More recent history of the region involvesdisputes and concerns over management, irriga-tion, and distribution and delivery of upstreamwaters to downstream users in an attempt at fairsharing between concerned parties. Because of the

4

This file was created by scanning the printed publication.Errors identified by the software have been corrected;

however, some errors may remain.

long history of agricultural activity, Rio Grandewater is tied to public laws governing its convey-ance, storage, and use. The close connection be-tween legislation and flow of water through theRio Grande is largely responsible for the presentphysical state of the river, floodplain, and associ-ated riparian community. Changes in the flood-plain ecology probably began shortly after humansettlement in the region, and change has continuedrelatively unabated with increasing population.(Bullard and Wells 1992).

Colorado. Above Velarde the drainage basin areais about 27,325 sq. km., including the Closed Basin.The Rio Chama, one of the most important tribu-taries in the study area, has its headwaters in theJemez, Conejos, and San Juan Mountains of NewMexico and Colorado. The MRG extends fromCochiti Dam downstream 260 river km (160 mi) toSan Marcia1 (fig. 3). The MRG constitutes 8% df theRiver’s total length and 34% of its length in NewMexico. The middle valley’s direct drainage ac-counts for 7% of the total Rio Grande drainage andabout half of New Mexico’s direct tributary drainage.

LOCATIONPHYSIOGRAPHY AND GEOLOGY

The MRG is part of the larger Rio Grande fluvialsystem (fig. 1). The Rio Grande headwaters liealong the Continental Divide at elevations rangingfrom 2,440 m to 3,660 m in the San Juan Mountainsof southern Colorado. The entire area of the RioGrande drainage basin is about 470,000 sq. km ofwhich about 230,OO sq. km. are in the United Statesand the remainder in Mexico (Hunt, 1974). Theriver flows south from Colorado through thelength of New Mexico and then forms the interna-tional boundary between Texas and the Republicof Mexico. The drainage basin area above ElephantButte is about 76,275 sq. km., including 7,615 sq.km. in the Closed Basin of the San Luis Valley in

Hydrologic characteristics of the Rio GrandeBasin, such as infiltration, runoff, and sedimentdischarge, are dependent on the geology, geomor-phic evolution of tributary basins and late Tertiaryand Quaternary geologic and climatic history.Structural geology (such as faults and folds) of aregion governs spatial and geometric relations ofrock units in that region. Geologic structures andlithology influence the development of topographicfeatures, river and tributary position, and landscapeevolution. Tectonic activity can produce measur-able effects on channel and sediment transportcharacteristics (Ouchi 1983, 1985; Schumm 1986).

Figure 1. Middle Rio Grande study area (from Bullard and Wells, 1992).

5

3b8CG

Pta

154016001700 1800165018801896 1925 1960 1970 1980 1990

Y E A R

Figure 2. Historical account of acres of land undercultivation in the Middle Rio Grande Valley (acres/2.47zhectares) (From Crawford et al, 1993).

The location of the Rio Grande is controlled bythe dominant geologic structure of the region, theRio Grande Rift. The Rio Grande Rift is a lineartopographic feature that separates the Great Plainsfrom the Colorado Plateau (Hawley 1978) moun-tain ranges, which can influence weather patterns,are a direct result of geologic processes. The rift,active for at least 18 million years (Wilkins 19861, ischaracterized by extension, seismicity, local tec-tonic uplift, and volcanism (Loainski et al. 1991).The location of early trade routes was influencedby the spatial arrangement of mountain rangesthat were natural barriers to travelers. Indigenouspopulations and early settlers in the region soughtareas of suitable climate, access, and availability ofwater. Thus, the presence of the Rio Grande Rifthas influenced human settlement patterns in theregion.

The extent and type of bedrock can influenceinfiltration and runoff characteristics. These factorscan dramatically influence tributary basin evolu-tion, discharge characteristics, main stem flow, andmain stem evolution and integration (Leopold etal. 1964; Schumm 1977; Richards 1982; Kelson 1986;Wells et al. 1987, Bullard and Wells 1992). Bedrocktype influences vegetation types and densities,

Figure 3. Setting and institutional boundaries in theMiddle Rio Grande (from Crawford et al, 1993).

which in turn influence infiltration and runoff,landscape stability, soil development, and sedi-ment supply. Soil development is importantbecause natural, progressive changes in physicalproperties of soils occurring through time alter thenature of the land surface, including vegetationcommunities, infiltration (decreases with increas-ing age), erosion, and runoff and discharge.

PHYSIOGRAPHIC REGIONS

The Rio Grande basin lies in five physiographicprovinces: the Coastal Plain, the Great Plains, theBasin and Range, the Colorado Plateau, and theSouthern Rocky Mountains (Hunt 1974). The MRGand its tributaries are located within the latterthree provinces. From about Santa Fe southward,the rift is in the Basin and Range PhysiographicProvince which separates the Colorado Plateau

6

Province to the west from the Great plains Prov-ince to the east. (Crawford et al. 1993).

The MRG valley is actually a series of basins.These grabens (depressions) formed a series oflinked, but slightly offset, depositional basins, eachof which contained its own ephemeral lake. Overtime, the surface water eroded canyons betweenthe intervening bedrock sills that defined thebasins, integrating the area into the Rio Granderiver system (Bullard and Wells 1992). Thethrough-flowing ancestral Rio Grande drainagedeveloped into a single river about 5 million yearsago (Lozinski et al. 1991). The basins in the MiddleRio Grande are:

l Santo Domingo Basin

l White Rock Canyon to San Felipe

l Albuquerque Basin

0 San Felipe to Isleta

l Belen Basin

l Isleta to San Acacia

l Socorro Basin

l San Acacia to San Marcia1

CLIMATE

The hydrology and morphology of the RioGrande are ultimately dependent on the climateand geology of the area. An overview of thesetopics will create a foundation of understandingfor later discussions.

The valley’s climate is characterized as havingmoderate temperatures and being semiarid aboveBernalillo to arid south of Bernalillo (Tuan et al.1973). Temperatures increase and precipitationdecreases from north to south. Annual averagemaximum temperatures, which usually occur inJuly, range from 21°C (69°F) at Cochiti Dam to24OC (76°F) at Bosque de1 Apache National Wild-life Refuge (NWR). Annual average minimumtemperatures (January) are about 4°C (40°F)throughout the valley. The growing season alsoincreases southward through the valley. InBernalillo and Albuquerque, the typical frost-freeperiod begins in early May and extends through

mid-October, lasting on average 160 days. InSocorro, the average period is 197 days, beginningin Mid-May and lasting through late October.

The Rio Grande drainage basin is located in atransitional climatic zone between the Gulf ofMexico and the Pacific rainfall provinces. Complexmeteorological conditions exist in this region, andthese conditions are further complicated by theorographic influence of surrounding mountainranges and global circulation patterns.

The MRG basin has an arid to semiarid climatetypical of the southwestern United States. Theclimate is characterized by abundant sunshine, lowrelative humidity, light precipitation, and widediurnal temperature fluctuations. The averageannual precipitation varies from 178 MM (7 in.) to380 mm (15.25 in.) over two-thirds of the basin andmay exceed 635 mm (25 in.) only in the highmountain areas. Winters are generally dry, andsnow rarely remains on the ground at low eleva-tions for more than 24 h. Snowfall in the highmountains composes 30-75% of the total annualprecipitation; in the remainder of the basin snow-fall composes less than 25% of the annual precipi-tation. Summer precipitation supplies almost halfof the annual moisture. Most of the rain falls inbrief, though sometimes intense, convective thun-derstorms (fig. 4). These summer thunderstormshave a considerable moderating effect on daytimetemperatures. Prevailing winds are from thesouthwest and typically are continuous during thespring months. Evaporation rate is high through-out the lower elevations of the basin and is highestin the southern part of the basin, where aridconditions exist.

Storms in the region are of two types; localthunderstorms that result from orographic orconvective lifting, and frontal storms resultingfrom the interaction of two or more air masses.Generally, precipitation during storm periods lastsless that 24 h, although precipitation intensity maybe extremely high at some locations within thegeneral storm area. Precipitation periods lastingmore than 24 h are generally associated withtropical disturbances related to hurricane activityin the Gulf of Mexico or in the Pacific Ocean off thewest coast of Mexico.

Storms are seasonal with respect to type andmagnitude. Summer months, June through Au-gust, are normally characterized by intermittent

.

7

50 1

Figure 4. Monthly average precipitation distribution in theMiddle Rio Grande Valley (after Crawford et al,1993).

importations of tropical maritime air masses.Orographic lifting or convective action results inisolated shower activity, which is often intense butgenerally localized.

Vigorous air-mass activity occurs during wintermonths, November through February, and ischaracterized by a series of cold fronts of polarPacific air moving eastward or northeastward(Maker et al. 1972). This results in snow in thehigher elevations and rain in the lower elevations.Due to the northerly path of the storms precipita-tion in the southern part of the basin is generallylow.

The periods transitional to summer and winter(March through May and September throughOctober) are associated with some of the largestflood-generating storms. Greater temperaturedifferences between air masses are reflected inincreased air-mass instability.

Runoff in the basin comes largely from springsnowmelt and spring and summer convectivethunderstorms; it ranges from ~25 mm (1 in.) to>255 mm (10 in.) in the mountains. About 70% ofthe runoff occurs from May to August duringsnowmelt and thunderstorm activity (fig. 5).

FLUVIAL CHARACTERISTICS

The physical nature of the Rio Grande, and itstributaries, varies with its position in the drainagebasin. This is a direct reflection of the geology andtopography of the physiographic regions traversedby the river. Gradient, channel pattern and width,

3 4,000 -

z

z 3.000 -

z

ztl 2,000 -P

xi

2 1,000 -0zj

” Ott NW Dee Jan Feb Mar Apr May Jun Jul Aug Sep

- Mean cfs per Month

Figure 5. Mean monthly discharge in cubic feet persecond (cfs) of the Rio Grande at the Otowi gaugeabove Cochiti Lake (U.S. Geological Survey data,1895-l 991).

discharge, and sediment load are variable through-out the length of the river. Discharge and sedimentload characteristics will be discussed in moredetail in separate sections.

GRADIENT OF THE RIO GRANDE

Relief is high in headwater regions, and tribu-tary streams characteristically flow through steepcanyons on their way to the San Luis Valley;gradients locally may be tens of meters per kilome-ter. The river has a gradient of about 0.56 m/kmthrough the San Luis Valley. Through the RioGrande Gorge, river slope ranges from 2.25 to~28.4 m/km. From Velarde to Cochiti Reservoirthe downstream end of White Rock Canyon) rivergradient is about 1.9 m/km. From below CochitiDam to Elephant Butte the gradient is about 0.76m/km.

CHANNEL OF THE RIO GRANDE

The channel of the Rio Grande varies dramati-cally with geographic location within the riverbasin. Channel characteristics such as width andsinuosity are strongly influenced by positionwithin the drainage basin and proximity to tribu-taries that discharge large volumes of sedimentinto the main stem.

8

The width of the Rio Grande Valley ranges from~200 m (656 ft.) in the Rio Grande Gorge to 1.5-10km (1-6 mi) from Velarde to Elephant Butte, withthe exception of White Rock Canyon and the SanMarcia1 Constriction. Short canyons or narrowsalso exit at San Felipe, Isleta, and San Acacia at theboundaries of the sub-basins within the RioGrande Rift The floodplain of the Rio Granderanges from 150 m (492 FT) or less in the RioGrande Gorge to greater than 1 km c.62 mi) in thereaches from Velarde to White Rock Canvon and

.from Cochiti Dam to San Acacia.The channel is narrowest in the bedrock canyons

and widest in the broad alluvial valleys down-stream from Bernalillo. Generally, the channel is60-90 m (196-295 ft) wide, flows on shifting sandand gravel substratum, and has low, poorlv de-fined banks (Lagasse 1980). Within the MRG thefloodway is largely confined between earthenlevees and is cleared for much of the length,especially in urban areas and areas prone to high-est aggradation. The floodplain contains a mixtureof cottonwood (Pq711114.5 F~r)r?lolztii), willow (%/ixspp.), Russian-olive (Elneaby~zus m~~ustifolia), and saltcedar (7‘al~lnris chinc~sis), which together form adense growth of riparian woodland (known asbosque), interspersed with pasture and cultivatedland (Lagasse 1981). The existing contiguousbosque that abuts the Rio Grande is generallylimited by the system of levees or natural bluffswhere such features are present. In the southernhalf of the valley where the bosque is at its widest,the bosque is up to 4-5 km (2.5-3 mi) wide.

The Middle Rio Grande is slightly sinuous withstraight, meandering, and braided reaches. Theriver is generally characterized by a shiftingsandbed in the reaches and bv a gravel riverbed inthe Cochiti Reach. Although k perennial river,there are reaches of the Rio Grande that experienceno surface flow during some summer months indry climatic periods (Crawford et al. 1993). Theformation of sediment bars in the channel duringlow-flow periods and, in particular, during therecession of flood flows, together with rapidgrowth of vegetation, generally determine thechannel configuration within the levees. In someplaces the floodway is unstable (i.e. the channel isnot confined to a fixed position). In these areas, thechannel has virtually no banks, and the bed of theriver is at or above the land surface outside the

levees due to sediment deposition between thelevees. Braided meandering patterns are especiallycommon downstream from major .scdiment-supplving tributaries such as the Rio Puerto andthe Rio Salado and other small, unregulated, high-sediment-discharge tributaries in the reach belowCochiti Dam.

The addition of numerous flood control ands.ediment control structures on the Rio Grande andtributaries has eliminated some of the problemsformerly associated with flood-transported sedi-ment discharged into the main stem. On the otherhand, flood control structures have added to theproblem of channel migration in some reaches ofthe river downstream of dams (Lagasse 1980,1981;Bullard and Wells 1992).

DISCHARGE OF THE RIO GRANDE

The Rio Grande is a perennial river that receivesthe majoritv of its discharge from late spring snow-melt and rain storms. Summer convective stormsproduce runoff in isolated parts of the basin, whichmav alter the hvdrolog:v for bnr>f p’rrlod>.

The malorltv ot the discharge fair thr &lRGcomes from the headwaters of the Rio Grande inColorado and from the Rio Chama. The Rio Chamajoins the Rio Grande 3S miles north of CochitiReservoir. The Rio Chama is assured of perennialdischarge because of the San Juan-ChamaTransmountain diversion (SJO Project and damsalong the Rio Chama and its tributaries (U.S.Bureau of Reclamation 1981). Average annualdischarge for the Rio Grande into the Gulf ofMexico is about 9,000,OOO acre-feet (Hunt 1974).The annual runoff in headwater regions rangesfrom 215,000 to l,lOO,OOO acre-feet, with andaverage of 660,000 acre-feet (U.S. Corps of Engi-neers 1989).

The Rio Grande has some of the longest streamgaging records in the United States; however,these records are not necessarily the most reliable(Bullard and Wells 1992). The Embudo gage nearthe southern end of the Rio Grande Gorge wasinstalled in 1889 and has nearly 100 years ofrecord, although not continuous. Reliability andcontinuity of stream gaging station data are prob-lems throughout the United States, and The RioGrande is no exception.

9

The main stem discharge of the MRG can becharacterized by 10 gaging stations: Embudoupstream from Velarde, San Juan Pueblo (discon-tinued in 1987), Otowi Bridge near San Ildefonso,below Cochiti Dan, San Felipe, Albuquerque, RioGrande Floodway near Bernardo, RIO GrandeFloodway at San Acacia, Rio Grande Floodway atSan Martial, and below Elephant Butte Reservoir.Annual average flow at Otowi Bridge (fig. 6), isabout l,lOO,OOO acre-feet; downstream at SanMarcia1 above Elephant Butte Reservoir, theannual average flow is 745,000 acre-feet (U.S.Army Corps of Engineers 1989).

Due to extensive agricultural activity in theMRG nearly all Rio Grande water is appropriated.Releases from upstream reservoirs, under non-flood conditions, are regulated to make reservoiroutflows equal to inflows in order to meet waterdemands. Irrigation accounts for about 90% ofdemands. Irrigation accounts for Rio Grande waterused in the region; however, water diverted foragricultural purposes is not fully utilized. About67% of all diverted water does not reach farm-lands. This water consists of transportation losses(spills, seepage losses to unlined canals), evapo-transpiration, and groundwater recharge. About

2.600

2,400

2,200

3 2,000

238 1,800

58 1,600

6 1,400

2

i 1,200 1,000t=i i 8 0 025 6 0 0

iG5

4 0 0

I- 2 0 0

0

45% of all water diverted eventually returns to theriver. About 33% of water diverted reaches thefarms; crops use about 55% of this amount (orabout 20% of the total diverted from the river).About 35% of the total diverted water is lost toevapotranspiration or groundwater recharge (U.S.Army Corps of Engineers 19793.

SEDIMENT LOAD OF THE RIO GRANDEAND TRIBUTARIES

Suspended sediment loads for the Rio Grandeand tributaries are variable. These are regulated toa certain degree by flood and sediment controlstructures, especially in the regions above Albu-querque. Tributaries, however, can be majorcontributors of sediment to the Rio Grande. Anincrease in sediment supplied to the Rio Grandecan have dramatic effects on river behavior andgeomorphology both upstream and downstream(Schumm 1977; Lagasse 1980,1981).

Based upon data reviewed by Bullard and Wells(1992), the Rio Puerto which has half of the drain-age area of that of the Rio Chama and the JemezRiver contributes far more sediment than that of

Figure 6. Total annual flow, Rio Grande at the Otowi gauge (from Allen et al, 1993).

1 0

6

-4 ’ ::; Exvew

I5 La Nha

Years/

- 6 I I I 1 .t I I I, I I i1 8 9 5 1909 1923 1937 1951 1955 1979 1 9 9 3

Year

Figure 6a. Drought severity index 1895 through 1988 and El Nina and La Nina events over the past 50 years for theMiddle Rio Grande Valley (after U.S. Army Corps of Engineers, 1991).

4,000r - o- Post-dosure

3,500

t

-c Pre-closure 0,I -s

, 0\

I \

3.000c

$ 2,500

!W2 2,000

8L 1,5000.-30 1,000

500

I \I \

I \, I

, \\

Jan Feb Mar Apr May Jun Jul Aug Sep Ott Nov Dee

Month

Figure 6b. Fifteen-year average of Rio Grande monthly mean discharge for pre- and post-Cochiti Dam closure periods atSan Felipe gauge (from Crawford et al, 1993).

11

the former. The authors attribute this to the factthat the Chama and Jemez rivers are dammed andthe Rio Puerto and Rio Salado are not. The conclu-sion is obvious and has merit. However, therationale may be a bit misleading since the charac-ter, including soils, elevation, vegetative groundcover of the respective watersheds are not similar.Flow regimes are also very dissimilar in that theChama is perennial, the Jemez tends to be, and theRio Puerto lacks discharge of water or sedimentfor many days of the year (Heath 1983). Leopoldfound that 82% of the sediment transported bv thistributarv occurs during events that recur aboutonce per year (Leopold et al. 1964).

CONTAMINANTS

An additional effect related to sedimentationand river siltation is the accumulation of toxicmaterials in the sediments. Popp et al. (1983) andBrandvold et al. (1984) conducted studies on thesediments of the Middle Rio Grande system. Thevfound that substantial quantities of cadmium,mercury, lead, uranium, and pesticides (78 differ-ent concentrations ranging from undetectable to>500 micrograms per liter are being transported bythe Rio Grande and deposited in Elephant ButteReservoir. These materials are primarily bound tosediment, although an unknown amount of cyclingfrom sediments to the water column occurs in thereservoir (Bullard and Wells 1992).

ANTHROPOGENIC FACTORS

Some references have already been made to theanthropogenic factors that have played a role increating the riparian habitat as currently evidencedby the “bosque” found within the MRG. Thesestructural changes and dewatering combined withthe abiotic features previously discussed combineto create a highly modified and controlled system.

Although there is no evidence of any majorclimatic changes within the past 5,000 years (Gully1977), there are indications of climatic variability(fig 6a).

The river has been a focus of human settlementand development since prehistoric times. Thissection addresses the hydrologic resource trends

from about 5,000 years ago up to the present.Generally the MRG was a braided, slightly sinuousaggrading river with a shifting sand substrate. Inthe past, as now, the slope of the riverbed de-creased from north to south and tributaries’ contri-butions of water and sediment were important indefining the river’s local and overall morphology.

Because there were no diversions and because ofthe relative hydrologic stability of the system,Crawford et al. believe that the Rio Grande gener-ally supported perennial flows. Exceptions couldhave occurred during periods of prolongeddrought and would have been more prevalentfarther downstream. With no water regulation, theriver’s hvdrograph would have reflected theseasonal events of snowmelt runoff and summer/fall precipitation (Fig. 6; note that these riverdischarge records do not reflect “natural” flowsbecause.upstream storage and diversions werealreadv in place during the period of record, butthev do indicate the general shape of thehydrographi.

The total flow in the MRG also fluctuated fromvear to vear in response to annuai climatic \:ariabii-itv. Figure 6 graphs the total annual Rio Grandeflows at the Otowi gauge above Cochiti over thepast 100 vears (fig. 6). Although these data alsoinclude the eftects of human water managementpractices, they too are indicative of this annualvariability. Figures 6a and 6b show temporalclimatic variability and the effects of Cochiti Damon the mean discharge respectively.

As human settlement and irrigated agricultureexpanded in the middle vallev and upstream in theupper Rio Grande Basin, more irrigation water wasdiverted from the river reducing total river dis-charges. The further downstream one proceeded inthe system, the less water there was. Prior to theconstruction of storage and flood control facilities,diversions from the Rio Grande and some of itstributaries were limited to the growing season.Other seasonal flows, peak runoff, and precipita-tion flows were not affected. By 1913, storagereservoirs in the headwaters of the Rio Grande hadbeen built, and in 1935 the MRGCD completed ElVado Reservoir on the Rio Chama (Shupe andFolk-Williams 1988). These facilities began to takepeaks off of some of the high river discharges andto increase the duration of lower flows. The expan-sion of these reservoirs and the addition of the

1 2

flood and sediment control dams and reservoirsfurther accentuated this trend.

Other water management facilities have influ-enced the hydrology of the MRG. The 120 km (75mi) long Low Flow Conveyance Channel, itsdownstream half operational in 1954 and its fulllength completed in 1959, reduced flows in theriver channel in the Cicero Reach. The San Juan-Chama Project, completed in 1971, imports up to110,000 acre-feet of San Juan River water from theColorado River Basin to the Rio Chama/RioGrande basins, 69,100 acre-feet of which is deliv-ered to or through the middle valley. The effect ofthis importation has been to increase mean dailyflows. In addition, the City of Albuquerque’sannual treated wastewater discharge into the RioGrande is currently about 60,000 acre-feet (R.Hogrefe, pers. comm in Crawford et al. 1993).

In all discussions regarding river morphology, itis important to recognize the differences withinspatial and temporal scales. To describe a riversystem as being in a state of dynamic equilibrium(or energy balance) does not mean that it is static.To the contrary, this equilibrium results from acollection of processes that are by definition predi-cated on change. For example, even during periodswhen the entire river system is considered to be ina state of dynamic equilibrium, changes constantlvoccur in subareas as small as the outside band of ameander, or as large as manv river kilometersupstream and downstream from a tributary inflow.Likewise, this state of dvnamic equilibrium canaccommodate climatic deviations from the normdistinguished between natural and human-causedperturbations. The geomorphic processes triggeredin response to a change in magnitude or durationof a variable, regardless of the cause, will be thesame (Leopold et al. 1964; The river constantlvadjusts, always trying to establish a new equilib-rium between its discharge and sediment load(Bullard and Wells 1992).

Prior to measurable human influence on thesystem, up to the 14th century (Biella andChapman 1977), the river was a perennially flow-ing, aggrading river with a shifting sand substrate.As stated, its pattern was, as a rule, braided andslightly sinuous. The river would freely migrateacross the floodplain, the extent being limited onlyby the valley terraces and bedrock outcroppings.The Rio Grande’s bed would aggrade over time;

then, in response to a hydrologic event or series ofevents, it would leave its e1evate.d channel andestablish a new course at a lower elevation in thevallev. This process is called river avulsion(Leopold et al. 1964). Although an aggradingsvstem, the Rio Grande was in a state of dvnamicequilibrium, providing periods of stabilitv thatallowed riparian vegetation to become establishedon riverbends and islands alternating with periodsof instabilitv (e.g. extreme tlooding;, that provided.bv erosion and deposition, new locations forriparian vegetation.

The earliest phase of significant water develop-ment activities (from about A.D. 1400 through theearly part of this century) progressively decreasedriver flows as irrigated agriculture increased. Moreinfluential on the morphology of the river, how-ever, was the increased sediment deposition intothe ecosystem resulting from land-use activities inthe watershed. When coupled with natural cli-matic variability, the net effect was to acceleratethe raising (aggregation) of the riverbed and ,accordingly, the frequency of overbank floodingand the river avulsion. The channel configuration,while still braided and sinuous, began to broadenand became shallower. Because the increasingrapiditv of channel movement, riverbanks andislands were as a rule less stable. This likelv con-tributed to an increased frequency of floods.Between 1822 and 1941, a total of 46 moderatefloods was recorded along the reach (Crawford etal. 1993). During nonflood periods, diminishedriver flows caused the active channel to retreat tofewer, narrower channels within the wide andshaiiow sandv riverbed.

During the next phase of human interacticonwith the river, from the mid-1920’s through 1950, asystem of levees were constructed to constrain theriver to a single floodwav through portions of themiddle valley. Concurrently, water diversions inthe middle vallev and upstream in the Rio GrandeBasin increased.-This had the net effect of furtheraccelerating channel aggradation, especially inthose areas where levees concentrated the deposi-tion of sediment in the floodway.

In the contemporary phase of human watermanagement beginning in the early 1950’s, thesediment and flood control structures constructedin the upper portion of the MRG valley acceleratedthe reversal of channel aggradation in the Cochiti

13

and Albuquerque reaches. The lowering riverbedis resulting in a more incised and sinuous single-channel river (see Fig. 7 for a visual example in theBelen Reach). This process becomes less pronouncedwith downstream distance from Cochiti and JemezDams. With reduction of the peak flows, whereunregulated tributaries and arroyos such asCalabicillas Arroyo discharge into the Rio Grande,adequate flows are not available to transport thesediment. Sediment deltas are more persistent;they reduce river gradient upstream (tending toincrease aggradation) and increase the gradientdownstream (tending to reduce aggradation).

The channel modification process, describedabove, immediately affected the river’s channelmorphology. To increase the water delivery effi-ciency and flood flow capacity within the flood-way the BOR initiated a river channel maintenanceprogram in 1953. This included Bank stabilization,river training, sediment removal, and vegetationcontrol. Although the techniques have evolvedover the years, the program continues. Within the

Figure 7. Changes from braided to single channel, 1935-89, portions of Belen Reach, Middle Rio Grande(from Crawford et al., 1993).

stabilized floodway, reaches of the MRG have beenstraightened, the irregularity of the channel widthhas been reduced, and the riverbanks have beenstabilized.

INSTITUTIONAL INFRASTRUCTURE

The waters of the Rio Grande are managed byan interwoven fabric of federal, state, interstate,and international water laws, agreements, andregulations. The fabric defines how water is re-leased through the system, influencing not onlythe quantity of water, but often the timing of thereleases as well. The following are the principalmanagement components.

THE TREATY OF 1906 - Provides for the annualdelivery of 60,000 ac/ft to Mexico. Prompted by theReclamation Act of 1902 and the resulting studyidentifying construction of Elephant Butte Damwhich was authorized in 1905 and completed in 1916.

THE RIO GRANDE COMPACT - Initiated in1923 and agreed upon in 1929, approved by Con-gress in 1939, allocates Rio Grande water betweenthe states of Colorado, New Mexico, and Texas viaa complex set of delivery schedules that relaterunoff volumes to delivery obligations at set riverindex points. In normal years New Mexico mustassure delivery of 60% of the flow passing Otowigage reaches Elephant Butte Reservoir which is thedelivery point for Texas’ allocation. In wet yearsthe percentage is 80%. The Compact also providesrules for accruing and repaying water credits anddebits, water storage restrictions, and operation ofreservoirs. The compact does not affect obligationsto Mexico or to Indian tribes (Shupe and Folk-Williams 1988).

MIDDLE RIO GRANDE CONSERVANCYACT 1923 - Formed the Middle Rio Grande Con-servancy District in 1925 in response to decrease inproductive, irrigated farmland and increasedflooding along the MRG. Channel Aggradation,flooding, and waterlogging of arable lands re-sulted from Rio Grande water infiltrating thegroundwater system of the lower, surroundingfloodplains. This resulted in a dramatic decrease inproductive farmland from 50,000 ha to 16,000 haby 1925 (Nanninga 1982). From 1925 to 1935 theMRGCD constructed, operated, and maintainedfour major diversions dams (Cochiti, Angostura,

1 4

Isleta, and San Acacia), two canal headings, andmany miles of drainage canals, river levees, andmain irrigation canals (fig 8). Initial flood controlstructures were 2.5 m spoil levees that paralleledthe Rio with a mean channel width of 450 m. From1951 to 1977 a system of Kellner jetty fields wasinstalled along the MRG to protect levees and toaid in flood control and channel stabilization.

In recognition of continued flooding and sedi-mentation problems on the MRG, the COE andBOR jointly prepared the “Rio Grande Compre-hensive Plan”. The COE’s portion of the plan pro-vided Jemez Canyon Dam in 1953, Abiquiu Damand Reservoir in 1963, Galisteo Dam in 1970, andCochiti Dam and Reservoir in 1973. The systemconsisting of Abiquiu, Jemez Canyon, Galisteo,and Cochiti dams and the levees along the RioGrande provides flood control and protection forthe MRG valley (Lagasse 1980).

El Vado Reservoir on the Rio Chama was pro-posed by the MRGCD in 1928, providing irrigationwater and flood control to the MRG. Under anagreement with the Department of Interior, ElVado also provided irrigation water to the A Indian

pueblos in the area (Cochiti, Isleta, San Felipt,Santa Ana, and Santo Domingo). The dam wascompleted in 193b5 and rehabilitated in 1958.Operating responsibility was transferred to BOR in1956.

FLOOD CONTROL, DIVERSION PROJECTS,AND PUBLIC LAWS

CABALLO DAM, located 27 km below El-ephant Butte Dam, was authorized in 1933. Thisprovides flood control for El Paso and the JuarezValley. It is managed by both the BOR (conserva-tion operations), and the International Boundaryand Water Commission (IBWC) (flood control).

PLATORO DAM AND RESERVOIR, on theConejos River in Colorado, was authorized in 1940for conservation and flood control and completedin 1951.

FLOOD CONTROL ACT OF 1948 authorizedconstruction of Jemez Canvon Reservoir and thelow-flow convevance channel from San Acacia tcrElephant Butte Reservoir. It also authorized the

Figure 8. Schematic map of an irrigation network on the Middle Rio Grande (Bullard and Wells, 1992).

1 5

BOR to maintain the Channel of the Rio Grandefrom Velarde to Caballo Reservoir to accommodateflows of about 5,000 CFS.

The Low Flow Conveyance Channel is used totransfer water through its 82 km length moreefficiently during periods of low flow, whichminimizes water losses to infiltration and phreato-phytes. The low-flow conveyance channel isnormally operated to convey the entire flow in theRio Grande up to about 2,000 cfs; when flowsexceed about 2,000 cfs, the remainder is carried bythe natural channel. Water is also allowed to flowin the natural channel when the silt load is high.

FLOOD CONTROL ACT OF 1960 - The FloodControl Act of 14 July 1960 (PL48-645) contains thecriteria governing operations of the four MiddleRio Grande Project flood control reservoirs: JemezCanyon, Abiquiu, Cochiti, and Galisteo. Portionsof the operating criteria include:

The reservoirs are operated only for floodcontrol.

Cochiti spring outflow will be at the maxi-mum rate of flow without causing flooding ofleveed protected areas.

Provided there is at least 212,000 ac ft of storageavailable for regulation of summer floods andinflow is less than 1500 cfs, no water will bewithdrawn from storage in Cochiti Reservoir.

Jemez and Galisteo will be managed duringJuly through October to only handle summerfloods.

All Reservoirs will be evacuated by March 31,each year.

When it benefits Colorado or New Mexico inCompact, deliveries of a flow of 10,000 cfs isauthorized through the Albuquerque reach.

No departure from the foregoing schedule isallowed without consent of the Rio GrandeCompact Commission.

In the event of an emergency, the COE mustadvise the Compact Commission in writing,and the foregoing rules of operation may besuspended during the period of emergency.

SAN JUAN-CHAMA TRANSMOUNTAINDIVERSION PROJECT - 1963 - The SJC Projectimports water from the San Juan River basin (in

1 6

the Colorado River basin). This water is not subjectto Rio Grande Compact, and can thus be used forbeneficial use (COE 1989). Annual diversion ofabout 110,000 ac ft is authorized. The importedwater is stored and released at Heron Reservoir.This water is allowed to be used for municipal,irrigation, domestic, and industrial purposes, andto provide recreation and fish and wildlife benefits.

ALBUQUERQUE METROPOLITAN ARROYOFLOOD CONTROL AUTHORITY (AMAFCA) -Following several large, damaging floods east ofthe Rio Grande in urban Albuquerque in 1955,1961, and 1963 AMAFCA was created in 1963 toaddress and alleviate the problems of urban flood-ing from unregulated ephemeral tributaries. Aseries of concrete lined drainage structures wereconstructed from arrovos at the foot of the SandiaMountains and feed into the Rio Grande.

THE CLOSED BASIN PROJECT - The ClosedBasin Project m Colorado was authorized by I’L 92-514 in 1972. The purpose is to help Colorado meetits required deliveries to New Mexico, and to helpall three Rio Grande Compact States meet theirdelivery requirement to Mexico. The closed basinProject‘was justified and funded 100% by thefederal government on the basis of honoring theTreaty of 1906. The project consists of 170 salvagewells that remove groundwater from the uncon-fined aquifer in the Closed Basin and discharge thewater into the Rio Grande. The water wouldnormally be consumed by evapotranspiration.Approximately 60,000 to 140,000 ac ft of water isdelivered to the Rio Grande at rates up to 140 cfswhen fully operational.

CURRENT HYDROLOGIC REGIME AND ITSEFFECTS ON THE RIPARIAN VEGETATION

Present conditions in the Rio Grande includelevees, dams, and channelization. Cochiti Dam hashad a major impact on the river and riparian zonebelow it by reducing peak flows and sediments inthe system (fig. 6b). The timing and duration ofreleases of peak flows may not be suitable forgermination and establishment of native species(Fenner et al. 1985, Szaro 1989). In contrast tounmodified riverine systems (fig. 9), levees haverestricted the lateral movement of the river, andchannelization has occurred along some reaches

(fig. 10). The consequence of all these actions fornative riparian vegetation, once areas have becomevegetated, is a drastic reduction in numbers of sitesand opportunities for further recruitment.

Probably as a result of the construction ofCochiti Dam, the northern reaches (Cochiti andAlbuquerque) of the Middle Rio Grande are nowdegrading. Because sediments are trapped at thedam, released waters have high potential forerosion and the channel is deepening. Vegetationis stabilizing the riverbanks, enhancing the nar-rowing and deepening of the channel. Comparisonof 1935 and 1989 aerial photos indicates that theriverine, or river channel portion of the MRG , hasbeen reduced by 49%(8,920 ha [22,032 ac] in 1935 to4,347 ha [ 10,736 ac] in 1989 (fig. 7). For nativeriparian plant species, there is little or no recruit-ment, except for banks and bars adjacent to the

OXboWmay ikmd ahgh water

old sandsYoung stand

Figure 9. Zones of cottonwood and other riparian speciesestablishment along an unmodified river (Crawfordet al, 1993).

main channel of the river that are exposed afterhigh flows. These areas may be scoured by the nexthigh flows and are often subject to mowing tomaintain the floodwav. This lack of recruitment isa consequence of the presence of existing riparianvegetation and the absence of high magnitudeflows to remove established vegetation and createbarren areas tor coloruzation.

In the southern reaches (Belen and Socorro) ofthe MRG, large amounts of sediment are intro-duced into the system at the confluence of the RioPuerto and Rio Salado (Lagasse 1980). Some areasare without levees, and waters spread out here anddeposit sediments. In these reaches, decreases inpeak flows prevent sediments in the channel frombeing moved downstream. At the southern end ofthe MRG, Elephant Butte Dam has caused the baseelevation to rise upstream enhancing deposition,channel widening, river braiding, and aggradingin some areas. Sediment deposition creates sub-strate for recruitment of native cottonwoods andwillows and introduced salt cedar.

Much of the riparian zone along the MRG isdominated by cottonwood trees, which form asparse to dense canopy cover along the river. Inthe understory, native species include the shrubcoyote willow, seepwillow, false indigo bush, NewMexico olive, and others. Introduced species havebecome increasingly important in numbers, fre-quently becoming dominant species in the under-storv and occasionallv in the canopy. In the north-ern reach, the major introduced species in Russianolive. In the south (below Bernardo), salt cedar isprevalent in the understory, and it also forms largemonotypic stands along the river and adjacentfloodplain. Other introduced species (e.g. Siberianelm, tree-of-heaven, china-berry tree, mulberry, andblack locust) are found in the bosque, mostly alonglevee roads and in other disturbed communities nowdominated by native species. These exotics have thepotential for becoming the primary species therethrough time.

Six structural types of plant communities wererecognized by Hink and Ohmart (1984) (fig.ll),based on the overall height of the vegetation andthe amount of vegetation in the understory orlower layers. Type I had vegetation in all layers,with trees 15-18 m ((50-60 ft) high. Type I areaswere mostly mixed to mature age class standsdominated by cottonwood/coyote willow, cotton-

1 7

Riverside Drain

Explanation

- Jetty jack line

- Wire rope

GroinChannel clearing

Trees and brush

Figure 10. Channel stabilization works on the Middle Rio Grande (after Bullard and Wells, 1992).

wood/Russian olive, and cottonwood/juniper.Type II areas consisted of mature trees from 15 to18 m (50-6- ft) with a sparse understory. Intermedi-ate age stands of cottonwood trees with a denseunderstory were classified as Type II, while simi-larly aged trees with open understory were calledType IV. Type V was characterized by densevegetation up to about 4.6 m (15 ft) often withdense grasses and annuals. Type VI had low, oftensparse vegetation, typical of sandbars with cotton-wood, willow, and other seedlings. This type alsoincluded sparsely vegetated drains.

Hink and Ohmart (19841, described three cotton-wood-dominated community types based on theoverstory species and on the type and abundanceof the understory species. The cottonwood/coyotewillow community, cottonwood/Russian olive,dnd cottonwood/juniper found in the northernreach. New Mexico olive, false indigo bush andother species were also found.

Other plant communities also occurred in thestudy area (Hink and Ohmart 1984). Russian olive

occurred along the river channel in narrow, 15-60m (50-200 ft wide bands. Cattail marshes, domi-nated by cattails with some bulrush and sedge, arefound in areas that are inundated or have a highwater table. Wet meadows with saltgrass andsedges were also designated as marsh communi-ties. In the southern reach, salt cedar was theprimary component of the plant communityalmost to the complete exclusion of other species.

Hink and Ohmart (1984) also delineated sand-bars in and adjacent to the river, and the riverchannel. Most of the sandbars were bare, but somehad developed vegetation consisting of grasses,forbs, cottonwood and willow seedlings, and otherspecies. Many of these bars were scoured duringeach year’s high fiows. If not removed by scouring,vegetation in these locations is periodically mowedby the BOR to keep the floodway clear.

While the structure and diversity of native plantcommunities appear to be significant to the diver-sity of species in animal communities, introducedplant species that have become naturalized in the

1 8

19

region also provide shelter and sometimes food.The fruits of the Russian olive, a species which isprominent in the community types in the northernreach of the MRG, appear to be a significant part ofthe diet for some resident, migrant, and breedingbird species. Salt cedar found throughout the studyarea but particularly abundant in the southernportion, provides cover for birds and mammalsand habitat for many insect species (Hink andOhmart 1984).

CONCLUSION

The MRG has been the center of considerableactivity by man for over 10,000 years. Until onlyrecently man’s activities have not had a significantimpact on the character of the riparian area adja-cent to the Rio Grande in this vicinity. With adventof irrigation, control structures such as levees andJetty Jacks, water diversions and control structuressuch as Cochiti Dam, the hydrograph and subse-quent river morphology have been dramaticallyaltered. This alteration in flow regimes and chan-nel configuration continues to have ramificationsand effects upon the native flora and fauna of theMRG.

ACKNOWLEDGMENTS

Much of what is contained in this paper comesdirectly from primarily two documents which Ihighly recommend to the reader. The first “Hy-drology of the Middle Rio Grande from Velarde toElephant Butte Reservoir, New Mexico” by Bullardand Wells 1992. The second indispensable docu-ment is the Middle Rio Grande Bosque BiologicalManagement Plan, Crawford et al. 1993. TheFigures have been taken from the Bosque Biologi-cal Management Plan, Crawford et al. 1993.

LITERATURE CITED

Allen, C., B. Hanson, and C. Mullins (eds.). 1993.Cochiti Reservoir reregulation interagencybiological report. Report submitted to RioGrande Joint Initiatives Committee, NationalPark Service, Bandolier National Monument,Los Alamos, New Mexico.

Biella, T.V., and R.C. Chapman teds.) 1977. Archeo-logical investigations in Cochiti Reservoir, NewMexico. Volume 1: a survey of regional variabil-ity. Report submitted to national Pard Survey,Santa Fe, for U.S. Armv Corps of Engineers,Albuquerque.

Brandvold, D.K., C. J. Popp. T.R. Lynch, and L.A.Brandvold. 1984. Heavy metals and pesticides inwater, sediments, and biota in the Middle RioGrande Valley. Pages 14-23 in W.J. Stone, com-piler. Selected papers on water quality and pollu-tion in New Mexico. New Mexico Bureau of Minesand Mineral Resources, Hydrological Report 7.

Brown, D.E. and C.H. Lowe. 1980. Biotic communi-ties of the Southwest. U.S. Forest Service, RockyMountain Forest and Range Experiment Station,General Technical Report RM-78.

Bullard, T.F. and S.G. Wells. 1992 Hydrology of theMiddle Rio Grande from Velarde to ElephantButte Reservoir, New Mexico. U.S. Fish andWildlife Service, Resource Publication 179.

Crawford, C.S., A.C. Cully, R. Leutheuser, M.S.Sifuentes, L.H. White, J.P. Wilber. 1993 MiddleRio Grande Ecosystem: Bosque Biological Man-agement Plan.

Cully, A.C. 1977. Paleoclimatic variability in theNorth-Middle Rio Grande, New Mexico. Pp. 97-101 in J.V. Biella and R.C. Chapman (eds.)Archeological investigations in Cochiti Reser-voir, New . Volume 1; a survey of regionalvariability,. Report submitted to National ParkSurvey, Santa Fe, for U.S. Army Corps of Engi-neers, Albuquerque.

Fenner, I’., W.W. Brady, and D.P. Patton. 1985.Effects of regulated water flows on regenerationof Fremont Cottonwood. Journal of RangeManagement 381.35138.

Hawley, J.W., compiler. 1978. Guidebook to RioGrande Rift in New Mexico and Colorado. NewMexico Bureau of Mines and Mineral Resources,Circular 163.241 pp.

Heath, D.L. 1983. Flood and recharge relationshipsof the Lower Rio Puerto, New Mexico. Pp. 329-337 in New Mexico Geological society Guide-book, 34 Field conference, Socorro Region II.Socorro, New Mexico.

Hink, V.C. and R.D. Ohmart. 1984. Middle RioGrande biological survey. Report submitted toU.S. Army corps of Engineers, Albuquerque,New Mexico.

20

Hunt, C.B. 1974. Natural region of the UntiedStates and Canada. W.H. Freeman & Co., SanFrancisco, Calif. 725. pp.

Kelson, K.I. 1986. Long-term tributary adjustmentsto base-level lowering, northern Rio Grande Rift,New Mexico. M.S. thesis, Universitv of NewMexico,

Albuquerque. 210 pp.Lagasse, P.F. 1980. An assessment of the response

of the Rio Grande to dam construction-Cochitito Isleta Reach. Technical Report submitted toU.S. Army Corps of Engineers, Albuquerque,New Mexico.

Leopold, L.B., M.G. Woldman, and J.P. Miller.1964. Fluvial processes in geomorphology. W.H.Freeman and Co., San Francisco.

Lozinski, R.P., J.Hawley, and D. Lowe. 1991.Geologic overview and Pliocene-Quaternaryhistory of the Albuquerque Basin, New Mexico.Pp 157-162 in B. Julian and J. Zidek (eds.) Fieldguide to geologic excursions in New Mexico andadjacent areas of Texas and Colorado Bulletin137. New Mexico Bureau of Mines and MineralResources, Socorro.

Maker, H.J., J.M. Downs, and J.U. Anderson. 1972.Soil associations and land classification forirrigation, Sierra County. New Mexico StateUniversity Agricultural Experiment ResearchStation Report 233.

Nanninga, R.S. 1982. Middle Rio Grande valleyprior to the advent of the conservancy district.Pages 99-101 in J.A. Grambling and S.G. Wells,editors. Albuquerque Country II. New MexicoGeological Society Guidebook 33.

Ouchi, S. 1983. Effects of uplift on the Rio Grandeover the Socorro magma body, New Mexico.Pages 54-56 in C.E. F. Chapin, editor. SocorroRegion II. New Mexico Geological SocietyGuidebook 34.

Ouchi, S. 1985. Response of alluvial rivers to slowactive tectonic movement. Geological Society ofAmerica Bulletin 96:504-515.

Popp, C.J., D.K. Brandvold, T.R. Lynch, and L.A.Brandvold. 1983. An evaluation of sediments inthe Middle Rio Grande, Elephant Butte Reser-voir and Caballo Reservoir, as potential sources

for toxic materials. New Mexico Water Re-sources Institute WRRl Report 161.97 pp.

Richards, K. 1982. Rivers: Form and process inalluvial channels. Methuen, London. 358 pp.

Schumm, S.A. 1977. The Fluvial System. JohnWilev and Sons, New York.

Schumm, S.A. 1986. Alluvial river response toactive tectonics. Pages 80-94 in Active tectonics;studies in geophysics. National Research Council.National Academy Press, Washington, D. 266 pp.

Shupe, S.J., and J. Folk-Williams. 1988 The UpperRio Grande: a guide to decision-making. West-ern Network, Santa Fe, New Mexico.

Szaro, R.C. 1989. Riparian and forest and scrublandcommunity tvpes of Arizona and New Mexico.Desert Plants 9:70-135.

Tuan, Y-F., C.E. Everard, and J.G. Widdison. 1973.The climate of New Mexico. State PlanningOffice, Santa Fe, New Mexico.

U.S. Army Corps of Engineers. 1979. Albuquerquegreater urban area, urban studies program,water supply, Appendix III. U.S. Armv Corps ofEngineers, Albuquerque District.

U.S. Armv Corps of Engineers. 1989. Reevaluationof the Rio Grande operating plan. AlbuquerqueDistrict Ottice.

U.S. Bureau of Reclamation. 1981. Project data. U.S.Government Printing Office, Washington, D.C.1463 pp.

Wells, S.G., K.I. Kellson, and C.M. Menges. 1987.Quaternary evolution of fluvial svstems in thenorthern Rio Grande Rift, New Mexico andColorado: implications for entrenchment andintegration of drainage systems. Pages 55-69 inC. Menges, editor. Quaternary tectonics, land-form evolution, soil chronologies and glacialdeposits-northern Rio Grande Rift of NewMexico. Friends of the Pleistocene-Rocky Moun-tain Cell Field Trip Guidebook.

Wozniak, F.E. 1987. Irrigation in the Rio GrandeValley, New Mexico: a study of the developmentof irrigation systems before 1945. The NewMexico Historic Preservation Division, Santa Fe,and U.S. Bureau of Reclamation, SouthwestRegional Office, Amarillo, Tex. Contract BOR-87-l. 191 pp.

21