By Scott P. Schell

A Thesis Submitted to the Department of Plant, Soil and Insect Sciences and The Graduate School of the University of Wyoming in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE in ENTOMOLOGY Laramie, Wyoming July 1994



Spatial Analysis of Ecological Factors Related to Grasshopper Outbreaks in Wyoming

ABSTRACT Spatial analysis was used to test the hypotheses that the frequency of grasshopper outbreaks (>9.6 grasshoppers per m2 as defined by USDA-APHIS) in Wyoming are dependent on: vegetation, precipitation, elevation, evapotranspiration, land form, and soil type. Thirty-one survey maps, compiled yearly by state and federal agencies, showing two levels of grasshopper densities (areas with <9.6 grasshoppers per m2 and areas with >9.6 grasshoppers per m2) were electronically overlaid with a geographic information system (GIS, ERDAS 7.5 PC). The resulting map showed that grasshopper outbreaks occur repeatedly in some areas but rarely in others, even though they had suitable vegetation and endemic populations of grasshopper species capable of reaching high densities. Maps of all the ecological factors were digitized and the areas of geographic overlap with the grasshopper outbreak frequency map were determined with GIS. One-way Chi-square analysis was performed on the distributions of the areas that supported outbreaks 0, 3, 6, 9, and 12 yr out of 31 yr with respect to each of the ecological factors. These analyses showed that the distributions were significantly (P < .0001) non-random for the five different outbreak frequencies on all ecological factor layers. Cross-tabulation of vegetation type and grasshopper outbreak frequency revealed that ponderosa pine (Pinus ponderosa ) vegetation type (which covers only 5.5% of Wyoming) is dominant in the areas that had outbreaks in 9 to 13 yr out of 31 yr. Cross-tabulation of soil type and outbreak frequency showed that soil types in areas of Wyoming's eastern plains with steep hills dominate areas that have been infested in 2 to 13 yr out of 31 yr. A model for describing the study area's susceptibility to grasshopper outbreaks using slope, soil depth to bedrock and soil erodibility was developed. Regression analysis showed the model to be highly significant (P<.001) with an r2 value of 0.885.

KEY WORDS: spatial analysis, grasshopper outbreaks, ecological factors 


Rangeland grasshopper survival and reproduction are dependent on exogenous biotic and abiotic factors in their environment. Weather, food quality and quantity, soil type, predation, and disease are acknowledged as factors that may regulate grasshopper population dynamics (Hewitt 1985). The precise influences these factors and their interactions have in promoting or preventing outbreaks in grasshopper species complexes are not well understood. Outbreak densities are defined here as the action threshold of >9.6 grasshoppers per m2 used by the United States Department of Agriculture (USDA) Animal and Plant Health Inspection Service (APHIS) (USDA-APHIS 1987). Although the >9.6 grasshoppers per m2 is an administrative definition, it also has ecological meaning. The APHIS outbreak density slightly exceeds the carrying capacities of 6.3 to 8.9 grasshoppers per m2 that Kemp and Dennis (1993) determined for the three major physiographic regions in Montana. Wyoming and Montana have similar climates and vegetation (Lockwood et al. 1988) so the grasshopper carrying capacities should be comparable. The actual economic threshold for grasshopper outbreaks is dependent on range and economic variables that vary widely across the western U.S. (Davis et al. 1992). Abiotic factors (e.g., precipitation and soil type) and sessile biotic factors ( e.g., vegetation) can be mapped and related spatially to grasshopper outbreaks more readily than ephemeral or mobile biotic factors such as grasshopper predators and pathogens. Mapping grasshopper population densities requires expensive yearly survey efforts; the effort that would be required to map yearly density changes in all the organisms that prey on or parasitize grasshoppers is currently inconceivable. A more practical approach to understand the ecological constraints on grasshopper population dynamics is to utilize maps of relevant ecological factors that have been produced for other purposes.

Maps of five major environmental factors are available for Wyoming: vegetation type, precipitation, elevation, evapotranspiration and soil type. First, vegetation type plays an important role in the distribution of grasshoppers (Isely 1938). Second, precipitation has complex effects on grasshopper populations. Precipitation determines the quality and quantity of food available and the survival and transmission of grasshopper fungal pathogens (Pickford and Riegert 1964). Extreme (dry or wet) soil surface moisture levels can cause grasshopper egg mortality (Hewitt 1985). Third, elevation plays an important role in the distribution of precipitation in Wyoming (Martner 1986). In Idaho, elevational differences in vegetation were associated with differences in grasshopper densities (Fielding and Brusven 1993). Fourth, mean annual potential evapotranspiration is a measure of water loss from the soil surface and vegetation, based on temperature (Marston ed. 1990). Temperature also has a direct affect on grasshoppers by determining the speed of development of nymphs and fecundity of adults (Pickford 1970, Hewitt 1985). Finally, the single best ecological predictor of grasshopper outbreak frequency might be the soil because it integrates the previous four factors. The soil of an area is a function of the five soil-forming factors: climate, parent material, organisms, topography, and time (Singer & Munns 1987). Ironically, of the aforementioned ecological factors, soil has been given the least attention by acridologists. Food plants and weather have been extensively studied (Pfadt 1949, Dempster 1963, Pickford 1970, Hewitt 1979, Hewitt et al. 1982, Joern and Gaines 1990). However, the importance of soil to grasshopper population dynamics has long been suspected. Soil plays a critical role in the life cycle of most grasshopper species. All rangeland pest species oviposit their eggs in the top 2-5 cm of soil (Pfadt 1988). There, the eggs must undergo partial embryonic development, enter diapause for the winter, break diapause (at the appropriate time of year for the species), complete development, and hatch. In principle, the properties of the soil can affect all of these processes and ultimately determine the survival of the grasshopper eggs. These theoretical constraints are supported by empirical evidence. King (1939) observed that Melanoplus sanguinipes (F.) grasshoppers must produce approximately twice the egg population in heavy clay soils as they do in light sandy soils to achieve the same population density of nymphs. Isely (1938) showed that certain species of grasshoppers are very discriminating about the soil type they select for oviposition. Isely (1938) described a habitat dominated by the food plant (Evax multicaulis DC.) of a monophagous grasshopper (Acrolophitus variegatus (Thos.)). However, the habitat did not support any of this grasshopper species, a phenomenon that Isely attributed to the type of soil on the hill. Johnson (1988) used spatial analysis to determine that grasshopper abundance was significantly related to soil type in Alberta, Canada. However, Johnson ascribed this relation to "geographic covariables" such as vegetation type, agricultural practices, and weather, rather than the particular physical qualities of the soil.

Geographic information system (GIS) technology provides a method to summarize and analyze data that are spatially related (Liebhold et al. 1993). GIS has been widely used in insect ecology to characterize habitat susceptibility to pest outbreaks (Liebhold et al. 1993). This technology has been used for spatial analysis of grasshopper outbreaks in Alberta,(Canada) Colorado, Idaho, and Montana (USA) (Johnson and Worobec 1988, Johnson 1989, Carter 1992, Fielding and Brusven 1993, Cigliano et al. 1994). Schell (1994) used GIS to construct a map of grasshopper outbreak frequencies from 31 of Wyoming's annual adult grasshopper survey maps. The map showed that the frequency of grasshopper outbreaks varied markedly across the state, even in areas recognized as suitable habitat, (Schell 1994). The purpose of the current study was to analyze the geographic overlap of grasshopper outbreak frequencies with vegetation, precipitation, elevation, evapotranspiration and soil data for the state of Wyoming. This approach was intended to provide a greater understanding of the ecological factors that affect rangeland grasshopper population dynamics. It must be kept in mind that this analysis is not an experiment with controls, so it is not possible to isolate and examine the effects of one ecological data layer at a time. The data layers used in this analysis are naturally correlative and uncontrollable (Martinat 1987). For example, precipitation varies with elevation, and the vegetation and soils vary with both precipitation and elevation. The goal of this analysis is to determine if each of the aforementioned ecological factors is spatially related to the grasshopper outbreaks. A model of grasshopper outbreak susceptibility was developed based on the most informative factors.

Materials and Methods

Grasshopper Data. The grasshopper outbreak frequency map generated from 31 years of Wyoming annual adult grasshopper surveys map by Schell (1994) was used in this study. This map is composed of Grasshopper Outbreak Frequency Classes (GOFC) that represent the number of years, from 1960 to 1993 (except for 1982-1984), that grid cells (i.e. finite map areas 1000 m square) have had grasshopper outbreaks. The maximum possible GOFC for any grid cell would be 31, but the highest GOFC observed on the map was 15. The minimum GOFC value for any grid cell would be 0, and large areas of the map had this value.

Ecological Factor Data. Each ecological data layer was produced by digitizing and converting of maps of ecological factors into raster format (grid cells) with the DIGPOL and GRDPOL routines (ERDAS 1991). The vegetation and mean annual precipitation maps were digitized from Roberts (1989). The vegetation map classes were: alpine tundra, sand dunes, sagebrush steppe, desert shrublands, mixed-grass prairie, juniper woodland, oak woodland, ponderosa pine forest, spruce-fir forest, douglas fir forest, lodgepole pine forest, and aspen forest. The precipitation map classes were: <20 cm, 20-30 cm, 30-40 cm, 40-50 cm, 50-100 cm, and >100 cm. The elevation data were digitized from Martner (1986). The elevation map classes were: <1200 m, 1200-1500 m, 1500-1800 m, 1800-2100 m, 2100-2700 m, 2700-3350 m, and >3350 m. The mean annual potential evapotranspiration (the calculated rate of water loss to the atmosphere from plants and the surface of the soil related to ambient temperature) was also digitized from Marston (1990). The evapotranspiration map classes were: <43 cm, 43-48 cm, 48-53 cm, 53-58 cm, 58-63 cm, and >63 cm. Soil is the most detailed ecological data layer available for Wyoming. The soil data were taken from Young and Singleton (1977). Three different digital maps were generated from the soil map. One map was composed of generalized soil types based on landform and climate (Young and Singleton 1977). The landform-soil map classes were: mountains (moist all year), mountains (moist in summer), mountains (dry), basins, river valleys and sand hills, steep uplands, and gentle uplands. The second soil map was composed of 69 associations of soil taxonomy's Great Group classification (Young and Singleton 1977); each association contained 1 to 3 Great Groups. Young and Singleton (1977) provided a table of the subgroups, families, representative series and selected properties of each soil association on the map. A limitation of this map is that different soil associations may differ by only one Great Group from each other. Another problem with the map is that soil associations on the state map contained many different soil series, but no information was provided regarding what percentage of a soil association was occupied by each soil series. The third soil map was made up of just those associations found in the eastern plains area of Wyoming (Young and Singleton 1977). Twenty-two soil associations made up this area. Four similar soil associations were combined into two classes in order to generate a map with 20 classes so it could be used with a version of MSUSTAT Chi-square statistical program that is limited to twenty classes (MSUSTAT 1986).

All maps were converted to a common projection system (Lambert Conformal Conic) after being digitized. This conversion allowed all map grid cells of every data layer to correspond geographically to the GOFC map upon superimposition. The grid cell size of all maps was standardized at 1000 m square so all ecological factor maps would be compatible with the GOFC map.

GIS Analysis. A cross-tabulation routine, SUMMARY (ERDAS 1991), was used to produce list files with the area of geographic overlap of each ecological factor's classes correspondence with all of the Grasshopper Outbreak Frequency Classes (GOFC) from the 31 yr outbreak map. The SUMMARY routine also calculated the mode, diversity, and modal density for each cross-tabulation. The modes were the classes within each ecological factor with the most geographic overlap in each GOFC. Diversity was the total number of different classes within each ecological factor that can be found in each GOFC. Modal density was the geographic overlap of the mode class within each ecological factor category on each GOFC; the percentage of each GOFC's total area can be directly calculated.

Statistical Analysis. Chi-square analysis was used to determine if the distribution of infested grid cells (for GOFCs 0, 3, 6, 9, and 12) in the ecological factor classes (vegetation, mean annual precipitation, elevation, mean annual potential evapotranspiration, landform-soil and eastern plains soil associations) differed significantly from chance. This analysis could not be conducted on the complete state soil association data because the 69 soil classes exceeded the capacity of the statistical package (MSUSTAT 1986). The GOFCs 0 and 3 represent those areas of Wyoming in which grasshopper outbreaks are rarely a problem (62.55% and 4.64% of the state respectively). The GOFCs 6 and 9 represent areas of the state where grasshopper outbreaks are a chronic problem (2.29% and 0.51% of the state respectively). GOFC 12 represents the small percentage of Wyoming where outbreaks are extremely frequent (0.05% of the state).

Outbreak Susceptibility Model. In an effort to predict outbreak susceptibility in Wyoming using the available soil properties, a regression model was developed:

Years Infested = A(Slope) + B(Depth) + C(Erodibility) + Error

These particular soil properties were chosen based on the characteristics of outbreak areas observed in southeastern Wyoming from 1990 to 1993 (Lockwood and Schell 1994). Each of the 69 soil associations found in Wyoming has a list of representative soil series (the most specific classification in soil taxonomy) that are found within the association (Young and Singleton 1977). Selected properties, (e.g., percent slope range, soil erodibility, and surface texture) of the soil series are presented in a tabular form (Young and Singleton 1977). The three variables chosen for the model are numeric, and average values were calculated for each mode soil association using the following procedures. Slope was calculated as the mean of the upper value in the percent slope range of all of the soil series present for each GOFC's mode soil association. The upper value in the percent slope range was used because it emphasized the presence of steep terrain in the soil association. Steep terrain was emphasized based on field observations, made from 1990 to 1993, in southeastern Wyoming where most of the grasshopper outbreaks were observed in association with steep terrain (Lockwood and Schell 1994). Slope is a good indicator of the extent to which the local terrain is dissected. Grasshoppers select oviposition sites with warmer soils (Fisher 1993), and slope captures the possibility that grasshoppers select oviposition sites or are more successful in developing or reproducing in habitats with southern aspects. Soil depth to bedrock indicates the type of terrain found in an area. Level plains usually have very deep soils, and steep areas usually have shallow soils. Every soil series is given a range of depth to bedrock, unless the depth to bedrock exceeds 60 inches (152.4 cm). Depth for each modal soil association was calculated as the mean of midpoints of the soil series ranges for depths less than 60 inches (152.4 cm) and 60 inches (152.4 cm) for all series with depth to bedrock greater than 60 inches (152.4 cm). The variable K, the soil erodibility factor, is the mean of all of the representative soil series K values present in a modal soil association. Soil erodibility (K), represents soil texture factors that could play a role in the success of hatching (King 1939). Mean K values ranged from .27 to .33 for the modal soil associations.

Results and Discussion

Vegetation. The results of the cross-tabulation of the GOFC map with the vegetation data revealed several ecological associations. The domination of the GOFC 0 and 1 by sagebrush steppe vegetation was expected because that vegetation type dominates a large part of the state that has had no recorded outbreaks (Fig. 1 and Fig. 2). Although sagebrush steppe vegetation is not strongly associated with grasshopper outbreaks in Wyoming, outbreaks of pest species common to Wyoming frequently occur on sagebrush steppe vegetation in Idaho (Fielding and Brusven 1990).

The cross-tabulation also showed that two vegetation classes dominated the other GOFCs. The Ponderosa pine forest class covers approximately 5.5 percent of the state while the mixed-grass prairie class covers approximately 22.1 percent (Table 2). The Ponderosa pine forest class was strongly associated with chronic grasshopper outbreaks in Wyoming. The modal vegetation class for GOFCs 9-13 was Ponderosa pine forest (Table 1). The Ponderosa pine forest class is misleadingly named; Roberts (1989) explains that this class actually contains open woodlands with an understory of grasses and herbs. Ponderosa pine vegetation also accounted for a major proportion of most of the other non-zero GOFCs as well (Table 3 ). In field surveys from 1990 to 1993, I have seen very few grasshoppers in dense ponderosa pine forest, but I have observed many grasshopper outbreaks in the ecotone between the ponderosa pine forest and mixed-grass prairie in southeastern Wyoming (unpublished data). Mixed-grass prairie was the modal class for GOFCs 2-8, and 14-15. This is an expected result as grasshoppers outbreaks are often associated with prairie vegetation.

Precipitation. The cross-tabulation of GOFCs with mean annual precipitation data shows that GOFCs 1-10 occur most frequently in the 30-40 cm precipitation class (Table 4). The 30-40 cm precipitation class is the dominant class in Wyoming (Table 5). However, the modal value of the five highest GOFCs (11-15) was the 40-50 cm precipitation class, which covers only 10.5% of the state (Table 5). The eastern plains of Wyoming, where most grasshopper outbreaks occur, is dominated by the 20-30 cm and 30-40 cm precipitation classes (Fig. 1and Fig. 3). The annual precipitation levels in Wyoming can vary widely from mean levels (Martner 1986). A cross-tabulation of GOFC and precipitation variance, if the latter data were available, might be more revealing than using mean annual precipitation data. Capinera and Horton (1989) found, that depending on latitude, grasshopper outbreaks were favored by weather conditions that vary from the norm. The grasshopper outbreaks in Wyoming and Montana are favored by abnormally hot and dry conditions, while grasshopper outbreaks in Colorado and New Mexico are favored by increased spring and summer moisture (Capinera and Horton 1989). African locust outbreaks are often associated with the termination of droughts (Dempster 1963).

Elevation. The cross-tabulation of GOFC and elevation showed that, except for GOFC 12, all non-zero GOFC modal elevation classes were < 1500 m in elevation (Table 6). The mode for GOFC 12 was the 1500-1800 m elevation class (Table 6). Most of Wyoming is higher than 1500 m in elevation (Table 7). The elevation mode of GOFC 11 and 13-15 is the <1200 m class; this class covers only 3.6% of Wyoming (Table 7). Elevations in Wyoming decrease from west to east, with the lowest areas being found in the northeastern corner of the state (Fig. 4). The northeastern corner of the state was the location of the highest GOFC (Fig. 1). Spring temperatures are higher earlier in the lower elevation areas of Wyoming (Martner 1986), and hatching of grasshopper eggs and development of nymphs are hastened by warmer temperatures (Hewitt 1985). The speed of development to sexual maturity may be very important to maximize egg production (Pickford 1966, Hewitt 1985). Both the fecundity of female grasshoppers and survival of the eggs is enhanced when oviposition occurs before temperatures start to fall in the late summer (Pickford 1966, Hewitt 1985).

Evapotranspiration. The cross-tabulation of GOFC and mean annual potential evapotranspiration showed that the 53-58 cm and 58-63 cm evapotranspiration classes were the modes for all non-zero GOFCs (Table 8). These two evapotranspiration classes cover 54.2% of the state (Table 9). This indicates that most grasshopper outbreaks occur in areas with very similar temperature regimes. Temperature in Wyoming is affected by elevation (Marston 1990). The modal evapotranspiration class for GOFC 0 was 48-53 cm, a value associated with the cooler regions that cover most of the lower mountain ranges in the state (Fig. 4 and Fig. 5). These regions with cooler temperatures are able to support some of the common pest species (Lockwood et al. 1993), but these conditions may prevent the grasshoppers from developing and reproducing at a rate necessary to reach outbreak levels (Hewitt 1985).

Landform-Soil. The results of the cross-tabulation of the GOFC and landform-soils was very similar to the vegetation cross-tabulation in one respect. The mode of GOFCs 0 and 1 is the basins soil class; this landform-soil is associated with sagebrush steppe vegetation. The basins soil class and sagebrush steppe vegetation cover approximately 37.6% and 46.0% of Wyoming, respectively (Table 1 and Table 10). The basin soil class contains the soils that have formed in the intermountain basins and foothills of Wyoming from transported or residual materials (Fig. 6) (Young and Singleton 1977), and the modal class for GOFCs 2-8 was steep uplands (Table 11). The modal class for GOFCs 9-15 was gentle uplands (Table 11). Both steep uplands and gentle uplands soil classes are found on the eastern plains of Wyoming. The gentle upland class is composed of those soils that have formed on level to gently rolling upland plains (Young and Singleton 1977). The steep upland class includes those soils formed from residual material in areas with slopes ranging from 5 to 50% (Young and Singleton 1977). Young and Singleton (1977) put all of Wyoming's eastern plains into one climatic zone with widely varying daily and seasonal temperatures and a semi-arid precipitation regime. These parameters generally agree with the other ecological factors addressed in this study.

Soil Associations. The cross-tabulation of the GOFC and soil associations revealed that out of 69 soil classes, 59 classes were infested at least once (Soil Assocation Map). However, only four soil classes were the modes for all of the non-zero GOFC classes (Table 12). All of the modal soil associations are found on the eastern plains except for BF-10, the modal soil of GOFC 0. BF-10 is the map code of the Torriorthents-Haplargids-Natrargids association; the percentage that each Great Group comprises of the association is not given. BF-10 is a Basins landform soil from the sagebrush steppe region of western Wyoming. Torriorthents are aridic, common soils of the Entisol soil order (Entisols are soils that lack developed pedogenetic horizons) (Munn 1992). The lack of soil development may be due to an aridic moisture regime or the topographic position of the soil (e.g. steep south facing slopes). The Haplargids Great Group include simple soils (i.e., lacking complex horizons) of the Aridisol order (soils that form under arid climates) with a diagnostic argillic horizon. An argillic horizon is an arbitrarily designated, relative increase in clay content in a subsurface horizon caused by leaching of clay (Munn 1992). An argillic horizon is an indicator of stability in the soil forming processes because of the time it takes for the leaching of clay in the soil profile formed under an arid climate (Munn 1992). The Natrargids Great Group is an Aridisol with a diagnostic natric horizon. A natric horizon is an accumulation of clay and sodium that has strong columnar structure (Munn 1992). If a natric horizon occurs close to the soil surface plant growth can be severely impeded due to the limited infiltration of precipitation caused by the sodium affected clay. This would limit the food supply available to all herbivores.

The modal soil of GOFC 1 was P-13, which is the Haplargids-Paleargids-Torriorthents association (Table 12). This association has two of the three Great Groups found in the BF-10 association, even though it occurs in the eastern plains. The Paleargids Great Group indicates a very old Aridisol soil with an argillic horizon. The Haplargids Great Group found in both the BF-10 and P-13 associations is indicative of stability in the soil forming factors over time in those areas.

The modal soil of GOFC 2-8 was P-4, the Torriorthents shallow association (Table 12). The Torriorthents Great Group is often associated with steep slopes where erosion removes soil faster than it can form and "shallow" indicates that the depth to parent material is minimal. It is surprising that these GOFCs were not dominated by Great Groups of the Mollisol soil order (grassland soils with a thick organic matter rich surface horizon called a mollic epipedon) or associations containing at least one Mollisol Great Group that would indicate more abundant food resources for rangeland grasshoppers.

The modal soil of GOFC 9-13 was P-11, which is the Torriorthents-Argiustolls association (Table 12). Although listed by Young and Singleton (1977) as a Gentle Upland landform soil association, P-11 contains soil series with slopes as steep as 65%. The Argiustolls Great Group are soils of the Mollisol order with an argillic horizon formed under a ustic moisture regime ( a climate that is usually hot in the summer and dry, but wetter than aridic). The ustic moisture regime matches the climatic conditions that have a positive influence on grasshopper populations in Montana and Wyoming (Capinera and Horton 1989).

P-12, the Torriorthents-Argiustolls-Haplustolls association was the modal soil association of GOFC 14 and 15 (Table 12). The Haplustolls Great Group is a simple soil (no other diagnostic horizons besides the mollic epipedon) of the Mollisol order that has a ustic moisture regime.

The common occurrence of the Torriorthents Great Group in all of GOFC modal soil associations could simply reflect the high frequency of this soil type in Wyoming. The Torriorthents Great Group occurs in 36 of the 69 soil associations (Young and Singleton 1977). However, 27 of the 69 soil associations contain Great Groups of the Mollisol order (Young and Singleton 1977). The presence of Torriorthents in areas with ustic moisture regimes indicates that the lack of soil development is due more to topographic position rather than insufficient moisture for weathering to occur. The soil forming factors that Torriorthents form under may also favor grasshopper outbreaks. These factors are time (recent exposure of parent material to weathering), organisms (sparse vegetation with a low rate of incorporation of organic matter), topography (steep slopes where erosion exceeds soil formation), parent material (resistant to weathering), and climate (low precipitation). Of the aforementioned factors climate, organisms, and topography would have the most influence on grasshoppers. Arid and semi-arid climates define the areas of North America where grasshopper outbreaks can occur. Sparse vegetation would keep humidity low at ground level inhibiting the transmission of some grasshopper pathogens and increase soil temperatures for oviposition and embryonic development (Pickford and Riegert 1964, Fisher 1993). Topography also influences soil temperatures with grasshopper eggs laid on aspects more exposed to the sun developing faster and hatching earlier. Early hatching has been found to enhance the fecundity of grasshoppers (Pickford 1966).

Eastern Plains Soils. The cross-tabulation of GOFC and the eastern plains soil associations (including areas of the Eutroboralfs-Haploborolls soil association that covers the mountainous Black Hills) showed that all of the non-zero GOFC modal soil classes were the same as the previous cross-tabulation (Table 12 and Table 13). The GOFC 0 mode was P-13 on the eastern plains instead of BF-10. P-13 covers over 15 percent of the eastern plains (Table 14). Forty-three point five percent of the eastern plains of Wyoming have had < 1 recorded grasshopper outbreak in 31 yearQ„4 h€ ‰DATU€€abitat that have not been infested are emphasized, because large areas of unsuitable grasshopper habitat that have never been infested are eliminated. The modal soil association for GOFC 0, the Haplargids-Paleargids-Torriorthents association (P-13), accounted for most of the area in the northeastern quarter of the state that has never been infested (Fig. 1and Fig. 7). P-4, the mode of GOFC 2-8, covers 23 percent of the eastern plains area (Table 14).

The four highest GOFCs were clearly dominated by their modal soils, but every other GOFC contained 11 or more soil associations (Table 13). Grasshoppers outbreaks can occur on a wide variety of soil types. The data also suggests that chronic infestation areas are actually uncommon on the eastern Wyoming plains. The proportion of eastern plains that have had six or more recorded infestations is only 15.6 percent (Table 13). As more detailed soil data such as the Soil Conservation Service's State Soil Geographic Data Base becomes more user friendly further investigation of how soils are related to grasshopper outbreaks is warranted.

Chi-Square Analysis. The chi-square analyses of the 0, 3, 6, 9 and 12 GOFC frequencies as a function of the six ecological factor data layers (vegetation, mean annual precipitation, elevation, mean annual potential evapotranspiration, landform-soil, and soil associations) were highly significant (P<.0001) in all cases. Thus, grasshopper outbreaks are not spatially random with respect to any of the ecological factors chosen for this analysis. That grasshopper outbreaks are not distributed by random chance is not surprising. All species have environmental parameters that define optimal habitats (Smith 1986). The optimal habitat for rangeland grasshoppers can probably be defined with elements of vegetation, precipitation, evapotranspiration, elevation, and soil.

Outbreak Susceptibility Model. The regression analysis showed the model for predicting an area's susceptibility to grasshopper outbreaks using slope, soil depth and soil erodibility was highly significant (P < .001, r2 = 0.885).

Years Infested = -141.8 + 1.056(Slope) - 0.718(Depth) + 345.4(Erodibility)

The independent variables in this case were purely observational; they were not under human control, as in a typical experiment. However, it is reasonable to assert that these variables, or variables highly correlated to the ones used in the analysis, are important to grasshopper outbreaks.

Pickford (1970) found that egg production and survival has the greatest influence on grasshopper population levels. Soil temperatures and the accumulation of degree days, so important to grasshopper egg survival, are directly related to topography (Pickford 1966, Hewitt 1985). At the present time, I lack the analytical tools necessary to derive slope and aspect from United States Geological Survey Digital Elevation Model (DEM) data for Wyoming. In an attempt to model the effect of slope and aspect on grasshopper populations, the available properties of the soil associations were used as surrogates for the physical factors that might directly influence the hatching date and survival of grasshopper eggs and the development of nymphs. The current results are encouraging and the development of a more precise model is warranted.

Summary. Steeper slopes, as shown by the outbreak susceptibility model and the soil association cross-tabulations, are associated with grasshopper outbreaks. Precisely how steep slopes are related to grasshoppers is a matter of speculation. Local topography largely governs the amount and distribution of precipitation in a mid-continent location like Wyoming (Martner 1986). Weather has been found to play a significant role in the population dynamics of pest grasshoppers (Pickford 1966, Capinera 1987). Topography in a region that often has cold late springs, like Wyoming, may be the key factor in the production of outbreak populations through its influence on the accumulation of degree-days and soil temperatures. Uvarov (1977) presented data showing that the mean daily temperature in western Siberia exceeds 10 0C 2 wks sooner in the spring than in Wyoming. No single factor, like slope, can be used to explain grasshopper population dynamics because of complications arising from heterogeneity of multiple species. Many pest grasshoppers species are very adaptable and can tolerate considerable variation in ecological conditions (e.g., Melanoplus sanguinipes F. has a continental distribution [Pfadt 1988]). Thus, particular environmental conditions appear to release the normal constraints on grasshopper populations (Capinera and Horton 1989).

Climate, time, parent material, organisms, and topography are the five major soil forming factors (Singer & Munns 1987). These five factors have varying potentials as indicators of an area's susceptibility to grasshopper outbreaks. Climate defines an area's susceptibility to grasshopper outbreak (i.e., outbreaks rarely occur in alpine climates). However, climate is a poor indicator of yearly changes in grasshopper populations because yearly weather patterns are highly variable and hard to predict in Wyoming (Martner 1986). The time over which soil develops reflects the long term environmental conditions of an area, but grasshopper outbreaks occur in the short term. The presence of certain parent materials can indicate areas unsuitable for grasshopper outbreaks because of the vegetation it supports. For example, salt laden shale parent material and greasewood (Sarcobatus vermiculatus) dominated vegetation would not be suitable habitat for most grasshopper pest species. A highly visible sessile organism, such ponderosa pine (Pinus ponderosa) has excellent potential as an indicator of areas susceptible to grasshopper outbreaks in Wyoming. Although ponderosa pine has the widest range of any cone-bearing tree in North America (Platt 1972), it is only found in 5.5% of Wyoming. Ponderosa pine occurs on 28 different soil associations state-wide, so soil alone is not the determining factor in its distribution. The unique combination of environmental factors that enable ponderosa pine to grow in Wyoming may be linked with grasshopper outbreaks. Defining the environmental factors that ponderosa pine needs to survive in Wyoming may prove as difficult as defining the conditions necessary for grasshopper outbreaks, but its value as an indicator of grasshopper outbreak susceptibility is promising. An area's topography, especially slope and aspect, could be an indicator of susceptibility to grasshopper outbreaks. Linking climate, ponderosa pine distribution and more detailed topographic information may prove to be as good of an outbreak susceptibility indicator as possible at this time for pest managers.

With advances in remote sensing technology and its analysis with GIS more thorough and timely grasshopper surveys may be conducted. For example, if changes in plant reflectance caused by concentrated grasshopper feeding can be detected with satellite imagery then large areas of rangeland inaccessible to the present roadside survey technique can be monitored. By concentrating on areas that have been identified as being susceptible to grasshopper outbreaks and utilizing field personnel for ground truthing grasshopper outbreak survey could be carried out more efficiently. Research into the actual economic losses caused to ranchers by individual grasshopper species is needed so more accurate assessments of outbreaks can be made. A case in point is Melanoplus occidentalis (Thomas), considered a minor rangeland pest (Pfadt 1988), this species has been found to feed heavily on the reproductive structures in cactus flowers ( Lockwood and Bomar 1992). Is M. occidentalis a beneficial insect or a competitor? Questions of this nature need to be answered with basic ecological research so the application of space age technology to the age old problem of grasshopper outbreaks makes sense.


I wish to thank D. E. Legg and L. C. Munn for their help in developing the regression model. I wish to thank the Geological Survey of Wyoming, The University of Nebraska Press, and the University of Wyoming for allowing me to reproduce and modify figures from their publications.

References Cited

Capinera, J. L. 1987. Population ecology of rangeland grasshoppers. pp. 162-182. In J. L. Capinera [ed.], Integrated pest management on rangeland. Westview Press, Boulder, Colorado.

Capinera, J. L. and D. R. Horton. 1989. Geographic variation in effects of weather on grasshopper infestation. Environmental Entomology. 18: 8-14.

Carter, M. R., T. O. Holtzer, and W. P. Kemp. 1992. Characterizing the Colorado rangeland landscape: implications for grasshopper ecology. Presented at Informal Conferences Orthopterists Society. ESA Annual Meeting 1992, Baltimore, MD.

Cigliano, M. M., W. P. Kemp and T. Kalaris. 1993. Spatiotemporal characteristics of rangeland grasshopper (Orthoptera: Acrididae) regional outbreaks in Montana. J. Orthoptera Res. In Press.

Davis, R. M., M. D. Skold, J. S. Berry and W. P. Kemp. 1992. The economic threshold for grasshopper control on public rangelands. J. of Agricultural and Res. Economics. 17: 56-65.

Dempster, J. P. 1963. The population dynamics of grasshoppers and locusts. Biol. Rev. 38: 490-529.

ERDAS, Inc. 1991. ERDAS Field guide. ERDAS Inc., Atlanta, Georgia.

Fielding, D. J. and M. A. Brusven. 1990. Historical analysis of grasshopper (Orthoptera: Acrididae) population responses to climate in southern Idaho, 1950-1980. Environ. Entomol., 19: 1786-1791.

Fielding, D. J. and M. A. Brusven. 1993. Spatial analysis of grasshopper density and ecological disturbance on southern Idaho rangeland. Agric., Ecosys. and Environ. 43: 31-47.

Fisher, J. R. 1993. Location of egg pods of Aulocara elliotti (Orthoptera: Acrididae) in a field of crested wheatgrass in Montana. J. Kan. Entomol. 65: 416-420.

Hewitt, G. B. 1979. Hatching and development of rangeland grasshoppers in relation to forage growth, temperature, and precipitation. Environ. Entomol. 8: 24-29.

Hewitt, G. B. 1985. Review of factors affecting fecundity, oviposition, and egg survival of grasshoppers in North America. USDA-ARS Rangeland Insect Laboratory. Bozeman, Montana.

Hewitt, G. B., A. C. Wilton, and R. J. Lorenz. 1982. The suitability of legumes for rangeland interseeding and as grasshopper food plants. J. Range Mangt. 35: 653-656.

Isley, F. B. 1938. The relations of Texas acrididae to plants and soils. Ecological Monographs. 8: 553-604.

Joern, A. and S. B. Gaines. 1990. Population dynamics and regulation in grasshoppers. pp. 415-482. In R. F. Chapman and A. Joern [eds.], Biology of grasshoppers. Wiley, New York.

Johnson, D. L. 1988. Spatial analysis of the relationship of grasshopper outbreaks to soil types. pp. 347-359. In L. McDonald, B. Manly, J. Lockwood, and J. Logan [eds.], Estimation and analysis of insect populations, Lecture notes in statistics 55. Springer-Verlag, New York.

Johnson, D. L. and A. Worobec. 1988. Spatial and temporal computer analysis of insects and weather: grasshoppers and rainfall in Alberta. pp. 107-128. In Paths from a Viewpoint: The Wellington Festschrift on Insect Ecology. Mem. Entomol. Soc. Can.

Kemp, W. P. and B. Dennis. 1993. Density dependence in rangeland grasshoppers (Orthoptera: Acrididae). Oecologia. 96: 1-8

King, K. M. 1939. Population studies of soil insects. Ecological Monographs. 9: 270-286.

Liebhold, A. M., R. E. Rossi and W. P. Kemp. 1993. Geostatistics and geographic information systems in applied insect ecology. Annu. Rev. Entomol. 38: 303-327.

Lockwood, J. A. and C. R. Bomar. 1992. Consumption of prickly pear cactus flowers by Melanoplus occidentalis: A coevolutionary Association. Environmental Entomology. 21: 1301-1307.

Lockwood, J. A., W. P. Kemp and J. A. Onsager. 1988. Long-term, large-scale effects of insectidal control on rangeland grasshopper populations (Orthoptera: Acrididae). J. Econ. Entomol. 81: 1258-1264.

Lockwood, D. R. and J. A. Lockwood. 1988. Application of catastrophe theory to population dynamics of rangeland grasshopper. pp. 268-277. In L. McDonald, B. Manly, J. Lockwood, and J. Logan [eds.], Estimation and analysis of insect populations, Lecture notes in statistics 55. Springer-Verlag, New York.

Lockwood, J. A. and S. P. Schell. 1994. Outbreak dynamics of rangeland grasshoppers: eruptive, gradient, both or neither? Proceedings of the 6 th International Orthopterists' Society Meeting. Hilo, Hawaii. In press.

Lockwood, J. A., T. J. McNary, J. C. Larsen and J. Cole. 1993. Distribution atlas for grasshoppers and the Mormon cricket in Wyoming 1988-92. University of Wyoming, Laramie, Wyoming.

Marston, R. A. 1990. Wyoming water atlas. Wyoming Water Development Commission and University of Wyoming, Laramie, Wyoming.

Martner, B. E. 1986. Wyoming climate atlas. University of Nebraska Press, Lincoln, Nebraska.

Martinat, P. J. 1987. The role of climatic variation and weather in forest insect outbreaks. pp. 241-268. In Barbosa, P. & J. C. Schultz [eds.], Insect outbreaks. Academic, New York.

MSUSTAT. 1986. A user's guide to MSUSTAT statistical analysis package. Research & Development Institute Inc. MSU. Bozeman, Montana.

Munn, L. C. 1992. Soil 5120, Genesis, morphology and classification of soil. Class packet. University of Wyoming. Laramie, Wyoming.

Pfadt, R. E. 1949. Food plants as factors in the ecology of the lessor migratory grasshopper Melanoplus mexicanus. Bulletin 290, Agriculture experiment station, University of Wyoming, Laramie.

Pfadt, R. E. 1988. Field guide to common western grasshoppers. Wyoming Agricultural Experiment Station Bulletin 912.

Pickford, R. 1966. The influence of date of oviposition and climatic conditions on hatching of Camnula pellucida (Scudder) (Orthoptera: Acrididae). Can. Entomol. 98: 1145-1159.

Pickford, R. 1970. The effects of climatic factors on egg survival and fecundity in grasshoppers. Proc. Int. Study Con. Current and Future Problems of Acridology, London. 257-260.

Pickford, R. and P. W. Riegert. 1964. The fungus disease caused by Entomophthora grylli Fres., and its effects on grasshopper populations in Saskatchewan in 1963. The Canadian Entomologist. 96: 1158-1166.

Platt, R. 1952. A pocket guide to trees. Pocket Books. New York.

Roberts, S. 1989. Wyoming geomaps. The Geologic Survey of Wyoming, Laramie, Wyoming.

Schell, S. P. 1994. Spatial characteristics of grasshopper outbreaks in Wyoming, 1960-1993. Environmental Entomology. Submitted.

Singer, M. J. & D. N. Munns. 1987. Soils, an introduction. McMillan Publishing, New York.

Smith, R. L. 1986. Elements of ecology, second edition. Harper & Row. New York.

USDA-APHIS. 1987. Rangeland grasshopper cooperative management program: Final environmental impact statement. 1987. FEIS 87-1. USDA-APHIS-PPQ, Hyattsville, Maryland.

Uvarov, B. 1977. Grasshoppers and locusts. Centre for Overseas Pest Research, London.

Young, J. F. & P. C. Singleton. 1977. Wyoming general soil map. Ag. Exp. Stat., Univ. of Wyoming. Research Journal 117. 


Table 1. Cross-tabulation of Vegetation type and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring Vegetation class in each GOFC (Mode), the total number of different Vegetation classes present in each GOFC (Diversity) and percent of each GOFC covered by the Mode.

GOFC Mode Diversity
% of GOFC covered by mode
0 Sagebrush steppe 12
1 Sagebrush steppe 11
2 Mixed-grass prairie 11
3 Mixed-grass prairie 11
4 Mixed-grass prairie 11
5 Mixed-grass prairie 10
6 Mixed-grass prairie 9
7 Mixed-grass prairie 9
8 Mixed-grass prairie 8
9 Ponderosa pine forest 8
10 Ponderosa pine forest 8
11 Ponderosa pine forest 5
12 Ponderosa pine forest 4
13 Ponderosa pine forest 4
14 Mixed-grass prairie 3
15 Mixed-grass prairie 1


Table 2. Vegetation types and the areas they cover in Wyoming.

Vegetation Type % of Wyoming
Alpine Tundra 1.3
Sand Dunes 0.2
Sagebrush Steppe 46.0
Desert Shrublands 4.4
Mixed-Grass Prairie 22.2
Juniper Woodland 2.3
Oak Woodland 0.2
Ponderosa Pine Forest 5.5
Spruce-Fir Forest 4.7
Douglas Fir Forest 3.1
Lodgepole Pine Forest 9.9
Aspen Forest 0.2


Table 3. Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31), the area and percent of Wyoming each covers, and the percentage of each GOFC covered by the Ponderosa pine vegetation class. 

Area (ha)
% of Wyoming
% of GOFC Covered by Ponderosa Pine Vegetation Class


Table 4. Cross-tabulation of Mean Annual Precipitation and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring Mean Annual Precipitation class in each GOFC (Mode), the total number of different Mean Annual Precipitation classes present in each GOFC (Diversity) and percent of each GOFC covered by the Mode. 

GOFC Mode (cm) Diversity % of GOFC covered by mode
0 20-30  6 38
1 30-40  6 48
2 30-40  6 62
3 30-40  6 72
4 30-40  6 73
5 30-40  4 73
6 30-40  4 72
7 30-40  4 69
8 30-40  4 68
9 30-40  4 52
10 30-40  4 48
11 40-50  3 70
12 40-50  3 82
13 40-50  2 85
14 40-50  2 85
15 40-50  1 100


Table 5. Mean annual precipitation classes and the percentage of Wyoming covered by each class. 

Precipitation Classes (cm)
% of Wyoming
Under 20
Over 100 


Table 6. Cross-tabulation of Elevation and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring Elevation class in each GOFC (Mode), the total number of different Elevation classes present in each GOFC (Diversity) and percent of each GOFC covered by the Mode. 

GOFC Mode class (m) Diversity % of GOFC covered by mode
0 2100-2700  7 38
1 1200-1500  7 33
2 1200-1500  7 44
3 1200-1500  7 47
4 1200-1500  6 49
5 1200-1500  6 49
6 1200-1500  49
7 1200-1500  6 47
8 1200-1500  1200-1500  6 39
10 1200-1500  6 43
11 <1200  5 31
12 1500-1800  4 39
13 <1200  4 41
14 <12000  2 88
15 <1200  1 100


Table 7. Elevation categories and the percentage of Wyoming covered by each category.

Elevation Category (m) 
% of Wyoming
Under 1200 
Over 3350 


Table 8. Cross-tabulation of Mean Annual Potential Evapotranspiration (MAPE) class and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring MAPE classes in each GOFC (Mode), the total number of different MAPE classes present in each GOFC (Diversity) and percent of each GOFC covered by the Mode. 

GOFC Mode 

MAPE (cm)

Diversity % of GOFC covered by mode
0 48-53  6 33
1 53-58  6 39
2 58-63  6 50
3 58-63  6 49
4 58-63  5 47
5 58-63 4 43
6 58-63  4 42
7 58-63  4 47
8 58-63  4 44
9 58-63  4 41
10 58-63  4 44
11 53-58  4 56
12 53-58  3 57
13 53-58  3 66
14 53-58  2 81
15 53-58  1 100


Table 9. Mean Annual Potential Evapotranspiration (MAPE) classes and the percentage of Wyoming covered by each class.

MAPE (cm) 
% of Wyoming
Under 43 
Over 63


Table 10. Landform-soil classes and the percentage of Wyoming covered by each class.

% of Wyoming
Mountains, Moist Year Round
Mountains, Moist in Summer
Mountain, Dry
River Valleys and Sand Hills
Steep Uplands
Gentle Uplands


Table 11. Results of the cross-tabulation of Landform-soil class and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring Landform-soil class in each GOFC (Mode), the total number of different Landform-soil classes present in each GOFC (Diversity) and percent of each GOFC covered by the Mode.

GOFC Mode  Diversity % of GOFC covered by mode
0 Basins 6 49
1 Basins 6 30
2 Steep uplands 6 30
3 Steep uplands 6 33
4 Steep uplands 6 39
5 Steep uplands 6 38
6 Steep uplands 6 40
7 Steep uplands 6 40
8 Steep uplands 6 37
9 Gentle uplands 6 35
10 Gentle uplands 6 41
11 Gentle uplands 6 65
12 Gentle uplands 3 82
13 Gentle uplands 3 83
14 Gentle uplands 2 88
15 Gentle uplands 1 100


Table 12. Cross-tabulation of State Soil associations and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring State Soil associations in each GOFC (Mode), the total number of different State Soil associations present in each GOFC (Diversity) and percent of each GOFC covered by the Mode. 

GOFC Map code of Mode  Diversity % of GOFC covered by mode
0 BF-10a 65
1 P-13b 59
2 P-4c 50
3 P-4 44
4 P-4 37
5 P-4 36
6 P-4 32
7 P-4 26
8 P-4 23
9 P-11d 21
10 P-11 19
11 P-11 15
12 P-11 6
13 P-11 5
14 P-12e 3
15 P-12 1

a Torriorthents-Haplargids-Natrargids soil association

b Haplargids-Paleargids-Torriorthents soil association

c Torriorthents, shallow soil association

d Torriorthent-Argiustolls soil association

e Torriorthents-Argiustolls-Haplustolls soil assocation


Table 13. Cross-tabulation of Soil associations of the Eastern Plains and Grasshopper Outbreak Frequency Classes (GOFC) (the number of years a grid cell has been infested out of 31) showing the most commonly occurring Soil associations of the Eastern Plains in each GOFC (Mode), the total number of different Soil associations present in each GOFC (Diversity) and percent of each GOFC covered by the Mode. 

GOFC Mode  Diversity % of GOFC covered by mode % of eastern plains covered by each GOFC
0 P-13a 18 27.4 23.9
1 P-13 19 20.9 19.6
2 P-4b 20 20.6 15.2
3 P-4 19 25.5 10.4
4 P-4 18 28.9 8.5
5 P-4 18 28.7 7.0
6 P-4 17 27.8 5.8
7 P-4 17 28.9 4.6
8 P-4 15 26.3 2.5
9 P-11c 15 24.3 1.2
10 P-11 13 32.5 0.6
11 P-11 11 59.9 0.3
12 P-11 4 81.7 0.1
13 P-11 3 66.7 0.1
14 P-12d 2 60.9 <0.1
15 P-12 1 100.0  <0.1

a Haplargids-Paleargids-Torriorthents soil association
b Torriorthents, shallow soil association
c Torriorthent-Argiustolls soil association
d Torriorthents-Argiustolls-Haplustolls soil assocation

Table 14. Soil Associations of Wyoming's Eastern Plains and the percentage of the Eastern Plains covered by each. 

Soil Associations
% of Eastern Plains Covered
Eutroboralfs-Haploborolls (MF-1) 
Torripsamments (P-1)
Torrifluvents-Haplargids & Torrifluvents-Haplargids-Torriorthents (P-2 & P-3)
Torriorthents, shallow (P-4)
Torriorthents-Haplargids (P-5)
Torriorthents-Torriorthents, shallow (P-6)
Torriorthents-Torriorthents, shallow-Rock outcrop (P-7)
Torriorthents (P-8)
Torriorthents, fine, acid (P-10)
Torriorthents-Argiustolls (P-11)
Torriorthents-Argiustolls-Haplustolls (P-12)
Haplargids-Paleargids-Torriorthents (P-13)
Haplargids-Torriorthents (P-14)
Torriorthents-Haplargids-Camborthids (P-15)
Argiustolls-Haplustolls (P-16)
Haplustolls-Argiustolls-(Torriorthents or Torripsamments) (P-17 & P-18)
Haplargids (P-19)
Argiustolls-Haplustolls-Torriorthents (P-20)
Torriorthents, fine-Torrifluvents (P-21)