10 S O I L S U R V E Y H O R I Z O N S

Vertisols are a group of soils distinguished by high amounts of

shrink–swell clays; minimal horizon differentiation due to pedo-

turbation; pronounced changes in volume with changes in water

content resulting in deep, wide cracks in the dry seasons; very

plastic and sticky soil consistency when wet; and unique subsur-

face structure features called slickensides (Mermut et al., 1996).

The shrink–swell process, caused by wetting and drying of the

expansive clay minerals associated with Vertisols can generate

swelling pressures on the order of 1 to 5 kg cm−2 (Ahmad, 1983).

Associated with the swelling, surface vertical movement (heave)

has been observed on the order of 10 to 20 cm in field studies in

Texas (Miller and Bragg, 2007). However, extreme heaves of 45

to 90 cm have also been reported in the literature, based on lab-

oratory analysis and case studies of building foundation failures

(Fredlund, 1996).

Vertisol Morphology, Classification, and Seasonal Cracking Patterns in the Texas Gulf Coast Prairie

Wesley L. Miller, Andrea Sz. Kishné, and Cristine L. S. Morgan

Vertisols have unique morphology and the capacity to change volume with changes in water content; there-fore, they are well known for their spatial and temporal variability. Our objectives were to address the spatial variability of Vertisol surface and subsurface features, as well as the spatial and temporal variability of Verti-sol cracking. First, we discuss surface and subsurface features of Vertisols based on the literature and more than three decades of interagency soil investigations in the Texas Gulf Coast Prairie. Second, we report on a 10-yr field study of location, length, width, depth, and duration of seasonal crack openings in a 100-m2 area of Laewest clay (fine, smectitic, hyperthermic, Typic Hap-ludert) with gilgai in a native grassland south of Victoria, TX. Results of the crack monitoring showed that cracks always started on, and crack density was greater on, microhighs or the upper part of microslopes. Under prolonged drying periods, cracks formed on microlows and were, on average, deeper in microlows. Cracks generally occurred in the same location within a year. Between years, crack locations shifted slightly, although they were clustered in the same general areas during the 10 yr of study. Vertisol cracking was affected not only by seasonal fluctuation of precipitation, but also by multiple-year precipitation cycles. Analysis of precipi-tation over several decades indicated that the Ustert/Udert classification criteria based on cracking in normal precipitation years was never met in this region of Texas because there were never eight normal months within a normal year. Normal yearly precipitation standards and months on a regional basis would provide better precipi-tation norms for soil classification.

abstract

Fig. 1. Normal gilgai in Laewest clay, Calhoun County, Texas showing evenly mixed areas of microhighs (light) and microlows (dark) in a recently burned native prairie. The height difference is 25 to 30 cm between the tops of micro-highs and bottoms of microlows.

W.L. Miller, USDA-NRCS, retired, P.O. Box 2252 Victoria, TX 77902 (wmiller@tisd.net); A.Sz. Kishné (akishne@ag.tamu.edu), and C.L.S. Morgan (CMorgan@ag.tamu.edu), Dep. of Soil and Crop Sciences, Texas AgriLife Research, Texas A&M Univ., Col-lege Station, TX 77843-2474. Published in Soil Surv. Horiz. 51:10–16 (2010).

11S p r i n g 2 0 1 0

Area, iron chlorosis and drought stress in corn and grain sorghum were

observed on truncated microhighs in Vertisol gilgai landscapes leveled

by cultivation (Wilding et al., 1990).

Field soil scientists have long recognized the variability of gilgai ele-

ments, lateral periodicity, and subsurface irregularities of Vertisols during

soil survey activities (Wilding et al., 1990). Field observations were com-

monly made on the microlow, or the microlow and microhigh. Less

attention was paid to the microslope position. Soil characterization pits

were usually small and centered on a microlow or a microlow and associ-

ated microhigh. Therefore, the knowledge base for short range variability

in Vertisols is limited. Newman (1986) was one of the first in Texas to

construct block diagrams of Vertisol surface and subsurface morphol-

ogy based on soil characterization studies in Texas, but did not include

the chimney and puff. Gilgai, cracking patterns, variability of subsurface

horizonation, amplitude of horizon waviness, microclimate, and drying–

wetting cycles were illustrated and discussed. However, further field

studies were needed to understand better the mechanisms of cracking.

In the late 1980s there was renewed interest in Vertisol characteriza-

tion due to several International Soil Correlation meetings to modify the

USDA Soil Taxonomy (Soil Survey Staff, 1975) system. The goal was to

improve Soil Taxonomy as a vehicle to transfer agrotechnology (Kimble,

1990). As a part of this international study, the USDA sponsored several

interagency and university cooperative soil characterization studies of

Vertisols in the Texas Gulf Coast Prairie. Additional Vertisol characteriza-

tion studies were also conducted in the 1990s in the Texas Gulf Coast

Prairie as part of the investigation of Paleovertisols (Driese et al., 2005).

One of the first detailed Vertisol characterization studies in the Texas

Gulf Coast Prairie was conducted in Victoria County in 1989 (Wild-

ing et al., 1990). A critical step in a Vertisol microvariability study was

making sure that the site was sampled adequately and represented the

field conditions of a given soil. A study site in native grassland (28° 44¢ 17² N, 96° 50¢ 30² W) was selected in an area of Lake Charles clay (Soil

Survey Staff, 1975) that was reclassified later as Laewest (Soil Survey

Staff, 1999). Ground truth was established by transects at 0.6-m lateral

intervals across at least three consecutive microhighs and microlows

in a representative area. Once a specific site location was selected and

marked, a pit was dug across three microhighs and two microlows. The

The physical movement of Vertisols commonly results in formation of

surface mound and depression microrelief features called gilgai. Gilgai

is an Australian aborigine term meaning “little water hole” (Paton, 1974;

Ahmad, 1996). Normal (or circular) gilgai is a term used to describe gilgai

with round, sometimes interconnected, depressions and mounds devel-

oped in an irregular pattern on nearly level landscapes and is a feature

found most commonly in the Texas Gulf Coast Prairie (Fig. 1). Melon hole

gilgai, a term found in Australian literature, refers to gilgai on nearly level

landscapes; however, the depressions (microlows) are very distinct, iso-

lated areas surrounded by mounds (microhighs). Linear gilgai occurs on

sloping land where microhighs and microlows are orientated parallel to

the slope gradient (up and down slope).

Gilgai surface mound and depression features are further subdivided

into microhigh, microslope, and microlow (Fig. 2; Miller et al., 2005; USDA-

NRCS, 2005; Miller and Bragg, 2007). The microhigh is convex and occurs

along the top part of the mound. The microlow is concave and occurs

along the bottom part of the depression and may pond water for brief peri-

ods after precipitation events. The gently sloping area between the lower

part of the microhigh and the upper part of the microlow is the microslope.

Distances between microhighs and adjacent microlows are commonly 2

to 5 m. Elevation differences measured between the top of microhighs and

the bottom of microlows range from 15 to 50 cm in the Texas Gulf Coast

Prairie (W.L. Miller, unpublished data).

Subsurface features associated with microlow, microslope, and

microhigh formations are bowl, intermediate, and chimney (Fig. 2). The

bowl position is a cup- or trough- shaped subsurface feature under

the microlow and is bounded along its base by concave slickensides.

The intermediate position is a transition zone between the chimney and

bowl, and is under the microslope surface feature. The chimney position

is composed of subsurface material that appears to upwell and reach

close to the surface under the microhigh or upper part of the microslope

surface features. In some Vertisol landscapes, the chimney material

is exposed at the surface and is termed a puff (Paton, 1974). Chimney

material is usually lighter in color, more alkaline, and has less organic

matter than the associated bowl material. In many Vertisols of the Texas

Gulf Coast Prairie, chimney materials are also calcareous and contain

pedogenic carbonates in the form of calcium carbonate nodules.

Pedogenic carbonates in Vertisols of the Texas Gulf Coast Prairie

develop from leaching of detrital carbonate in the fine earth fraction of

the sediments to just below the effective leaching depth (Stiles, 2001).

Carbon isotope values were determined within individual pedogenic

carbonate nodule samples from the chimney and bowl to document for-

mation and relative age of the nodules (Nordt et al., 1998; Stiles, 2001;

Miller et al., 2007). The results of the carbon isotope study indicate that

there are two types of pedogenic carbonates in the chimney of these

Vertisols. The first type is the smaller carbonate nodule that formed

pedogenically in place. The second type is the larger carbonate nodule

that initially started formation deep in the bowl, and with time moved

upward along major slickensides toward the chimney and the surface.

Differences in plant communities in native grasslands on Vertisol

microhighs and microlows are documented (Thompson and Beckmann,

1982; Wilding et al., 1990). Generally, more xerophytic plant spe-

cies are found on microhighs and mesophytic or hydrophytic species

in microlows. Vertisols in forested areas tend to have trees on micro-

highs and mixed forbs/grasses on microlows (Russell et al., 1967; Miller

and Bragg, 2007). In the Texas Gulf Coast Prairie Major Land Resource

Fig. 2. Schematic block diagram showing microvariability of a Vertisol with gilgai in the Texas Gulf Coast Prairie. Surface features are microhigh (A), microslope (B), and microlow (C). Subsurface features bounded by major slickensides are chimney (D), intermediate (E), and bowl (F). A surface exposure of the chimney is called a puff (G). Reprinted with permission from USDA-NRCS (Miller and Bragg, 2007).

12 S O I L S U R V E Y H O R I Z O N S

ensides (Williams et al., 1996). Several of the larger slickensides were

traced from a near vertical position close to the surface along the chim-

ney, continued downward, and were nearly level and slightly concave

with depth underneath the bowl, and then up again to a near vertical

position along the adjacent chimney.

Between the bowl and chimney was an extensive zone of calcar-

eous gray clay, darker in color than the chimney, but lighter than the

bowl (Wilding et al., 1990). It contained common carbonates as soft

segregations, hard concretions encased in soft calcareous rinds and dis-

seminated forms. These calcareous materials were commonly banded at

angles corresponding to bordering slickensides and oriented toward the

chimney and microhigh.

The cross-sectional area of the pit face was composed of 19%

microhighs, 52% microslopes, and 29% microlows (Wilding et al.,

1990). In a 100-m2 area adjacent to the pit, microlows covered 24%

of the area (W.L. Miller, unpublished data). Of the seven individual

pedons described and sampled from the pit, four classified as Typic

Hapluderts (57%) and three classified as Chromic Hapluderts (43%).

Using current USDA-NRCS standards for classifying soils where hori-

zons are intermittent or cyclic and occur in linear intervals of 2 to 7

m (Soil Survey Staff, 1999), 66% of the pit face was Typic Hapluderts

and 34% was Chromic Hapluderts (W.L. Miller, unpublished data).

During the Victoria County soil survey field work, a similar analysis

of soil cores investigated at every 0.6 m across three microhighs and

two microlows of a Laewest clay showed that 75% of the individual

pedons classified as Typic Hapluderts and 25% as Chromic Haplud-

erts (Wilding et al., 1990). Considering the linear proportions of the

total transect distances, 65% of the linear distance was Typic Hap-

luderts and 35% was Chromic Hapluderts. Comparing procedures for

determining the composition percentage of Chromic and Typic Hap-

luderts based on the number of individual pedons from either field

transects or pit faces and analysis of the linear distance covered by

each soil, it seems that classification based on linear distance is the

more accurate method.

pit was ~240 cm deep and 750 cm long. Over a 4-d period, each distinct

soil zone or polyhedron was outlined using nails and white string (Fig.

3) and then diagramed to scale (Fig. 4). The soil morphology of the pit

face was described in detail according to Soil Survey Staff (1951), includ-

ing depth, orientation, and slope of slickensides. Seven pedons were

selected for detailed sampling of each horizon to a depth of 240 cm. The

seven pedons included two microhighs (chimney), three microslopes

(intermediate), and two microlows (bowl).

The soil exposed by the pit had pronounced morphological differ-

ences between chimney (microhigh), intermediate (microslope), and bowl

(microlow). The chimneys were narrow (30–70 cm wide) zones of gray-

ish, calcareous clays extending from the lower Bk horizon to the surface

(Wilding et al., 1990). These teepee-shaped structures appeared to have

been “pushed or squeezed” up from the lower Bw and Bk horizons at

the bottom of the bowls along slickenside planes at steep angles. The

chimneys were not always located in the center of microhighs, but were

slightly offset. Two chimneys (Pedons 1 and 5) had near-vertical, thin

zones of darker soil materials. These did not appear to be due to crack

filling of soil surface material, but rather seemed to be darker soil materi-

als of adjacent bowls thrust along slickenside planes into the chimneys.

The bowls were generally black clays with little or no carbonates,

were 2 to 5 m across and 1 to 2 m deep, and were outlined by large slick-

Fig. 3. Laewest clay profile in Victoria County, Texas (28° 44¢ 17² N, 96° 50¢ 30² W). Each distinct area (polyhe-dron) has been outlined with nails and string based on soil matrix color, slickensides, size, and arrangement of structure, kind and amounts of concre-tions, nodules, soft masses, and other features that distinguish one polyhe-dron from another.

Fig. 4. Sketch of cross-sectional profile showing the distribution of predominate soil matrix colors in Laewest clay.

13S p r i n g 2 0 1 0

The objectives of the crack study reported here were as follows: (i) to

measure and accurately diagram crack width, length, depth, and location

in a 100-m2 area; (ii) to determine where cracks occur in gilgai microrelief;

(iii) to determine if cracks occur in the same or similar locations from sea-

sonal and yearly dry–wet–dry periods; (iv) to take periodic soil moisture

samples to depths of 100 cm when cracks were present; and (v) record

depths and durations of ponding in microlows.

Materials and MethodsThe study of monitoring cracks was conducted in a representative

Texas Gulf Coast Prairie Vertisol in Victoria County, Texas (28° 39¢ 46² N, 96 o 46¢ 20² W) for a 10-yr period (Fig. 5). A 10- by 10-m square plot

was selected for the crack study in native grassland. The soil was first

classified as Lake Charles clay (fine, montmorillonic, thermic Typic Pel-

ludert) according to Soil Survey Staff (1975) and was later changed to

Laewest clay (fine, smectitic, hyperthermic Typic Hapludert) according

to Soil Survey Staff (1999). The crack study site was adjacent to the site

of the Victoria County Wet Soil Field Study as part of the USDA Interna-

tional Committee on Aquic Soils Wet Soil Study in Louisiana and Texas

(Kimble, 1990; Soil Survey Staff, 1990; Griffin et al., 1992). Figure 6 shows

the site used for field observations and crack measurements. The sites

were visited every 2 to 3 wk, with additional field visits after major precip-

Several Vertisol field and characterization studies in the past 30 yr

showed the need for more comprehensive documentation and larger

soil characterization pits to adequately describe and sample short-range

variability in Vertisols. The studies in Texas, on Pleistocene (Wilding et

al., 1990; Williams et al., 1996) and Holocene age landscapes (Miller and

Bragg, 2007) documented distinct morphological, chemical, and hydro-

logical differences between microhigh, microslope, and microlow soils.

Soil scientists who have participated in many detailed Vertisol field stud-

ies at the national and international level agree that if there is evidence of

surface gilgai, then there will be at least subtle associated morphological,

chemical, and structural differences between soils in the microtopogra-

pic positions (personal communications, Alan Newman, 1986; Warren

Lynn, 1990; DeWayne Williams, 1990; Larry Wilding, 1990).

Vertisol studies in California (Williams et al., 1996) and in a few areas

on uplands in the Texas Gulf Coast south of Corpus Christi, TX (W.L.

Miller, unpublished data) suggest that not all Vertisols exhibit such dis-

tinct spatial variability. However, data gathered from studies in Vertisols

with gilgai in Russia, central Texas, and the Texas Gulf Coast Prai-

rie (Lake Charles and Laewest soils) all show a strong need to identify

at least two taxonomic units to adequately identify, classify, and make

interpretations of these particular Vertisols exhibiting gilgai (Wilding et al.,

1990; Williams et al., 1996; Nordt and Driese, 2009).

Monitoring Cracking of a Vertisol with GilgaiSoil Survey Staff (1999) requires information on the duration of open-

ing of cracks and crack width at a specific depth for soil classification,

but does not specify their detailed description. Current USDA-NRCS

criteria for Vertisol cracks indicate that “a crack is a separation between

gross polyhedrons,” and that “a crack is regarded as open if it con-

trols the infiltration and percolation of water in a dry, clayey soil” (Soil

Survey Staff, 1999). At the suborder level of Vertisols, soils that “…have

cracks in normal years (when not being irrigated) that are 5 mm or more

wide, through a thickness of 25 cm or more within 50 cm of the mineral

soil surface, for 90 or more cumulative days per year” are classified as

Usterts. Vertisols that fail these criteria are classified as Uderts.

Vertisol cracking patterns are not merely the response of the soil mate-

rial to dehydration, but rather the differential moisture status of the soil at

different depths plays a role in determining the size, shape, and frequency

of the cracks (Wilding et al., 1990). Soil chemistry may also play a part in

cracking patterns, as some Vertisols with Ca-saturated clays form wide

cracks with rather low frequency, while other Vertisols with Na-saturated

clays have a higher intensity of fine cracks (Mermut et al., 1996).

Long-term studies of cracking patterns in Vertisol microhighs, micro-

slopes, or microlows are generally lacking. Thompson and Beckman

(1982) observed a network of cracks of Black Earths developed domi-

nantly in the microlows during a prolonged dry season in Australia.

Cracks were less on the microhighs and were generally finer and shal-

lower than the cracks in the microlows.

Systematic measurement of crack width, depth, and location in the

gilgai microtopography for an area greater than a pedon is often the

technique used, but there is not a standard method. In many cases, char-

acterization of cracks is done as part of a profile description and wider and

deeper cracks are emphasized. Wilding et al. (1990) suggested that crack-

ing patterns, crack closure hysteresis, and cracking depths as a function

of seasonal moisture changes in Vertisols with gilgai microtopography

have not been adequately studied and require further attention.

Fig. 5. Location of study site for the 10-yr monitoring of Vertisol cracking.

Fig. 6. Monitoring site of cracking in Laewest clay with normal gilgai.

14 S O I L S U R V E Y H O R I Z O N S

tation standards and months on a regional basis would provide better

precipitation norms for soil classification.

During the two years with normal precipitation, cumulative ponding

(i.e., the number of times ponding was observed based on 2-wk observa-

tions) in a representative microlow was 10% of the time. Ponded events

occurred in normal precipitation months 20% of the time and in above-

normal precipitation months 80% of the time. Average ponded depth

measured in the lowest part of the microlow was 6 cm and ranged from 2

to 9 cm. The duration of the ponding events included three “less-than-2-

continuous-week” events and one “2-continuous-week” event.

Spatial Variability of CrackingVertisol cracking was analyzed spatially according to gilgai cate-

gories. Gilgai was stratified based on a digital elevation model of the

100-m2 study site (Fig. 6). We found that microhighs were 38%, micro-

slopes 43%, and microlows 19% of the area (Kishné et al., 2009).

Calcareous puffs (surface outcrop of light-colored calcareous Bkss

clay chimney) comprised 5% of the surface area and were located on

microhighs and the upper part of microslopes. Puff areas were oval

itation events from 1989 to 1998. Surface width, length, and location of

cracks in the study plot were measured using a 1-m2 square frame with

0.1-m cell size. Cracks were measured as they occurred on 42 dates. On

each date, crack locations and dimensions were diagramed on engineer

graphing paper on a 0.0254 to 1-m scale. Crack widths were grouped

into categories with limits of 0.5, 1, 2, 5, and 7 cm, and drawn to scale

on each crack diagram using a color code for each of the different crack

categories. Periodically during the study, soil water was measured gravi-

metrically from samples at 10-, 25-, 50-, 75-, and 100-cm depths in two

to three replicates on microhighs and microlows in areas immediately

adjacent to the crack study plot.

The vertical crack depth relative to the surface was measured with

a 200-cm-long pointed steel tape (6.35 mm wide, 0.79 mm thick) that

had a 5-mm-wide tip attached to one end (known by the field name

“crackometer”). The pointed end of the steel tape was used to mea-

sure maximum crack depth. Additionally, the 5-mm-wide tip was used

to measure crack depth that was at least 5 mm wide on 18 dates (Soil

Survey Staff, 1999). Locations for crack depth measurements were

selected to characterize the deepest cracks.

The crack diagrams were scanned and digitized in ArcMap v.9.1.

A digital elevation model was created based on elevation data mea-

sured using standard surveying techniques. The data, including digitized

cracks, measured soil water content, elevation, and field notes, are pub-

lished at http://soilcrop.tamu.edu/research/pedology/crack.html (verified

29 Dec. 2009).

The long-term mean annual and monthly precipitation was deter-

mined based on 30-yr averages for 1971 through 2000. For the region,

the yearly and monthly mean precipitations and their normal thresh-

olds values were found in the WETS (Wetlands Determination) table of

the National Water and Climate Center (USDA-NRCS, 1995, 2002). The

normal precipitation range was defined as the 30th and 70th percentile

of a two-parameter distribution fitted to the 30-yr precipitation data.

Results and Discussion

Precipitation AnalysisThe current standard in Soil Taxonomy defines a normal precipita-

tion year according to the long-term mean annual precipitation with

eight normal months out of the 12 months (Soil Survey Staff, 1999). The

30-yr mean annual precipitation for the region was 1019 mm for the 10

yr of observation. Only two years, 1990 and 1994, were within normal

yearly precipitation limits (Fig. 7). In 1990, there were only four normal

precipitation months, and in 1994, there were only five months (Table 1).

Therefore, neither of the two years with normal annual precipitation met

the Soil Taxonomy criteria of a normal precipitation year also having eight

normal precipitation months.

A preliminary analysis of Victoria County National Weather Service

normal precipitation years during the past 30 yr before the end of the

study in 1996 indicated that the average number of normal precipitation

months was 4, with a range of to 2 to 7 mo in normal precipitation years

(18 of the 30 yr). The current Soil Taxonomy criteria of a normal precipi-

tation year including eight normal precipitation months is too restrictive

and impossible to meet in the central Texas Gulf Coast Prairie MLRA.

The Soil Taxonomy criteria of normal precipitation year with eight normal

precipitation months should be re-evaluated, and normal yearly precipi-

Fig. 7. Annual precipitation of Victoria County for 1989 through 1998, with thresholds for normal precipitation years shown.

Table 1. Monthly precipitation is listed for the measurement site in normal precipitation years, 1990 and 1994. Thresholds of normal monthly rainfall are also listed. The maximum values of crack area density measured at the study site in these years were 0.0041 m2 m−2 (21 Sept. 1990) and 0.0137 m2 m−2 (1 Aug. 1994).

Months Thresholds Total precipitation

Lower Upper 1990 1994————————————————————— mm ————————————————————

January 27 76 44 n† 36 n

February 23 63 52 n 21 bn

March 26 70 76 an 121 an

April 25 91 92 an 52 n

May 50 157 30 bn 113 n

June 54 155 21 bn 139 n

July 23 88 345 an 54 n

August 51 93 37 bn 99 an

September 74 154 91 n 73 bn

October 50 132 40 bn 266 an

November 27 82 59 n 8 bn

December 34 77 21 bn 128 an

† n, an, bn represent normal, above-normal, and below-normal, respectively.

15S p r i n g 2 0 1 0

changes in soil water content at the time of cracking and antecedent

soil water content before cracking. For example, the greatest crack den-

sity was found to be associated with above-normal precipitation and soil

water recharge conditions followed by a period of below-normal precipi-

tation and drying soil water conditions. In contrast, relatively small crack

density was associated with a 2-yr period of below-normal precipitation

in 1988–1989.

It is an open question how cracking in a normal precipitation year

might be affected by previous multiyear rainfall history. Figure 7 shows that

before the normal year 1990, there was below-normal precipitation. Con-

versely, before 1994, there were 3 yr of above-normal precipitation. The

maximum measured crack density for 1994 was about three times greater

than it was in 1990, 0.014 and 0.004 m2 m−2, respectively. Thus, different

recharged water conditions of bulk soil matrix in prior years may cause

significant difference in cracking in normal years. To improve understand-

ing of this mechanism and to standardize conditions for soil classification,

we suggest further investigations regarding the effect of antecedent soil

water content before cracking to fulfill the required criteria for Udic/Ustic

soil moisture regimes in Vertisols in different climatic regions.

Considering the occurrence of cracks in the same or similar loca-

tions, we found that cracks opened and closed in the same general

location during a season of drying–wetting cycles. However, over a

period of several years, some cracks shifted location slightly, or pre-

vious observed cracks remained closed and new ones opened. Even

though some specific crack locations shifted during the study, cracks

still clustered in the same microtopographic areas within the 100-m2

plot (Fig. 8). Above-normal precipitation events, prolonged dry periods,

and disturbances such as fire ant mounds in a crack may have contrib-

uted to changes in crack locations.

Summary and ConclusionsWe have reviewed the spatial and temporal dynamic properties of

Vertisols developed in the region of Texas Gulf Coast Prairie. In the

first part of the article, observations collected on Vertisol morphology

and classification over more than three decades were summarized and

to narrow-oblong in shape, and ranged in size from 20 by 50 cm to 40

by 400 cm. Elevation differences between the top of microhighs to the

bottom of microlows ranged from 15 to 25 cm.

Soil cracking always started on microhighs or on upper part of micro-

slopes (Fig. 8). Cracks that were at least 5 mm wide through a thickness

of ~25 cm and within 50 cm of the mineral soil surface were observed on

the microhighs 17 times, microslopes 15 times, and microlows 7 times out

of the 18 measurement dates. Average surface crack widths were simi-

lar on microhighs and microslopes (3.8 and 3.9 cm, respectively), and

slightly greater in microlows (4.2 cm). Average measured depths of cracks

that were at least 5 mm wide were 46 cm on microhighs, 52 cm on micro-

slopes, and 64 cm on microlows. Maximum depths of cracks that were

at least 5 mm wide were similar for all microtopography locations (83–85

cm). The deepest crack segment was 140 cm deep, 7 cm wide, and 93 cm

long and was situated across a microhigh and a microslope. A width of 7

cm was the maximum surface crack width measured during the study. The

longest surface crack segment was 145 cm long and 4 cm wide.

Temporal Variability of CrackingSeasonal dry and wet periods produced short-term cycles of soil sur-

face crack area density (crack density), and is reported in more detail in

Kishné et al. (2010). Crack formation occurred most extensively in the

summer and early fall months. Considering the relationship of soil water

content and the extent of cracks on the soil surface, current soil water at

10 to 25 cm showed the strongest correlation to soil crack density. The

greatest crack density was measured when gravimetric soil water ranged

15 to 20%, and the smallest crack density was measured when gravi-

metric soil water was 21 to 38% in microhighs and microlows.

In addition to the short-term seasonal variation in rainfall, multiyear

oscillations also seemed related to crack density through variation in the

antecedent soil water content just before the start of cracking events that

showed positive correlation with crack density (Kishné et al., 2010). Thus,

crack density was found to be dependent on both current and anteced-

ent soil water content. In other words, a dual fluctuation was observed

in cracking such that crack density fluctuated according to seasonal

Fig. 8. Superimposed crack locations illustrate the clustering and shifting of the crack openings in Laewest clay over the years. The microtopography categories and the location of chimney outcrops (puffs) are shown for reference.

16 S O I L S U R V E Y H O R I Z O N S

Kimble (ed.) Proceedings of eighth international soil correlation meeting (VIII ISCOM): characterization, classification, and utilization of wet soils. USDA-SCS, Natl. Soil Survey Ctr., Lincoln, NE.

Kimble, J.M. (ed.) 1990. Proceedings of sixth international soil correlation meeting (VI ISCOM): Characterization, classification and utilization of cold Aridisols and Vertisols. USDA-SCS, Natl. Soil Survey Ctr., Lincoln, NE.

Kishné, A.Sz., C.L.S. Morgan, and W.L. Miller. 2009. Vertisol crack extent associ-ated with gilgai and soil moisture in the Texas Gulf Coast Prairie. Soil Sci. Am. J. 73:1–10.

Kishné, A.Sz., C.L.S. Morgan, Y. Ge, and W.L. Miller. 2010. Antecedent soil mois-ture affecting cracking of a Vertisol in field conditions. Geoderma. In press.

Mermut, A.R., G.S. Daso, and G.N. Dowuona. 1996. Soil morphology. p. 89–114. In N. Ahmad and A. Mermut. (ed.) Vertisols and technologies for their manage-ment. Developments in Soil Science 24. Elsevier, Amsterdam, The Netherlands.

Miller, D.L,. C.I. Mora, and S.G. Driese. 2007. Isotopic variability in large carbonate nodules in Vertisols: Implications for climate and ecosystem assessments. Geoderma 142:104–111.

Miller, W.L., and A.L. Bragg. 2007. Soil characterization and hydrological monitoring project, Brazoria County, Texas, bottomland hardwood Vertisols. USDA-NRCS, Temple, TX.

Miller, W.L., A.S. Kishné, and C.L.S. Morgan. 2005. Using seasonal crack patterns to evaluate the criteria for ustic and udic moisture regimes for Vertisols in the Texas Gulf Coast Prairie. In 2005 Annual Meetings Abstracts. ASA, CSSA, and SSSA, Madison, WI.

Newman, A.L. 1986. Vertisols in Texas: Some comments. USDA-SCS, Temple, TX.

Nordt, L.C., and S.G. Driese. 2009. Hydropedological model of Vertisol formation along the Gulf Coast Prairie land resource area of Texas. Hydrol. Earth Syst. Sci. 13:2039–2053.

Nordt, L.C, C. Hallmark, L. Wilding, T. Boutton. 1998. Quantifying pedogenic car-bonate accumulations using stable carbon isotopes. Geoderma 82:115–136.

Paton, T.R. 1974. Origin and terminology for gilgai in Australia. Geoderma 11:221–242.

Russell, J.S., A.W. Moore, and J.E. Coaldrake. 1967. Relationships between subtropical semiarid forest of Acacia Harpophylla (Brigalow), microrelief, and chemical properties of associated gilgai soil. Aust. J. Bot. 15:481–498.

Soil Survey Staff. 1951. Soil survey manual. USDA Agric. Handb. 18. U.S. Gov. Print. Office, Washington, DC.

Soil Survey Staff. 1975. Soil taxonomy. A basic system of soil classification for making and interpreting soil surveys. USDA Agric. Handb. 436. U.S. Gov. Print. Office, Washington, DC.

Soil Survey Staff. 1990. Classification and management of wet soils. Guidebook for Texas. VIII Int. Soil Correlation Meeting, Baton Rouge, LA, and College Station, TX. 6–21 Oct. 1990. Dep. of Soil and Crop Sciences, TAMU and USDA Soil Conservation Service, USDA Natl. Soil Survey Ctr., Lincoln, NE.

Soil Survey Staff. 1999. Soil taxonomy. A basic system of soil classification for making and interpreting soil surveys. 2nd ed. USDA. Agric. Handb. 436. U.S. Gov. Print. Office, Washington, DC.

Stiles, C.A. 2001. Pedogenic relationships in a Vertisol climosequence defined by geochemical mass balance of whole soil and chemistry of iron-manganese nodules. Ph.D. diss. Univ. of Tennessee, Knoxville.

Thompson, C.H. and G.G. Beckmann. 1982. Gilgai in Australian Black Earths and some of its effects on plants. Trop. Agric. (Trinidad) 59:149–156.

USDA -NRCS. 1995. WETS table documentation.

USDA-NRCS. 2002. Yearly total precipitation for station TX9364, Victoria WSOUSDA-NRCS. 2000. Selected statistical methods. Chapter 18. In Natl. Engineering Handbook. Part 630 Hydrology.

USDA-NRCS. 2005. National soil survey handbook. Part 629. Glossary of landform and geologic terms.

Wilding L.P., D. Williams, W. Miller, T. Cook, and H. Eswaran. 1990. Close interval spatial variability of Vertisols: A case study in Texas. p. 232–247. In J.M. Kimble (ed.) Proceedings of sixth international soil correlation meeting (VI ISCOM): Characterization, classification and utilization of cold Aridisols and Vertisols. USDA-SCS, Natl. Soil Survey Ctr., Lincoln, NE.

Williams, D., T. Cook, W. Lynn, and H. Eswaran. 1996. Evaluating the field morphol-ogy of Vertisols. Soil Surv. Horiz. 37:123–131.

discussed in light of related studies. A major unsolved problem was

emphasized in that the Udic/Ustic classification criteria with eight normal

months within a normal year were never met in that region during the

study period. Because the climatic conditions in this region do not gen-

erally meet these criteria under normal circumstances, modifying the

criteria is recommended.

In the second part of this article, we summarized a field study of

close-range spatial and temporal dynamics of Vertisol cracking in a

100-m2 area of Laewest clay conducted during 1989–1998. This study

showed that seasonal cracks of a representative Vertisol in the Texas

Gulf Coast Prairie (Laewest clay) always started on, and crack area

density was greater on, microhighs or upper part of microslopes. With

prolonged drying periods, cracks also formed on the microlows. How-

ever, cracks were deeper on microlows on average. Data showed that

cracks generally occur in the same location during a given seasonal

period and were clustered in the same general areas during the 10-yr

study period.

Temporal analysis of cracking indicated that crack area density

depends not only on current soil moisture conditions, but on long-term

history of soil moisture as well. On both microhighs and microlows, the

greatest crack area density was associated with prior above-normal pre-

cipitation and soil water recharge followed by an extended period of

below-normal precipitation and dry soil water conditions. Further studies

are recommended to investigate the long-term effect of antecedent soil

water content on Vertisol cracking under different climatic conditions and

the possible influence on meeting the classification criteria of Udic/Ustic

soil moisture regime.

AcknowledgmentsThis research was funded by the Texas USDA-NRCS and Texas AgriL-

ife Research. Personally, we want to thank past and current USDA-NRCS Texas State Soil Scientists and MO Leaders, Richard Babcock, Gaylon Lane, Mike Golden, Mike Risinger, and Dennis Williamson, for their support of the field stud-ies in the Texas Gulf Coast Prairie. Grateful acknowledgments go to Alan Newman, DeWayne Williams, Dr. Larry Wilding, and Dr. Warren Lynn for their comments and helpful suggestions during the course of the field studies.

ReferencesAhmad, N. 1983. Vertisols. p. 91–124. In L.P. Wilding et al. (ed.) Pedogenesis and

soil taxonomy II. The soil orders. Elsevier, Amsterdam, The Netherlands.

Ahmad, N. 1996. Occurrence and distribution of Vertisols. p. 1–41. In N. Ahmad, and A. Mermut (ed.). Vertisols and technologies for their management. Devel-opments in Soil Science 24. Elsevier, Amsterdam, The Netherlands.

Ahmad, N., and A. Mermut (ed.) 1996. Vertisols and technologies for their manage-ment. Developments in Soil Science 24. Elsevier, Amsterdam, The Netherlands.

Driese, S., L. Nordt, W. Lynn, C. Stiles, C. Mora, and L. Wilding. 2005. Distinguish-ing climate in the soil record using chemical trends in a Vertisol climose-quence from the Texas coast prairie, and application to interpreting Paleozoic paleosols in the Appalachian basin, U.S.A. J. Sediment. Res. 75:339–349.

Fredlund, D.G. 1996. Geotechnical problems associated with swelling clays. p. 499–524. In N. Ahmad and A. Mermut (ed.) Vertisols and technologies for their management. Developments in Soil Science 24. Elsevier, Amsterdam, The Netherlands.

Griffin, R.W., L.P. Wilding, and L.R. Drees. 1992. Relating morphological proper-ties to wetness conditions in the Gulf coast Prairie of Texas. p. 126–134. In