Undisturbed Prairie Research Articles

GRASSLAND MANAGEMENT EFFECTS ON SOIL SURFACE PROPERTIES IN THE OZARK HIGHLANDS

Much of the research regarding the effects on soil properties of grazing and land-use conversion from native prairie to pasture and hay meadows has been concentrated in the northern Great Plains where the climate is relatively cool and dry. However, little research has been conducted to ascertain similar effects on soil properties in the much warmer and wetter mid-southern latitudes. The objective of this study was to characterize the effects of prairie conversion to agricultural grasslands on soil properties in the Ozark Highlands of northwest Arkansas. Three native prairie sites were identified with adjacent grazed or hayed grasslands on silt-loam soils. Four prairie-agricultural grassland comparisons were made for this study, with each comparison occurring on the same soil-mapping unit. Conversion of native prairie to grazing and hay harvesting created significantly (P < 0.05) higher bulk density, soil pH, and extractable P, Mg, and Mn within the top 10 cm. There was a significantly (P < 0.2) larger difference in electrical conductivity, extractable K, Na, and Fe, but a significantly smaller difference in total N and C and organic matter concentration, in the top 10 cm between prairie-grazed and prairie-ungrazed landuse combinations. Overall, soil surface properties of agricultural grasslands in the Ozark Highlands were similar to those of native, undisturbed prairie, indicating that either negligible responses occurred following conversion, parameters recovered to original equilibrium levels, or the magnitudes of agricultural inputs following conversion were too low to change equilibrium levels substantially. The baseline soil data generated in this study for undisturbed prairie and agriculturally managed grasslands can serve as standards for assessing impacts of high-input agriculture and future land-use changes in the Ozark Highlands.

Relationships Among Soil Carbon and Physiochemical Properties of a Typic Albaqualf as Affected by Years Under Cultivation

Soil organic carbon (SOC) is a dynamic soil property and a key component of the quality and sustainability of agricultural soils. Predictions of SOC change often rely on correlations between SOC and other soil physical and chemical properties that are assumed to be more stable over time. The objective of this study was to examine the long-term stability of relationships between SOC, physiochemical properties, and years under cultivation in a common rice (Oryza sativa L.) soil (i.e., a fine, montmorillonitic, thermic, Typic Albaqualf) of the Mississippi Delta region of the south-central United States. The field site consisted of an undisturbed native prairie and three agricultural fields that have been cultivated for 15, 26, and 44 years, respectively, in 2001. Neither SOC concentration nor mass was significantly correlated with sand, silt, or clay content, but both were significantly correlated with pH in 1987. Soil OC concentration and mass were significantly correlated with both silt and clay, but not sand content, and pH in 2001. However, the relationship between SOC concentration and sand content changed significantly, while the relationships between SOC and pH and silt and clay content remained stable and did not change significantly from 1987 to 2001. Soil OC decreased exponentially as years under cultivation increased, but the modeled relationship appeared to remain stable from 1987 to 2001. Results indicate that continuous cultivation of the Typic Albaqualf of this study will result in SOC stabilizing around 8–10 g kg−1, which is less than 50% of the SOC in the adjacent undisturbed, native prairie soil. The results of this study can be used to strengthen predictions of how ecosystems, particularly flood-irrigated rice agroecosystems, will respond to future changes in management practices towards more sustainable use of the soil resource.

Extirpation or Coexistence? Management of a Persistent Introduced Grass in a Prairie Restoration

Abstract Introduced perennial grasses are one of the greatest constraints to prairie restoration. Herbicides suppress but do not eliminate introduced grasses, so we explored the interaction of herbicide with two additional controls: heavy clipping (to simulate grazing) and competition from native species. A 50‐year‐old stand of the introduced perennial grass Agropyron cristatum (crested wheatgrass) in the northern Great Plains was seeded with native grasses and treated with herbicide annually for 7 years in a factorial experiment. Clipping was applied as a subplot treatment in the final 3 years. Both herbicide and clipping significantly reduced the cover of A. cristatum, but clipping produced an immediate and consistent decrease, whereas herbicide control varied among years. The cover of A. cristatum decreased significantly with increasing cover of a seeded native grass, Bouteloua gracilis (blue grama), suggesting that both top‐down (i.e., grazing) and bottom‐up (i.e., resource competition) strategies can contribute to A. cristatum control. No treatment had any effect on the seed bank of A. cristatum. Even in the most effective control treatments, A. cristatum persisted at low amounts (approximately 5% cover) throughout the experiment. The cover of B. gracilis increased significantly with seed addition and herbicide, and, after 7 years, was similar to that in undisturbed prairie. The total cover of native species increased significantly with clipping and herbicide, and species richness was significantly higher in plots receiving herbicide. Clipping season had no effect on any variable. In summary, no method extirpated A. cristatum, but clipping reduced its cover by 90% and doubled the cover of native species. Extirpation might not be a realistic goal, but relatively simple management allowed coexistence of native species.

Vegetation responses to different spatial patterns of soil disturbance in burned and unburned tallgrass prairie

Pocket gopher (Geomyidae) disturbances are created in spatiallypredictable patterns. This may influence resource heterogeneity and affectgrassland vegetation in a unique manner. We attempt to determine the extent towhich density and spatial pattern of soil disturbances influence tallgrassprairie plant community structure and determine how these disturbances interactwith fire. To investigate the effects of explicit disturbance patterns we createdsimulated pocket gopher burrows and mounds in various spatial patterns.Simulated burrows were drilled into the soil at different densities inreplicated plots of burned and unburned prairie. Separate plots of simulatedmounds were created in burned and unburned prairie at low, medium, or high mounddensities in clumped, uniform, or random spatial dispersions. In both burned and unburned plots, increased burrow density decreasedgraminoid biomass and increased forb biomass. Total-plant and graminoid biomasswere higher in burned than unburned plots while forb biomass was higher inunburned plots. Total-plant species richness was not significantly affected byburrow density or burning treatments, but graminoid species richness increasedin unburned plots and forb species richness increased in burned plots. Plant species richness was temporarily reduced directly on mounddisturbances compared to undisturbed prairie. Over time and at larger samplingscales, the interaction of fire and mound disturbance patterns significantlyaffected total-plant and graminoid species richness. The principal effect inburned and unburned prairie was decreased total-plant and graminoid speciesrichness with increased mound disturbance intensity. Although species richness at small patch scales was not increased by anyintensity of disturbance and species composition was not altered by theestablishment of a unique guild of disturbance colonizing plants, our studyrevealed that interactions between soil disturbances and fire alter the plantcommunity dominance structure of North American tallgrass prairie primarily viachanges to graminoids. Moreover, these effects become increasingly pronouncedover time and at larger spatial sampling scales.

Prairie vegetation and soil nutrient responses to ungulate carcasses.

The impact of large ungulate carcasses on grassland dynamics was investigated by monitoring vegetation and soil nutrients in 50-cm circular zones around the center of bison (Bos bison), cattle (B. taurus), and deer (Odocoileus virginianus) carcasses. An ungulate carcass creates an intense localized disturbance that varies with animal size and the season of death. Unlike other natural disturbances, carcasses deposit a concentrated pulse of nutrients into the soil. One year after death, inorganic nitrogen concentrations were significantly higher in the inner 50 cm at both adult and juvenile carcass sites than in surrounding prairie. Areas around a carcass became zones of fertility that favored different components of the vegetation and stimulated biomass production. Species richness and diversity at the center of carcass sites were lowest 1 year after death, but increased significantly in subsequent years. However, warm-season perennial grasses declined near the center of carcass sites and did not recover. Five years after death, ungulate carcass sites remained disturbed patches that harbored vegetation characteristically different in composition and stature from surrounding prairie. By providing a niche for species not normally found in undisturbed prairie, carcasses increased community heterogeneity and may play an important role in adding spatial complexity to grassland ecosytems.

LONG-TERM ECOSYSTEM IMPACTS OF AN INTRODUCED GRASS IN THE NORTHERN GREAT PLAINS

Introduced plants can have negative effects on native species and diversity, but their impacts on ecosystem function are less apparent. At the northern edge of the Great Plains, we examined five stands in each of undisturbed prairie, successional prairie (fields abandoned for ;50 yr that had undergone succession to native grasses), and abandoned fields of similar age planted with Agropyron cristatum, aC 3 perennial tussock grass intro- duced from northern Asia. We attributed differences between successional prairie and A. cristatum stands to the introduced species and not to cultivation. A. cristatum fields con- tained few native species, resulting in significantly lower species richness and diversity. Growth forms most similar to A. cristatum, i.e., C3 grasses, were most likely to be excluded. Soils under A. cristatum had significantly less available N, total N, and total C than soils under successional prairie. A. cristatum fields had significantly higher shoot mass than successional prairie, but root : shoot ratios in A. cristatum fields were less than half those under successional prairie, because root mass was significantly higher under native grasses. C:N ratios did not vary significantly between A. cristatum and native grasses, suggesting that the lower N and C content of soils under A. cristatum was not caused by differences in nutrient concentrations. Instead, the relatively low root mass of the introduced grass was probably responsible for decreased rates of N and C addition to the soil. Because total C was 25% less in soils under A. cristatum than under successional prairie, the planting of A. cristatum over millions of hectares of the Great Plains may have left 3.3-4.8 3 10 14 g of C in the atmosphere that would otherwise have been stored as soil organic matter by native grasses. In contrast to the large effects of A. cristatum, almost no significant dif- ferences were found between undisturbed and successional prairie. Thus, the identity of the species dominating after disturbance played a larger long-term role in determining ecosystem function than did the disturbance itself. The results suggest that the effects of this introduced species extend beyond the displacement of native species and the reduction of diversity, and include the alteration of pools and flows of energy and nutrients in the prairie ecosystem.

DETERMINANTS OF SOIL CO2FLUX FROM A SUB-HUMID GRASSLAND: EFFECT OF FIRE AND FIRE HISTORY

Soil CO2 flux (Jc02) was measured at midday over a 2-yr period in undisturbed tallgrass prairie (Konza Prairie, Kansas, USA) to quantify seasonal and annual budgets, to evaluate temperature and moisture as determinants of soil CO2 flux, and to assess the effect of a common land management tool, spring fire, and fire history on soil respiration. We

Sources of error in direct microscopic methods for estimation of fungal biomass in soil

Sources of error in direct microscopic measurement of fungal hyphae in soil were examined and fungal biomass estimates and associated variability obtained by the direct counting method and the ergosterol technique were compared. Nested random effects ANOVA indicated that the major source of variance in the direct microscopic counting method were the people examining the prepared microscope slides, accounting for 83% of the total variance. Sampling variability accounted for approximately 14% of the total variance. Fungal biomass values calculated from soil ergosterol concentrations were close to the range of values derived from hyphal length estimates but coefficients of variation were much lower for soil ergosterol determinations (6–13%) than for hyphal length estimates (16–32%). For one soil sample, we compared total hyphal length and fungal biomass estimates from our lab to those of another lab. Values obtained by the other laboratory were outside the range of values and 95% confidence intervals reached in our lab. Comparison of fungal hyphal length estimates from undisturbed prairie soil and an adjacent cultivated soil indicated that the undisturbed soil contained more than twice as much fungal biomass. Results of our study indicate: (1) extreme caution must be used when comparing hyphal length and fungal biomass estimates made by different laboratories using the direct counting method; and (2) soil ergosterol determinations can provide information on fungal biomass that may be useful in comparing direct count estimates by different labs.

Surface Burrow Densities in a Sample of Black-Tailed Prairie Dog Towns in Gray and Meade Counties, Kansas

We conducted surveys on sites (towns) of 27 individual colonies of blacktailed prairie dogs (Cynomys ludovicianus) in southwestern Kansas to determine burrow-entrance densities. Mean size of the prairie dog towns surveyed was 16 ha. Estimated mean active burrow-entrance densities were 206/ha, higher than reported for Colorado, Oklahoma, Wyoming, New Mexico, and South Dakota. Active prairie dog burrow-entrance densities are among the criteria used to select reintroduction sites for black-footed ferrets (Mustela nigripes) potentially making the existence of towns with high burrow-entrance densities in Kansas of interest for such efforts if other reintroduction criteria can be met. The black-tailed prairie dog (Cynomys ludovicianus) is a colony-forming rodent that occurs in mixed and shortgrass prairie areas of the Great Plains and is the only prairie dog species native to Kansas (Hall, 1981; Bee et al., 1981). The dependence of the endangered black-footed ferret (Mustela nigripes) on prairie dogs (Henderson, Springer, and Adrian, 1969; Linder, Dahlgren, and Hillman, 1972; Biggins and Schroeder, 1988) and the recent implementation of black-footed ferret reintroduction plans (U.S. Fish and Wildlife Service, 1988) have focused attention on prairie dog habitat. Scant data are available on prairie dog habitat in Kansas. Prairie dogs construct extensive burrow systems, some of which are per'Corresponding author. This content downloaded from 157.55.39.116 on Sun, 18 Sep 2016 06:10:27 UTC All use subject to http://about.jstor.org/terms VOLUME 97, NUMBERS 3-4 103 manent and are inhabited continuously (Smith, 1958). Although prairie dogs excavate some new burrow systems each year, others disappear from lack of use. The total number of burrow entrances changes by fewer than 10/ha annually in undisturbed prairie dog colonies (Hoogland et al., 1987). Although exact data are lacking on the relationship between burrow-entrance densities and black-tailed prairie dog populations, burrow-entrance density is considered an indicator of potential prairie dog abundance (Biggins et al., 1989). We conducted this study to determine burrow-entrance densities on prairie dog towns in southwestern Kansas.

VAM INNOCULATION OF SUDEX USING PRAIRIE REMNANT SOIL

Soil samples collected from four remnant prairie plant communities in Indiana were used as innoculum for pot cultures of Sudan grass seedlings (Sorghum sudanense). The growth of seedlings was correlated with the quality of the prairie remnant; the less degraded sites provided innoculum which resulted in the most vigorous plant growth in culture. Sudan grass root segments were harvested and prepared for assay of vesicular-arbuscular mycorrhizae. VAM was shown to be present in innoculated samples. Results indicate that soil collected from undisturbed prairie remnants may provide a better source of VAM innoculum for larger scale production of plants which associate with VA mycorrhizae.

CHANGES IN SOIL WATER REGIME AFTER PIPELINE CONSTRUCTION IN SOLONETZIC MIXED PRAIRIE RANGELAND

A study was conducted in the Solonetzic mixed prairie rangeland of southern Alberta to evaluate soil water responses to pipeline construction. Five adjacent natural gas pipelines on a series of rights-of-way (ROW) were studied at three sites. The pipelines were installed in 1957, 1963, 1968, 1972 and 1981. Soil water contents were monitored throughout the year using a neutron probe. Water-holding capacity of the soil was determined using the pressure-plate extraction method. Pipeline construction in Solonetzic mixed prairie ecosystems caused total soil water to a depth of 50 cm in the trench to increase over that in the undisturbed prairie. There were no significant effects on available water capacity. Berm construction over the trench impeded overland flow, particularly during springmelt, causing short periods of ponding. Construction activity had no significant effect on total water within zones of a given ROW. Soil water status was affected for a short period of time with a trend towards predisturbance conditions within 10 yr. Key words: Pipelines, Solonetzic soils, rangeland, reclamation, soil water

Microbiology of stored topsoil at North Dakota stripmining sites

Abstract The effect of topsoil storage on various microbial populations and parameters was determined. Topsoil piles that had been stored from 6 months to 12 years were sampled, as was an undisturbed prairie site. Samples were taken at all sites from the near‐surface (0 to 8 cm) and from core‐depth (110 to 120 cm). Counts of bacteria, actinomycetes, and fungi generally did not decline markedly with long‐term storage and were usually higher at core‐depth in such piles than in undisturbed prairie soil. However, alkaline phosphatase activity, an indicator of microbial biomass, and dehy‐drogenase activity, an indicator of biological activity, were both lower in the topsoil pile near‐surface samples than in the undisturbed prairie sample. Counts of nitrifying bacteria tended to be higher in the topsoil pile samples at core‐depth than in the undisturbed prairie soil sample. There were more asymbiotic nitrogen‐fixing bacteria in the storage piles than in undisturbed prairie soils. They were also more abundant at...

Vesicular–arbuscular mycorrhizal fungi associated with native tall grass prairie and cultivated winter wheat

More vesicular–arbuscular mycorrhizal (VAM) fungal species and significantly more fungal spores were recovered from undisturbed prairie soils than four winter wheat field soils in Kansas through the 1980–1981 growing season. Two previously undescribed sporocarpic species of Endogonaceae were found in prairie samples but have not been successfully established in pot culture, leaving the genus to which they belong unclear. Though variable, 11–50% VAM root colonization was evident in all prairie grass roots sampled throughout the year. In contrast, no identifiable VAM root colonization was evident in wheat until May after flowering when 27% root colonization was observed. During the 1981–1982 growing season, roots of two other wheat fields were sampled with similar results. No colonization occurred until May when 8% root colonization was evident. The possible influence of such low levels of root colonization occurring quite late in the growing season of winter wheat is discussed.

The Effect of the Pesticide Carbofuran on Soil Organisms and Root and Shoot Production in Shortgrass Prairie

To test the hypothesis that ectoparasitic phytophagous nematodes are a major consumer of primary production in arid grasslands, carbofuran was applied once in each of two successive years at 12 lb active ingredient/acre to plots of undisturbed prairie in northeastern Colorado. Bacterial feeding and phytophagous nematodes were reduced to 10% of the density on untreated plots but microand macro-arthropods were apparently unaffected. Species composition of the fungal flora of the treated soil was greatly changed and colony counts reduced but incidence of mycorrhizae was unaffected. Although root weight measured in the autumn was significantly greater under treated plots (by 130-150 g m-2 and about 25% of NPP) shoot production was not, possibly due to below average summer rainfall in both years. Nevertheless, our results suggest that nematodes may be the most significant consumer in shortgrass prairies.

Enumeration of acetylene-reducing bacteria in strip-mined reclamation sites in a temperate grassland

From 36 to 71% of bacteria, depending on the sampling site, that were isolated from the soil or rhizosphere of undisturbed prairie soil or reclamation sites of strip-mined grassland areas in western North Dakota were capable of reducing acetylene. These bacteria generally could be divided into two populations; one capable of acetylene reduction under aerobic conditions and another capable of acetylene reduction under anaerobic conditions. The reclamation site to which no topsoil had been applied, pH 8.5, had a bacterial population which generally was capable of higher levels of acetylene reduction than individual bacteria isolated from other sites.

EFFECTS OF TOPSOIL STORAGE DURING SURFACE MINING ON THE VIABILITY OF VA MYCORRHIZA

Storing topsoil for 3 years was shown to reduce substantially the levels of viable inocula relative to levels in adjacent, undisturbed prairie soils. The detrimental effect of storage on VA mycorrhiza is associated with the loss of viability of mycorrhizal fragments occurring in the stored soil. Data are also presented supporting an interaction between infected root segments and roots of uninfected plants as a major means of spreading mycorrhiza in these soils.

The natural history of a fugitive prairie plant (Mirabilis hirsuta (Pursh) MacM.).

Some perennial fugitive plants that colonize badger disturbances in xeric prairies have a limited dispersal capacity, and consequently propagules are dispersed over a small area. I hypothesized that high density-dependent mortality might occur early in the life history of such species, and thus increased survival might occur in subsequent age classes because intraspecific competition would be reduced. These hypotheses were tested using natural and experimental cohorts of Mirabilis hirsuta (Pursh) MacM. From these data and field observations, inferences were obtained concerning selective forces operating upon life history characteristics of this species.The distance between individuals of M. hirsuta increases in successive age classes; the greatest decrease in density occurs between the propagule and seedling age classes. Mortality of propagules due to predation by ants and mice was density-dependent. Predation rates were highest at high propagule densities and predation upon propagules located on badger disturbances was higher than the mortality of propagules at similar densities in undisturbed prairie. The results of mortality in the propagule age class are seedlings present only at low densities and located away from parent plants. Seedlings survive to maturity only if they are located on badger disturbances; this species apparently can not successfully compete with plants present in undistrubed prairie. On badger disturbances seedlings present at low densities have much higher survival (roughly 50%) to maturity than do seedlings present at high densities (essentially zero). Thus, if high densities of propagules occur on a disturbance, predation upon propagules results, indirectly, in increased survival of seedlings to maturity. Such predation potentially could have important effects upon interspecific competition of M. hirsuta with other fugitives also colonizing badger disturbances.Reproductive success of M. hirsuta on the Cayler Prairie Preserve is contingent upon successful colonization of disturbance sites. It would appear that selection has operated upon the life history characteristics to favor both successful immigration onto new sites and establishment of seedlings on those sites. Relatively few, but large propagules are produced annually over a long adult life span. While large propagules enhance seedling establishment on xeric sites, production of few propagules annually for a number of years increases the likelihood of immigration onto sites that are variable in the time of appearance within the dispersal range of the plant.

Remnant Grassland Vegetation in the Central Great Plains of North America

Numerous ecological studies concerning the vegetation of grasslands in the Great Plains have been conducted, from the prairies of south-central Canada to those of westcentral Texas. Studies such as those by Coupland (1950, 1961), Moss (1955) and Hulett, Coupland & Dix (1966) characterized much of the vegetation of the Great Plains region of Canada. Hanson & Whitman (1938), Tolstead (1941) and Larson & Whitman (1942) investigated the Great Plains vegetation of the northern United States. In the Southern Great Plains, Brumer (1931) described the vegetation of Oklahoma and Dyksterhuis (1946) and Launchbaugh (1955) wrote about Texas prairies. The area concerned in the present study is the Central Great Plains of North America. A detailed study portraying the ecology of remnant grassland vegetation in the region is unprecedented. The majority of the ecological studies in Kansas (Hopkins 1955) concern grazing, drought or both. Studies by Clements (1920), Albertson (1937), Albertson & Weaver (1942, 1946), Tomanek & Albertson (1953, 1957), and Albertson & Tomanek (1965) were important in characterizing grazing and drought ecology of western Kansas grasslands. Weaver & Albertson (1956) compiled the results of many studies concerning Great Plains grasslands. The vegetation of Nebraska has been qualitatively characterized in its entirety by Weaver (1965). Pool (1914), Tolstead (1942) and Burzlaff (1962) reported on areas in the Nebraska Sandhills and Weaver & Bruner (1948) and Hopkins (1951) characterized the vegetation of the loess hills in the central portion of the state. Most recently, information concerning the Pine Ridge vegetation in north-western Nebraska has been given by Nixon (1967). Lang (1945) reported upon stands in adjacent eastern Wyoming, while areas in Colorado have been characterized by Vestal (1914), Ramaley (1939), Costello (1944) and Livingston (1952). The primary objective in the present study is a detailed ecological evaluation of the vegetational characteristics of some grassland remnants and their sites in six of the major land resource areas (Austin 1965) in the Central Great Plains. When the criteria Austin used to delineate these resource areas were examined, we decided they would be a meaningful foundation upon which a large number of undisturbed prairie remnants could be characterized and compared.

Rate of Decomposition of Roots and Rhizomes of Certain Range Grasses in Undisturbed Prairie Soil

That grass is an excellent preventive of soil depletion and the best means of soil improvement as well as man's most efficient weapon against soil erosion have been pointed out in a previous paper (Weaver and Zink, '46). In that study the annual production of root materials of certain perennial grasses and the number of growing seasons necessary for maximum production were ascertained. An equally fundamental question in grasslegume and grain-crop rotations (alternate husbandry) is the rate of decomposition of these underground plant materials when the land is plowed, so that their constituents may be utilized by succeeding crops. The rate of decomposition of plant materials when incorporated with soil has been the subject of numerous investigations. These have dealt almost entirely with the aerial portions of plants and furnish little or no information pertaining to the decomposition of roots. Where the roots have been studied, portions of them almost invariably have been removed from the soil, ground or at least cut into small pieces to which soil and nutrients were added, and kept under optimum conditions of moisture, aeration, and temperature to promote rapid decomposition. Although such procedure yields valuable information

Some Edaphic Changes Due to Overgrazing in the Agropyron‐Poa Prairie of South‐Eastern Washington

The course of vegetational retrogression resulting from the overgrazing of native prairies has been described for many regions. Relatively little effort has been made, however, to carry these investigations far enough to include those subsequent changes in the soil which follow the changes in vascular plant populations. It is well known that the ill effects of excessive grazing pressure upon natural plant communities are not alone measured by the loss of valuable forage species; the environment may be so altered in the course of range deterioration that the process of restoration of a satisfactory forage cover becomes greatly complicated, and extensive damage through erosion and consequent deposition may be incurred. By providing an understanding of the nature and degree of soil deterioration, investigations in this field may be of material aid in a program of range restoration. Furthermore, from a theoretical standpoint, such studies cast much light upon vegetational influences, i.e., the degree to which a plant association determines its own environment. In an earlier paper (Daubenmire, '40) the major vegetational changes wrought by intensive grazing upon the Agropyron spicatum-Poa secunda prairie of southeastern Washington were described. For the region centering about the eastern extremity of Franklin County, a fairly detailed description was given of the general environment, nature of the initiating causes of succession, and the changes in vascular plant population. It became evident during the course of this study that the alteration in macroscopic vegetation, which is the immediate effect of heavy grazing, becomes in turn a causal factor which initiates a complicated chain of readjustments involving aerial, edaphic, and biotic phases of the ecosystem.1 In the present paper attention is directed to certain of these secondary, or even more remote, effects of grazing. An especially favorable opportunity to conduct studies along this line was afforded in eastern Franklin County by a circumstance in which a spur along the canyon of the Palouse River was isolated from the remainder of the uplands by the excavation of a railroad cut approximately 10 meters deep, in 1911. The vegetation of this artificial butte, which is about a quarter mile long and half as wide, is inaccessible to livestock, and, having had 28 years to recover from the effects of the light grazing to which it was subjected prior to its isolation, appears to be quite representative of the undisturbed prairie which dominated the landscape before the region was settled by the white man. On the opposite side of this railroad cut the prairie has been so heavily grazed that most of the valuable forage plants have disappeared. For the present investigation two stations were selected on opposite sides of the cut, and as close to it as was deemed advisable. In this way it was possible to make an intensive study of two small areas separated by a distance of 50 meters, one of which is in essentially virgin condition and the other severely overgrazed. Both areas are on the same soil type, which is essentially identical with Ritzville silt loam, and appear to differ significantly only in the grazing factor. To facilitate comparisons between the present study and others which may be made in different regions, the following description of the profile in the virgin area is offered.