UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The relationship of grazing to orthopteran diversity in the intermontane grasslands of the South Okanagan,… Griesdale, Peggy Liu 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-0456.pdf [ 4.3MB ]
Metadata
JSON: 831-1.0092083.json
JSON-LD: 831-1.0092083-ld.json
RDF/XML (Pretty): 831-1.0092083-rdf.xml
RDF/JSON: 831-1.0092083-rdf.json
Turtle: 831-1.0092083-turtle.txt
N-Triples: 831-1.0092083-rdf-ntriples.txt
Original Record: 831-1.0092083-source.json
Full Text
831-1.0092083-fulltext.txt
Citation
831-1.0092083.ris

Full Text

THE RELATIONSHIP OF GRAZING TO ORTHOPTERAN DIVERSITY IN THE INTERMONTANE GRASSLANDS OF THE SOUTH O K A N A G A N , BRITISH C O L U M B I A by P E G G Y LIU GRIESDALE B. Sc., The University of British Columbia, 1998 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (ZOOLOGY) THE UNIVERSITY OF BRITISH C O L U M B I A September 2005 © Peggy Liu Griesdale, 2005 ABSTRACT The antelope-brush shrub-steppe of the South Okanagan is small in size yet home to many of the unique and endangered flora and fauna of British Columbia and Canada. More insect species are found in this ecosystem than other grassland ecosystems. Antelope-brush ecosystems are dominated by bunchgrasses, antelope-brush, and a well-developed cryptogam crust, owing to the hot and dry summers of the South Okanagan. Urban and vineyard development are the most immediate threat to this fragile ecosystem, followed by unmanaged livestock grazing. Livestock grazing exposes soil, stunts plant growth, and fragments the cryptogam crust. Less than 9% of the antelope-brush ecosystem is relatively undisturbed and only two small ecological reserves exist. Orthopterans are the most important invertebrate herbivore in North American grasslands and are one of the main biotic influences on grasslands. While Orthopterans assist with biomass turnover and nutrient cycling processes of ecosystem functioning, they may add to the effects of livestock overgrazing. Numerous studies have shown contradictory results of the relationship between grasshopper abundances and grazing pressures. As part of a larger study of the biodiversity and impact of grazing on this threatened ecosystem, this study was conducted to determine how livestock grazing in the intermontane grasslands of the South Okanagan of British Columbia influenced the abundance and species assemblage of Orthopterans. Orthopterans were collected with pitfall traps in ten locations in the antelope-brush ecosystem of the South Okanagan over two years. The study sites were of three different grazing levels: 1) non-grazed; 2) moderately grazed; and 3) heavily grazed. Vegetation data were collected with Daubenmire plots at each site. Twenty-four orthopteran species were captured (seventeen grasshopper species and seven cricket species). A l l seventeen grasshopper species were previously known to occur in British Columbia, but the taxonomies of four of the cricket species are currently being revised. Grazing did not affect orthopteran species abundance or diversity. Regression analyses showed that the number of orthopteran species and Shannon-Wiener Index values increased with increasing bare soil. The effects of grazing on the vegetation community and structure, and its corresponding effects on the orthopteran species assemblage, are discussed. iii TABLE OF CONTENTS A B S T R A C T ii T A B L E OF CONTENTS iv LIST OF T A B L E S ." vi LIST OF FIGURES vii A C K N O W L E D G E M E N T S ix 1 INTRODUCTION 1 1.1 The South Okanagan Grassland Conservation Research Project 1 1.2 Grasslands 1 1.2.1 North American Grasslands 1 1.2.2 British Columbian Grasslands 2 1.2.3 Grassland Characteristics 3 1.2.4 Climate 3 1.2.5 Bunchgrass Subzones 4 1.3 Antelope-brush Ecosystem 4 1.3.1 Characteristics and Ecological Significance 4 1.4 Grazing and Grasslands 6 1.4.1 History of Grassland Disturbances in British Columbia 6 1.4.2 Cryptogam Crust 7 1.4.3 Overgrazed Sites 7 1.5 Grasshoppers and Grazing 9 1.6 Objectives 11 2 M A T E R I A L S A N D METHODS 12 2.1 Study Location 12 2.2 Overall Study Structure 14 2.3 Grazing Categorization and History 15 2.4 Grazing History by Site 16 2.5 Arthropod Collection 18 2.5.1 Sampling Method 18 2.5.2 Pitfall Trap Design 20 2.5.3 Emptying Trap Contents 21 2.6 Data Processing 21 2.7 Statistical Analysis 23 2.7.1 Data Grouping 23 2.7.2 Data Analysis 24 2.7.3 Biodiversity Measures 24 2.7.3.1 Species Diversity and Heterogeneity Measures 24 2.7.3.2 Analysis of Variance 26 2.7.3.3 Taxonomic Diversity Measures 26 2.7.4 Cluster Analysis and non-metric Multidimensional Scaling 29 2.7.5 Vegetation Data 31 2.7.6 Correlation and Regression Analysis 32 3 RESULTS 33 3.1 Descriptive Statistics and General Observations 33 3.2 Orthopteran Abundance 41 iv 3.3 Site Groupings 44 3.4 Trap Disturbance 46 3.5 Biodiversity Measures 47 3.5.1 Taxonomic Distinctness Measures 48 3.6 Cluster Analysis and non-metric Multidimensional Scaling 51 3.6.1 Vegetation Overlays 56 3.7 Correlation Analysis 63 3.8 Regression Analysis 64 4 DISCUSSION 67 4.1 Trapping Method 67 4.1.1 Sweep Netting 67 4.1.2 Pitfall Trapping 69 4.2 Orthopteran Study in the South Okanagan 71 4.2.1 Orthopteran Species Descriptions 71 4.3 Descriptive Statistics 72 4.4 Biodiversity Measures 74 4.5 Cluster Analysis and non-metric Multidimensional Scaling 75 4.6 Vegetation Overlays, Correlation Analyses and Regression Analyses 75 4.7 Effects of Grazing on Grasshoppers in Different Grassland Types 77 5 CONCLUSION. . 83 LITERATURE CITED 84 v LIST OF TABLES Table 1. The location, latitude, longitude, unofficial name, site label, site label used for vegetation survey, and the grazing category of the ten grassland study sites in the South Okanagan, BC 14 Table 2. The grazing history, definition of grazing history, and grazing history category of the ten grassland study sites in the South Okanagan, B C 16 Table 3. Dates of pitfall trap collections from ten grassland study sites in the South Okanagan, BC 20 Table 4. List of orthopteran species found at ten study sites in the South Okanagan, BC 35 Table 5. Species occurrence at ten grassland study sites in the South Okanagan, BC 37 Table 6. Total catch over the entire collection period from ten grassland study sites in the South Okanagan, BC (missing traps are not accounted for) 39 Table 7. List of orthopteran species captured at all three grazing categories from the ten grassland study sites in the South Okanagan, BC 45 Table 8. The calculated biodiversity indices for each of the ten grassland study sites and the biodiversity indices averaged for each grazing regime from the South Okanagan, BC 48 fable 9. List of percent bare ground, percent plant cover, and plant species richness for the ten grassland study sites in the South Okanagan, BC (raw data provided by Dr. P. Krannitz) 57 Table 10. Correlations between vegetation data and orthopteran data from the ten grassland study sites in the South Okanagan, BC 64 vi LIST OF FIGURES Fig. 1. Map of the ten grassland study sites in the South Okanagan, BC and neighbouring Washington State 13 Fig. 2. Diagram of the layout of pitfall traps at each of the ten grassland study sites in the South Okanagan, BC 19 Fig. 3. The total number of orthopteran species and the total number of orthopteran specimens collected per month from ten grassland study sites in the South Okanagan, BC 41 Fig. 4. The total number of Orthopterans (divided into nymphs and adults, grasshoppers and crickets) summed across all ten grassland study sites per month, captured in the South Okanagan, B C 42 Fig. 5. Average number of Orthoptera per pitfall trap per site (with standard error bars) captured from ten grassland study sites in the South Okanagan, BC, for all collection periods. The number of orthopteran species per site is listed above the standard error bars 43 Fig. 6. Mean number of Orthoptera per grazing category, captured from ten grassland study sites in the South Okanagan, BC (with 95% confidence intervals). Non-grazed (O, V , Z); moderately grazed (S, T, X , Y); and heavily grazed (P, U , W). The number of orthopteran species per grazing category is listed above the confidence intervals 44 Fig. 7. Venn diagram representing grasshopper species presence and absence according to grazing category and the overlap between grazing categories from ten grassland study sites in the South Okanagan, BC 46 Fig. 8. Simulated means (dashed line), 95% probability funnels (continuous line), and measured average taxonomic distinctness ( A + ) values for each of the ten grassland study sites in the South Okanagan, BC, plotted against the number of species for 1000 random simulations 49 Fig. 9. Simulated means (dashed line), 95% probability funnels (continuous line), and measured variation in taxonomic distinctness (A + ) values for each of the ten grassland study sites in the South Okanagan, BC, plotted against the number of species for 1000 random simulations 50 Fig. 10. Fitted 95% probability contours of the joint distribution of A + and A + , from 1000 random simulations (for species sublist sizes = 5, 10, and 15), calculated for each of the ten grassland study sites in the South Okanagan, BC 51 Fig. 11. Dendrogram for hierarchical clustering of the ten grassland study sites in the South Okanagan, B C , using group-average linking of Bray-Curtis similarities calculated on square root-transformed data 52 vn Fig. 12. Dendrogram for hierarchical clustering of the ten grassland study sites in the South Okanagan, B C , using group-average linking of Bray-Curtis similarities calculated on presence/absence data : ; 53 Fig. 13. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC 54 Fig. 14. MDS of Bray-Curtis similarities from presence/absence data from the ten grassland study sites in the South Okanagan, BC 55 Fig. 15. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent bare soil (arcsine transformed) 58 Fig. 16. MDS of Bray-Curtis similarities from species presence/absence data from the ten grassland study sites in the South Okanagan, B C , with superimposed circles of increasing size with increasing percent bare soil (arcsine transformed) 59 Fig. 17. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent plant cover (arcsine transformed) 60 Fig. 18. M D S of Bray-Curtis similarities from species presence/absence data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent plant cover (arcsine transformed) 61 Fig. 19. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC,with superimposed circles of increasing size with increasing plant species richness 62 Fig. 20. MDS of Bray-Curtis similarities from species presence/absence data from the ten grassland study sites in the South Okanagan, BC,with superimposed circles of increasing size with increasing plant species richness 63 Fig. 21. Fitted line plot of orthopteran species richness and percent bare soil (arcsine transformed) (Fjg = 15.518,/? = 0.004, r 2 - 66%) from the ten grassland study sites in the South Okanagan, BC 65 Fig. 22. Fitted line plot of Shannon-Wiener Index values and percent bare soil (arcsine transformed) (Fi;g = 9.51,/? = 0.015, r 2 = 54.3%) from the ten grassland study sites in the South Okanagan, BC 66 viii ACKNOWLEDGEMENTS I would like to thank my committee members: 1) Dr. Geoffrey Scudder, my supervisor, for giving me the opportunity to pursue my interest in entomology and for his confidence in me; 2) Dr. Judy Myers, for her support and extensive feedback through my writing process; and 3) Dr. Dolph Schluter, for his feedback regarding my statistical analyses. Thank you to my friends from the Scudder lab for all your support and encouragement: Susanna Carson, Jenny Heron, Jeff Jarrett, Suzie Lavallee, Launi Lucas, and Karen Needham. I would like to thank the following funding sources for this project: Canadian Wildlife Service, the World Wildlife Fund, Endangered Species Recovery Fund, the National Science and Engineering Research Council of Canada, the Vancouver Foundation, and the Habitat Conservation Trust Fund. Thank you to Dr. Pam Krannitz for providing me the vegetation data for this project. I would like to thank the following Orthopterists for verifying my species identifications: Dr. Vernon R. Vickery (Lyman Entomological Museum and Research Laboratory), Dr. Theodore Cohn (University of Michigan), and Dr. Dan Johnson (University of Lethbridge). Thank you to my good friend Gayle Pelman whose assistance with proofreading and formatting this thesis were invaluable to me throughout the writing process. Lastly, I would like to thank my parents and my husband, Donald, for all their support throughout the years. 1 INTRODUCTION 1.1 The South Okanagan Grassland Conservation Research Project This study of the relationship between grazing and Orthoptera took place in the grasslands of the Southern Interior of British Columbia and was part of the South Okanagan Grassland Conservation Research Project. This was a large, multi-disciplinary project with the overall goal of identifying strategies for land managers to use to improve degraded lands for the welfare of rare wildlife (Krannitz 1994). Three of the objectives of this project were to 1) identify which wildlife increased or decreased with livestock grazing in the antelope-brush shrub-steppe; 2) determine which wildlife species are affected by grazing and disturbance; and 3) examine the effects of grazing and disturbance on the invasion of exotic species. 1.2 Grasslands 1.2.1 North American Grasslands Grasslands are plant communities dominated by graminoids (grasses and sedges) and are void of trees, or communities where herbaceous plants have a continuous cover (Brink 1982). Throughout the world, grasslands are known by many different names, including campos, pampas, plains, prairie, steppe, veld, and savanna. North America has two basic types of grasslands: meadow and prairie. Prairies (also known as steppe) are considered climax communities whereas meadows are usually in an intermediate successional stage (in transition to a forest climax community) (Titlyanova et al. 1990). Three areas of grasslands occur in Canada: 1) short-grass, mixed-grass, and tall-grass prairies of Alberta, Saskatchewan, and Manitoba; 2) Aspen Parkland grasslands of Alberta and British Columbia (BC); and 3) intermontane grasslands of south and central BC (Carder 1970). 1 Canadian grasslands are mainly of the short-grass, mixed-grass, and tall-grass prairie types. Annual precipitation is the main determinant of the heights of these grasses; the more rainfall, the taller the grass. Areas of mixed-grass prairies have variable precipitation that allows both the short and tall grasses to coexist. During years of drought, the shorter grasses are more successful, while during years with greater rainfall, the taller grasses are more successful (Coupland 1992). The Aspen Parkland grasslands exist at the grassland-forest boundary and these occur on the edges of the Great Plains prairies to the northeast, north, and northwest. The intermontane grasslands lie between the Coast and Cascade Mountains on the west and the Rocky Mountains to the east. These grasslands extend south through the United States to Northern Mexico, still bounded by the Coast Mountains (Cascade-Sierra Nevada Mountain chain) on the west and the Rocky Mountains to the east (Carder 1970; Nicholson et al. 1991; Pitt and Hooper 1994). This grassland corridor is also referred to as the Great Basin or Intermountain West. They are characterized by bunchgrass and sometimes called palouse grasslands (an old term used more frequently in the United States). 1.2.2 British Columbian Grasslands The intermontane grasslands of the South Okanagan of BC are in the Bunchgrass biogeoclimatic zone (BG), the smallest of the fourteen biogeoclimatic zones in B C , covering less than 1 % of the southern interior valleys of the province (Alldritt-McDowell and Coupe 1998). Biogeoclimatic zones, or ecological zones, are large geographic areas classified according to similar climate, soil, and vegetation. 2 1.2.3 Grassland Characteristics Grasslands of the B G consist of bunchgrasses that are widely spaced with a well-developed cryptogam crust on the bare soil between vascular plants. This cryptogam crust (also known as grassland microphytes) is a thin, fragile layer of mosses, lichens, algae, and bacteria (Alldritt-McDowell and Coupe 1998). Climax communities in the B G Zone typically have a cover of 60% bunchgrasses, 10-15% shrubs, and 25-35% cryptogams. The climax grass is usually Agropyron spicatum (Pursh.) (bluebunch wheatgrass) (Nicholson et al. 1991). 1.2.4 Climate The B G sits in the rainshadow of the Coast and Cascade Mountains; therefore, this zone receives little rainfall. Summers are very hot and dry while winters are moderately cold with little snowfall (Nicholson et al. 1991). The B G receives approximately 300 mm of rain a year (Schluter et al. 1995), mainly during early winter and the month of June. As spring is usually dry, and summer rain evaporates before getting into the soil, plant growth depends on winter moisture. Winter precipitation is able to reach deep into the soil and recharge it, as evaporation and transpiration is minimal during the winter (Nicholson et al. 1991). The Rocky Mountains separate the B G from the weather fronts of the Great Plains (coming from the East Coast). The South Okanagan grasslands differ from the Great Plains, where most of the rainfall occurs during the summer (Schluter et al. 1995; West 1983) and this allows the sod-forming grasses to grow (Nicholson et al. 1991; Schluter et al. 1995). 3 1.2.5 Bunchgrass Subzones The B G is divided into two subzones: the Very Dry Hot subzone (BGxh) and the Very Dry Warm subzone (BGxw). The BGxh occurs from the valley bottoms to approximately 700 m while the BGxw occurs from approximately 700 m to 1000 m in elevation. This study was located in the BGxh (Nicholson et al. 1991). The BGxh is known to have widely spaced clumps of bluebunch wheatgrass, Artemisia tridentata Nult. (big sagebrush), and many lichen species that encrust on the soil surface. The soil type is usually silty clay loam to sandy loam Brown Chernozems (comes from glacial deposits) (Nicholson et al. 1991). Drier sites within the BGxh have a sandy soil that is low in nutrients. They are often dominated by Hesperostipa comata Trin. and Rupr. (needle-and-thread grass), Sporobolus cryptandrus (Torr.) Gray (sand dropseed), and Purshia tridentata (Pursh.) D.C. (antelope-brush) (Nicholson et al. 1991). 1.3 Antelope-brush Ecosystem 1.3.1 Character ist ics and Ecological Signif icance Many different ecosystems exist within the BG. This study was undertaken within the antelope-brush ecosystem or antelope-brush shrub-steppe. This ecosystem occurs mainly south of Penticton, BC to just over the Washington State border and covers sandy or gravelly soils at lower elevations. The antelope-brush ecosystem of British Columbia is very small (less than 5000 hectares) and home to many unique or endangered flora and fauna of both British Columbia and 4 Canada. This ecosystem is considered one of the four most endangered ecosystems in Canada because of its small area and serious threats to its integrity (Schluter et al. 1995). The antelope-brush ecosystem is the northern limit of this landscape corridor (Great Basin or Intermountain West) that extends south to northern Mexico. This ecosystem is particularly significant as genetic diversity is greatest at the boundaries of populations (Wilson 1992). While the antelope-brush ecosystem is an extension of the grasslands located immediately south in Washington, they are not the same, owing to differences in precipitation (Tisdale 1947) and place of origin (Daubenmire 1978). While the grasslands of the South Okanagan and the Great Basin both receive rainfall primarily during the winter, the South Okanagan grasslands also usually receives significant rainfall in late spring (June). Additionally, the South Okanagan grasslands have a larger percentage of plant species that are of northern origin as compared to southern origin, as a result of glacial coverage (Daubenmire, 1978). The Cordilleran ice sheet covered the South Okanagan during the Pleistocene and the southern limit of the ice sheet reached just south of the current Washington border. The soil of the South Okanagan formed mainly from the remaining glacial till, and antelope-brush thrives on this sandy soil (Tisdale 1947). As of 1995, approximately 9% of the antelope-brush ecosystem remained relatively undisturbed (Schluter et al. 1995), and only two small ecological reserves existed (4 ha and 101 ha in size) (Scudder 2000). Twenty-two percent of the endangered and threatened vertebrates of British Columbia occur in this ecosystem (Schluter et al. 1995) and more than 250 potentially 5 rare and endangered invertebrate species reside here. In addition, more species of insects are found in the antelope-brush ecosystem than other grassland ecosystems (Schluter et al. 1995). Currently, the most immediate threats to the antelope-brush ecosystem are vineyard and urban development, unmanaged livestock grazing, cultivation, and forest encroachment. More than 60% of this ecosystem has been destroyed, due to conversion to agriculture use or urban development (Dyer and Lea 2003). Livestock grazing exposes soil and stunts bunchgrass plants, while hooves and equipment easily fragment the cryptogam crust. This ecosystem is poorly understood (Schluter et al. 1995) and very fragile (Nicholson et al. 1991). Range re-seeding and off-road recreation have altered much of the remaining "undeveloped" grassland, and hydro-electric power projects also threaten the remaining grasslands (Pitt and Hooper 1994). 1.4 Grazing and Grasslands 1.4.1 History of Grassland Disturbances in British Columbia Livestock grazing in BC began after the discovery of gold in the Fraser River in 1858. By the early 1900's, overgrazing (overstocking and grazing over an entire season by cattle, sheep, and horses) had occurred, and weedy and introduced plant species increased on BC grasslands (Pitt and Hooper 1994; Tisdale 1947). Since the 1940's, grassland range conditions have improved due to better management and government-imposed decreases in livestock on lands (Pitt and Hooper 1994). For example, cattle are no longer dependent solely on grasslands, as grazing now occurs to a great extent in forests. 6 Some experts consider livestock grazing as the most widespread influence on native ecosystems of western North America (Crumpacker 1984; Wagner 1978), but deforestation receives much more public attention (Fleischner 1994). In the western United States, grazing is allowed to occur even in protected areas such as national parks and wildlife refuges (Fleischner 1994). In BC, "ancient" grasslands are much more endangered than old-growth forests. Additionally, the remaining "ancient" grasslands cover a much smaller area than old-growth forests in BC (Pitt and Hooper 1994). 1.4.2 Cryptogam Crust Recently, interest has been taken in the cryptogam crust, the organic crust of lichens, bryophytes, and cyanobacteria that grows over soil with no other vegetation growth (Pitt and Hooper 1994). Under relatively pristine conditions, the cryptogam crust is well developed between the dominant vascular plants of dry grasslands (Anderson and Rushforth 1977; Daubenmire 1978). Cryptogams help maintain the health of the antelope-brush ecosystem by stabilizing the soil (Anderson et al. 1982), improving soil moisture retention (Brotherson and Rushforth 1983), and enriching the soil with nitrogen through nitrogen fixation by cyanobacteria (Cameron and Fuller 1960). Additionally, cryptogams may help prevent germination and establishment of weedy plant species (Mack and Thompson 1982). 1.4.3 Overgrazed Sites Overgrazing usually increases abundance of unpalatable or weedy vascular plants, to the detriment of bunchgrasses, especially bluebunch wheatgrass. Decreases in density and diversity of native plant species due to livestock grazing has been widely observed in western ecosystems (Fleischner 1994). South Okanagan grasslands in poor condition often have increased 7 abundances of big sagebrush, needle-and-thread grass, Antennaria dimorpha (Nutt.) Torr. Gray (low pussytoes), Opuntia fragUis (Nutt.) Haworth (brittle prickly-pear cactus), and Bromus tectorum L . (cheatgrass) (Nicholson et al. 1991). The recovery time needed for grasslands of poor range condition to become excellent is approximately 20-40 years of complete rest for bluebunch wheatgrass-dominated communities (McLean and Tisdale 1972). Cattle prefer to forage on large bunchgrasses over smaller grasses and forbs. Bunchgrass communities grazed by cattle usually have lower proportions of the dominant bunchgrasses. Additionally, grazed areas often have increased amounts of smaller perennial grasses, annual grasses, and forbs, which indicate that the range condition is decreasing (McLean and Marchand 1968). In addition to the effects of foraging, trampling by livestock has significant effects on grassland health. Livestock trampling in particular has damaged or destroyed much of the cryptogam crust in areas of heavy grazing. The cryptogam crust recovers at a very slow rate (Alldritt-McDowell and Coupe 1998). Although a small amount of grazing by wild ungulates has occurred naturally in the evolution of the antelope-brush ecosystem, these grasslands did not evolve with the influence of large herds of bovid grazers as occurred with the short-grass, mixed-grass, and tall-grass prairies (Mack and Thompson 1982). As the B G did not evolve with bovid grazers, bunchgrass grasslands are particularly susceptible to the negative effects of livestock grazing. The following vertebrates naturally graze or browse in the BG: white-tailed deer (Odocoileus virginianus (Zimm.)), mule deer (Odocoileus hemionus (Rafinesque)), Nuttall's cottontain 8 rabbit (Sylrilagus nuttallii (Bachman)), and California bighorn sheep (Ovis canadensis californiana Douglas) (Pitt and Hooper 1994). 1.5 Grasshoppers and Grazing As the plant community changes with grazing, insect populations also change owing to alterations to their microclimate, living space, and food resources (Breymeyer and van Dyne 1980). Grasshoppers have been identified as the most important invertebrate herbivore in North American grasslands (Hewitt 1977), both in terms of biomass and plant biomass consumption (Baldi and Kisbenedek 1997; Fielding and Brusven 1995b; Mitchell and Pfadt 1974). They have been found to comprise more than half of the arthropod biomass on disturbed lands (Urbanek 1982). As primary consumers, grasshoppers serve an important role in ecosystems through biomass and nutrient cycling (Fielding and Brusven 1995b; Parmenter et al. 1991). Additionally, grasshoppers are a major food source for other grassland species, especially reptiles, birds, and small mammals (Mitchell and Pfadt 1974; Smith 1940). Besides their positive influences on ecosystem functioning, grasshoppers are also capable of causing significant economic losses and damages to crops and rangelands (Hewitt 1977) and can periodically assume epidemic proportions (Tisdale 1947). Some studies have shown that depleted or overgrazed rangelands are particularly favorable habitats for grasshoppers (Hewitt 1977; Hewitt and Rees 1974). Other studies have suggested that grasshopper outbreaks do not occur on well managed grasslands (Pepper 1955) or grasslands that are maintained in climax condition (Smith 1940). Nerney (1958) showed that grasshopper abundance can be increased by poor rangeland management practices, especially under the proliferation of weeds. Once grasshoppers reach outbreak densities, they compete with livestock and other wildlife for food 9 (Fielding and Brusven 1995b), which can then further compound the effects of overgrazing (Tisdale 1947). Owing to the ability of grasshoppers to cause significant economic loss, there is a great deal of interest amongst rangeland managers to identify how grasshopper abundances respond to livestock grazing. While many invertebrates decrease with overgrazing (Hutchinson and King 1970; Morris 1967; Smith 1940), grasshopper abundances have been observed to increase with overgrazing (Smith 1940). Numerous studies have shown that livestock grazing influences grasshopper density and/or grasshopper species composition, but the nature of this relationship is unclear. Most of these studies have taken place on short-grass, mixed-grass, or tall-grass prairie. Some studies have shown that grasshopper abundance increases under heavy grazing pressures (Campbell et al. 1974; Coyner 1938; Holmes et al. 1979; Joern 1982; Joern 2004; Knutson and Campbell 1974; Smith 1940; Weese 1939; Willms et al. 1985), while others have shown that grasshopper abundance decreases under heavy grazing pressures (Capinera and Sechrist 1982; Fielding and Brusven 1995a; Jepson-Innes and Bock 1989; Knopf 1957; Welch et al. 1991). Many studies have also linked changes in the grasshopper community to those of the plant community (Anderson 1964; Campbell et al. 1974; Capinera and Sechrist 1982; Coyner 1938; Fielding and Brusven 1993a; Joern 1979; Kemp 1992; Knutson and Campbell 1974; Quinn and Walgenbach 1990; Scoggan and Brusven 1973; Smith 1940; Weese 1939). As the effects of grazing on the grasshopper community in the intermontane grasslands of B C have not been examined, this study examines this relationship and compares it to the results of studies on other grassland types. 10 1.6 Objectives The objectives of this study were to: 1) identify the orthopteran species composition in the antelope-brush ecosystem of the South Okanagan; 2) analyze the influences of grazing on orthopteran community attributes; and 3) investigate the influence of the plant community on the orthopteran community. The hypotheses of this study predict that: 1) grazing changes orthopteran abundance; 2) grazing changes orthopteran species diversity; and 3) changes in orthopteran abundance and species diversity will be related to changes in the plant community. 11 2 MATERIALS AND METHODS 2.1 Study Location This study of Orthoptera was part of the South Okanagan Grassland Conservation Research Project located in the South Okanagan valley of British Columbia and neighbouring Washington State. Ten sites that differed in their grazing regimes were used in this study, located between Vaseux Lake in the north and Oroville, Washington, in the south (Fig. 1). A l l sites were located on the east side of the Okanagan valley, ranged in altitude of between 300-500 m above sea level, were dominated by antelope-brush, had a southwest aspect, and had similar slopes. Table 1 lists the specific location of each study site. 12 i j / "^enticton t i i I Brtsh CokJTtia 'i 1 Idaho "^i ! _i L .Oliver 4 (975 \ I N A British Columbia Washington .^Osoyoos Fig. 1 . Map of the Washington State. ten grassland study sites in the South Okanagan, BC and neighbouring Table 1. The location, latitude, longitude, unofficial name, site label, site label used for vegetation survey1, and the grazing category of the ten grassland study sites in the South Okanagan, BC. Site Label Other Label1 Unofficial Name Location Latitude Longitude Non-grazed sites O W A East Osoyoos Lake Oroville, W A 48° 58' 119°25' Non-grazed sites V B W Brights Winery Inkameep Reserve, Osoyoos, BC 49°13' 119°32' Non-grazed sites z K B Kennedy Bench Vaseux Creek, BC 49° 16' 119°30' Moderately grazed sites s BO Black Sage Road Osoyoos, BC 49° 07' 119°33' Moderately grazed sites T OS Inkameep Inkameep Reserve, Osoyoos, BC 49° 09' 119°32' Moderately grazed sites X K L Kennedy Flats Vaseux Creek, BC 49°15' 119°31' Moderately grazed sites Y CWS Canadian Wildlife Service Bench Vaseux Creek, BC 49°16' 119°30' Heavily grazed sites P ELO East Osoyoos Lake Inkameep Reserve, Osoyoos, BC 49° 04' 119°29' Heavily grazed sites U WT Watertower Inkameep Reserve, Oliver, BC 49° 10' 119°31' Heavily grazed sites w R G Near Mud Lake Osoyoos, BC 49°12' 119°31' 2.2 Overall Study Structure The South Okanagan Research Project was a large, collaborative project that involved experts from several different fields. This examination of Orthoptera was a small portion of the project. The main objective of this research project was to examine the effects of livestock grazing on the flora and fauna of the antelope-brush ecosystem. This project's multi-'Dr . P. Krannitz used these site labels in her study o f the plants at these sites (Karnnitz 1994). 14 disciplinary team came from the University of British Columbia, the Federal and Provincial governments, and the private sector (including First Nations). The experts and their area of study were: • Dr. Pam Krannitz (Research Scientist, Canadian Wildlife Service), examined the flora of the area (grasses, shrubs, and herbaceous plants); • Dr. Walt Klenner (Wildlife Ecologist, Ministry of Forest), studied the small mammals; • Dr. Geoff Scudder (University of British Columbia, Department of Zoology), investigated the ground-dwelling arthropods; • Dr. Rhonda Mill ikin (Biologist, Canadian Wildlife Service), studied the migrating and nesting birds; • Lynne Atwood (Genoa Environmental Consulting, Smithers, BC) examined the soil and cryptogamic layer; • Chris Shewchuk (Contractor) investigated the snake population. In addition to these experts, many summer students assisted with data collecting and processing. The grasshoppers and crickets (Orthoptera: Caelifera and Ensifera) were separated from the collected of ground-dwelling arthropods and were examined for this thesis. 2.3 Grazing Categorization and History The ten study sites were categorized into 3 grazing classes: non-grazed, moderately grazed, and heavily grazed (Table 2). Dr. P. Krannitz selected the ten sites, based on the known livestock grazing history of each site and the landowner's interest in allowing scientific research to occur on their property. The grazing history of each site was determined from historical 15 records and interviews with current and former landowners and ranchers (memories dating back to 1930's)(Krannitz 1994). Table 2. The grazing history, definition of grazing history, and grazing history category of the ten grassland study sites in the South Okanagan, BC. Grazing History Definition of Grazing History Study Sites Non-grazed (NG) Little or no livestock grazing in the past century o, v , z Moderately Grazed (MG) Some historical livestock grazing, no current livestock grazing S, T, X , Y Heavily Grazed (HG) Some historical livestock grazing, current, heavy, livestock grazing P, U , W In addition to livestock grazing, natural or background grazing by wildlife has occurred on these sites. Feces from Californian bighorn sheep, mule deer, white-tailed deer, and NuttalPs cottontail rabbit were found in all study sites (Krannitz 1994). 2.4 Grazing History by Site The following site descriptions come from Krannitz (1994). Site O (UG) was located in northern Washington and is privately owned. It was grazed minimally historically, and has not been grazed since the 1970's. This site is separated from the neighbouring ranch by a fence erected in the 1970's. Site V (UG) was located within Inkameep Reserve. It had not been grazed since the 1920's when the addition of an irrigation canal separated it from the rest of the Reserve. 16 Site Z (UG) is privately owned and had little historical grazing. A fence separated this ranch from the one next door. In the 1970's, the occasional horse or cattle would get through a hole in the fence and temporarily graze on the land. A large fire swept through the area sometime between 1930-19502. Site S (MG) was located next to Inkameep Reserve and is a wildlife reserve, owned by the government (British Columbia Ministry of Environment, Lands, and Parks). Grazing by cattle occurred here until the 1970's; but no further grazing has occurred since that time. Vineyards now surround this site. Site T (MG) was located within the Inkameep Reserve. Some horse grazing occurred as they used this as a corridor to travel to a water source. Cattle have not grazed this site. Site X (MG) was privately owned and located near site S. Grazing from cattle in the spring and fall took place here until the 1980's. The same large fire as in site Z swept through this area sometime between 1930-1950. Site Y (MG) is owned by the Canadian Wildlife Service (CWS) and is a wildlife reserve. Some grazing by cattle and horses occurred here until 1966, when it was purchased by the Okanagan and Similkameen Park's society, and then transferred to CWS in 1971. 2 The exact date of this fire is unknown, as it is based on the memory of one of the current landowners (Krannitz 1994). 17 Site P (HG) was in the Inkameep Reserve. Heavy grazing by cattle and horses has occurred here year-round, but not every year. This site has recently been converted into a vineyard. Site U (HG) was in the Inkameep Reserve. Heavy grazing by horses occurred here, and some cattle grazing may have occurred. Site W (HG) was in the Inkameep Reserve and is privately owned. Horses grazed this site heavily. 2.5 Arthropod Collection 2.5.1 Sampling Method Arthropod samples, including grasshoppers and crickets, were collected with pitfall traps. Twenty-five pitfall traps were placed in each of the 10 study sites for two years (1994 and 1995). Each site had 25 pitfall traps, with a total of 250 traps set out across all sites. The pitfall traps were placed in rings of five, with a distance of 4.5 m between each trap (Fig. 2). These rings were placed at 50 m intervals along the vegetation transects. Vegetation was surveyed in late May of 1994 using Daubenmire plots (20cm x 50cm, 40-55 per site), which were placed at 10 m intervals along each transect (Krannitz 1994). 18 Pitfall Trap 4.5 m between traps o 50 m 50 m 50 m 50 m o Trap Ring Fig. 2. Diagram of the layout of pitfall traps at each of the ten grassland study sites in the South Okanagan, BC. In 1994, the pitfall traps were set out in early May, then collected once a month until early October (Table 3). The traps remained open and operational over the subsequent winter months and were emptied in early April of 1995, but these insects were not analyzed for this study. Following this date, insects were again collected once a month until early October. I visited several of these study sites during the summer of 1998 to familiarize myself with the antelope-brush ecosystem. Additionally, I collected grasshoppers via sweep netting to compare the species composition found by this collection method to those collected by pitfall traps. 19 Table 3. Dates of pitfall trap collections from ten grassland study sites in the South Okanagan, BC. Collection Period Dates of Collections 1 May 5-June 3, 1994 2 May 31 - July 9, 1994 3 July 4 - August 4, 1994 4 August 2 - September 8, 1994 5 September 6 - October 6, 1994 6 October 3, 1994 - April 11, 1995 7 April 8 - M a y 4, 1995 8 May 2-June 9, 1995 9 June 7 - Ju ly 7, 1995 10 July 5-August 12, 1995 11 August 9 - September 10, 1995 12 September 6 - October 6, 1995 2.5.2 Pitfall Trap Design Each pitfall trap consisted of two plastic cups (500 ml), one slightly larger than the other, so that the smaller cup sat inside the larger one. The cups were placed in the ground, with the rim just below the ground's surface. The two cups allowed only the removal of the inner cup when emptying traps, and thus minimized disturbance to the ground. The inner plastic cup was one-third filled'with 50% diluted propylene glycol, which acts as both killing agent and preservative. Predation is prevented by the propylene glycol, as predatory insects drown when they fall in the trap, thus preventing them from eating or chewing on prey insect species. Propylene glycol does not evaporate or freeze, and it does not harm wildlife that could potentially drink the liquid (Mochida and Gomyoda 1987). 20 Each pitfall trap was covered by a square, grey piece of plastic (23 cm ) that kept out rain and insects that may enter from aerial fall-out. The cover sat on 4 aluminum strips (2.5 cm ) that were placed in an " X " shape radiating out from the cups. 2.5.3 Emptying Trap Contents When the traps were collected, the inner cup was removed, and the contents were strained through a fine-mesh plastic strainer. The propylene glycol was strained into a new inner cup, which was then placed back into the original outer cup. Extra 50% propylene glycol solution was added to the new inner cup if evaporation had occurred. The organisms left in the plastic strainer were placed into plastic specimen jars, which were then fdled with 70% ethyl alcohol. If a small vertebrate (mouse, vole, or shrew) was found in the trap, the alcohol was replaced entirely. 2.6 Data Processing The collected arthropods were brought to the University of British Columbia. Laboratory assistants sorted the arthropods to order and the specimens were placed in separate vials according to order. The grasshoppers and crickets were stored in these alcohol vials for approximately three years before they were identified to the species level. Once identified to the species level, the specimens were placed back into the alcohol-filled vials for storage. Most species did not need to be mounted for identificaton. A few species of Oedipodinae were mounted for more close examination of the wings for correct identification. These species were pinned and their left 21 fore (tegmina) and hind wings were spread and mounted like those of butterfly wings (Vickery andKevan 1985). Melanoplus foedus foedus Scudder resembles Melanoplus packardii packardii Scudder and these two species can only be differentiated by close examination of the male genitalia (shape of aedeagus) (Vickery and Kevan 1985). A l l male samples were identified as Melanoplus foedus foedus based on the shape of the aedeagus; therefore, it was assumed that all female specimens were of the same species. Taxonomic keys were used to identify the collected grasshoppers and crickets (Brooks 1958; Hubbell 1936; Otte 1981; Vickery and Kevan 1985). The orthopteran synoptic collection of the Spencer Entomological Museum at the University of British Columbia was also used to assist in identifying specimens. Dr. Vernon R. Vickery (Emeritus Curator, Lyman Entomological Museum and Research Laboratory) verified the majority of the Orthoptera identifications. Dr. T. Cohn (Adjunct Curator, University of Michigan, Insect Division, Museum of Zoology) confirmed the Rhaphidophoridae. Dr. D. Johnson (Professor, University of Lethbridge) examined some of the Melanoplus specimens. Specimens are currently stored at the University of British Columbia. Voucher material will be placed in the Spencer Entomological Museum, the Royal British Columbia Museum (Victoria), and the Canadian National Museum (Ottawa). 22 2.7 Statistical Analysis 2.7.1 Data Grouping The total number of each Orthopteran species was determined for statistical analyses. This was done by totalling each species per site, across the ten collection periods. Trap disturbance was accounted for by using a random number table (Zar 1984). The greatest number of missing or disturbed traps at one site was eight in one month; therefore only data from 17 pitfall traps per month were used for statistical analyses. Depending on the number of disturbed traps per site per month, data from eight or fewer traps were eliminated following the method described in Zar (1984). Each pitfall trap was assigned a number from one to twenty-five. The random number table was entered randomly. The table columns were envisioned as two digit combinations (i.e. 07 represented pitfall trap number seven). When a two digit combination between 01 and 25 was encountered, the data from the corresponding pitfall trap was eliminated. If the trap number was already one that was disturbed, then the process for selecting a random number would be repeated. Two digit combinations over twenty-five were ignored. Movement through the table was down the columns from left to right (after the table was randomly entered). As in other studies (Capinera and Sechrist 1982; Kisbenedek 1995; Porter et al. 1996) only data of adults were used for statistical analyses because identification to the species level for nymphs is difficult or impossible for many Orthoptera species (Vickery 1987). Due to a significant number of missing traps, August 1994 data was also eliminated. 23 Preliminary analyses were done to look for trends and to find the best approach to concisely represent the findings of this study. Other groupings not represented in this thesis were examined. For example, data were analyzed by looking at yearly differences, examining grasshoppers only, examining crickets only, eliminating the most abundant grasshopper species (M. sanguinipes sanguinipes Fabricius), and eliminating both the most abundant grasshopper species and the most abundant cricket species (Gryllus sp.). No significant differences were found, hence only data analyses of the complete collection period for the entire Orthopteran species assemblage are presented. 2.7.2 Data Analysis Two computer software packages were used for data analysis. Plymouth Routines in Multivariate Ecological Research (PRIMER) v5 was used for calculating biodiversity indices, cluster analysis, and ordination by non-metric multidimensional scaling. PRIMER was developed for the purpose of studying community structure. It allows for the analysis of abundance or biomass readings to be done either spatially (several sites at one time) or temporally (same site at different times) for a community subject to different treatments (Clarke and Warwick 1994). This thesis presents data analyzed spatially. Analysis of variance testing and Regression analysis were performed with Minitab 2000. 2.7.3 Biodiversity Measures 2. 7.3.1 Species Diversity and Heterogeneity Measures Species richness was calculated using Margalef s index (d) (Clarke and Warwick 1994). This measurement of species richness also accounts for the total number of individuals. If species richness is given only as the total number of species, then sample size is not accounted 24 for because the larger the sample size, the more likely you are to get more species. Margalef s index is measured as: d = ( S - l ) / l o g N Where S = total number of species N = total number of individuals Evenness, or equitability, describes whether all species in a community are represented in similar proportions. Most communities of plants and animals are dominated by a few species and the many other species are relatively rare in numbers (Krebs 1999). Pielou's evenness index (J) (Clarke and Warwick 1994) expresses evenness as: J' = H'(observed) /H'max Where FT (observed) = observed diversity FT max — maximum possible diversity i f all species were equally abundant (— log S) The Shannon-Wiener diversity index is the most commonly used diversity measure and incorporates both species richness and evenness (Clarke and Warwick 1994). This measure places more weight on the rarer species in a community (Krebs 1999). It is calculated as: H' = - ZiPi(logePi) Where Pi = proportion of the total individuals made up by the z'th species ( P i = Ni/N) 25 Simpson's Index of diversity is a nonparametric measure that is sensitive to changes in more abundant species (Clarke and Gorley 2001). Simpson's Index (1- A.') ranges from 0 (low diversity) to almost 1 (high diversity). It is calculated as: 1-X' = 1- {ZM(Nrl)/N(N-l)} Where X = Simpson's Index Nj = number of individuals of species i N = total number of individuals (= £ Nj) 2.7.3.2 Analysis of Variance Analysis of variance (ANOVA) was used to test for differences between grazing regimes for: 1) Orthopteran abundance; 2) Margalef s species richness; 3) Pielou's Evenness Index; 4) Shannon-Wiener Index; and 5) Simpson's Index. A N O V A assumptions were checked for graphically. 2.7.3.3 Taxonomic Diversity Measures The taxonomic distinctness index is part of a recently proposed suite of biodiversity measures that examines the relatedness of species within a sample (Clarke and Gorley 2001). Specifically, the taxonomic distinctness index is the average taxonomic 'distance' or path length, traced through a Linnean or phylogenetic tree, between any two randomly chosen individuals of different species (Clarke and Warwick 1998; Warwick and Clarke 1995). There are two particular advantages to using the taxonomic distinctness index. Firstly, taxonomic distinctness is not sample-size dependent like other commonly used diversity indices that are related to species richness (Clarke and Warwick 1998; Warwick and Clarke 1995). One 26 of the difficulties with these other standard diversity indices is that it is difficult to compare across studies because of the strong dependence on sampling effort. As taxonomic distinctness does not depend on sampling effort, comparisons across studies with different sampling effort can be made, as long as comparable taxonomic accuracy is met across these studies (Clarke and Gorley 2001; Clarke and Warwick 1998). Secondly, taxonomic distinctness may be a more sensitive univariate measure of community change than species diversity and be a more accurate measure of 'biodiversity' than Shannon-Wiener (Warwick and Clarke 1995). While the total genetic makeup of a community may remain fairly constant (i.e. the values of species diversity measures would remain relatively unchanged), differences in the division of hierarchy of taxonomic units may occur depending upon the successional stage or disturbance observed at a particular assemblage (which be measurable by the taxonomic distinctness index). For example, in communities that have been slightly or moderately disturbed by anthropogenic means, measures of species diversity have not shown any change (Warwick and Clarke 2001). However, a closer examination of the division of taxonomic units may reveal changes. The average taxonomic distinctness index ( A + ) , for presence/absence data (Clarke and Warwick 1998; Warwick and Clarke 1995), was calculated as: A + = EZ,<,coy]/[^-l)/2] Where: co,y = the 'distinctness weight' given to the path length linking species i and j in the hierarchical classification; CD = 1 for species within the same genus; co = 2 for species within the same subfamily but different genera; co = 3 for species within 27 the same family but different subfamily; oo = 4 for species within the same order but different suborder; oo = 5 for species within the same order s = the number of species present; the double summation is over the set {/ = l, . . .s;7 = l,...s, such that / <j} A further complement of the average taxonomic distinctness index is the variation in taxonomic distinctness index. This index is the variance of the pairwise path lengths calculated in the taxonomic distinctness index and reflects the unevenness of the taxonomic tree. For example, two communities may have the same average taxonomic distinctness A + , but different variations in taxonomic distinctness A + . This would occur i f one community has several different orders with only one species, but also has some genera that are very species rich (higher A + ) , while the other community has many species from different families but all of the same order (lower A + ) (Clarke and Warwick 2001). The variation in taxonomic distinctness index (A + ) (Warwick and Clarke 2001) is calculated as: - e ^ V [ ^ - l)/2] = [{ILW}/Ks-i)/2]- ST2 Where m = [{YZKJ %} / {s(s - l)/2] = A + A randomization test for taxonomic distinctness and variation in taxonomic distinctness can be done to test for differences between species assemblages at different sites. It tests for the null hypothesis that a species list from one site has the same taxonomic distinctness structure as the 'master' list for all sites from which it is drawn (Clarke and Gorley 2001; Clarke and Warwick 2001). 95% probability intervals are plotted for each of a range of species sublist (M) 28 sizes, spanning the actual observed species list sizes for each site. The computations for this test are done by repeated, random drawings of M from the master list, calculating the average taxonomic distinctness (or variation in taxonomic distinctness) for each random drawing, which leads to the 95% probability range of average taxonomic distinctness (or variation in taxonomic distinctness). These computations are done for a range of species sublist sizes, in increments of factor 1.2, and result in a funnel shape when the 95% limits are plotted against the species list size (Clarke and Gorley 2001; Clarke and Warwick 2001). The actual calculated average taxonomic distinctness value or variation in taxonomic distinctness value for each site are also plotted on the 95% probability funnel. If the calculated index value for a particular site falls within the probability limits, then the null hypothesis cannot be rejected; the species assemblage at that site is not different from the known species assemblage for the entire geographical area. Calculated index values falling below the 95% probability funnel are significantly lower than expected and suggests that the site is less diverse than others while calculated index values falling above the 95% probability funnel are significantly higher than expected and suggests that the site is more diverse than others. 2.7.4 Cluster Analysis and non-metric Multidimensional Scaling Cluster analysis attempts to find natural groupings of samples so that those samples within a group are more similar than samples from different groups. Samples are arranged into groups of similar communities. Cluster analysis begins with the formation of a similarity matrix. The Bray-Curtis Similarity Coefficient was used to generate a similarity matrix as it is commonly used in ecological work for terrestrial systems (Clarke and Warwick 1994). Several studies of Orthoptera have used the Bray-Curtis Similarity Coefficient (Kisbenedek 1995; 29 Parmenter et al. 1991; Van Jaarsveld et al. 1998). The grasshopper count data were square-root transformed and reduced to presence/absence data as recommended (Clarke and Gorley 2001). The presence/absence transformation gives equal weight to all species (Clarke and Warwick 1994). The similarity matrix begins with the calculation of similarity (5) between every pair of species. The similarity coefficient S is conventionally defined between the range (0, 100%) or (0, 1) and refers to between-sample similarity. 5* = 100% or 1 if two species have significant representation at the same set of sites while S = 0 if two species never occur within the same set of sites. The Bray-Curtis coefficient is calculated as: Sjk = 100(1- (ZP,=, \y,j - ylk\) / (27,=, (yl} + yik)} Where: yij = entry in the /th row and the /th column of the data matrix. For example, this means the abundance for the /th species in the /th sample (/ = 1,2,... ,p; j = 1,2,...,n) yik = entry in the /th row in the M i sample |... | = the absolute value of the difference Starting from the similarity matrix generated by cluster analysis, ordination by non-metric multidimensional scaling (MDS) (Clarke and Warwick 1994; Kruskal 1964) was performed. A rank similarity matrix is formed and then displayed graphically as an ordination plot or map. The placement of samples or sites on this map is such that sites with similar species composition are located in closer proximity than sites with very different communities. 30 2.7.5 Vegetation Data The following three vegetation variables (provided by Dr. P. Krannitz) were selected and examined closely based on previous studies of Orthoptera and biotic data: 1) plant species richness (Evans 1987; Kemp et al. 1990; Kisbenedek 1995; Parmenter et al. 1991;Pfadt 1982) 2) percent bare soil (Fielding and Brusven 1995b; Miller and Onsager 1991) 3) percent plant cover (Fielding et al. 2001; Fielding and Brusven 1995b; Kemp et al. 1990; Kisbenedek 1995; Miller and Onsager 1991; Parmenter et al. 1991; Pfadt 1982; Prendini etal. 1996). Percent bare soil and percent plant cover were arcsine transformed, as recommended for percentage data (Zar 1984). Plant species richness is the total number of plant species per study site. Plants provide both food and shelter to grasshoppers. Changes in the plant community, such as species richness, amount of plant cover, or amount of bare ground, may lead to changes in the orthopteran community (Joern 2004; Prendini et al. 1996). Several studies have shown that grasshopper community attributes change with changes in the plant community (Capinera and Sechrist 1982; Evans 1988; Joern 1982; Kemp et al. 1990; Otte 1976; Pfadt 1982; Quinn et al. 1991; Scoggan and Brusven 1973). A relationship was looked for between vegetation data and orthopteran data. Firstly, vegetation data were superimposed (one at a time) onto the MDS ordination plot. The vegetation data is represented by circles of differing sizes on the MDS ordination plot. The 31 larger the circle on the ordination plot, the greater the value of the vegetation variable at that particular study site. This technique allows for quick visualization to see i f a vegetation variable appears to be linked to a particular biotic cluster. Secondly, correlations were tested between vegetation variables and orthopteran data. 2.7.6 Correlation and Regression Analysis Correlation analyses (Pearson correlation) were performed between orthopteran data and the following vegetation variables: percent bare soil, percent plant cover, and plant species richness. The following orthopteran data were used for the analyses: total number of Orthoptera species per site; average number of Orthoptera per trap; Shannon-Wiener Index results; Simpson's Index results. Regression analysis was then carried out with those variables that strongly correlated. 32 3 RESULTS 3.1 Descriptive Statistics and General Observations In total, 9,537 Orthoptera specimens were captured. Twenty-four species of grasshoppers (Caelifera) and crickets (Ensifera) (Table 4), of which seventeen species were grasshoppers and seven species were crickets, were represented in the total catch. Two species of Orthoptera were found at all ten sites (Table 5): Melanoplus sanguinipes sanguinipes and Gryllus sp3. Sixty-three percent of the captured specimens were crickets (6,046 specimens) and thirty-seven percent (3,491 specimens) were grasshoppers (Table 6). Five grasshopper species comprised 96% of the total catch of grasshoppers. M. sanguinipes sanguinipes was the dominant species, comprising 57% of the total grasshopper catch. Amphitornus coloradus ornatus4 comprised 5% of the total grasshopper catch, Melanoplus foedus foedus and Arphia pseudonietana pseudonietana each comprised 4% of the total grasshopper catch, and Ageneotettix deorum comprised 3% of the total grasshopper catch. Twenty-three percent of the captured grasshoppers were immature nymphs, and nearly all belonged to the genus Melanoplus. Taking the captured nymphs into account, M. sanguinipes sanguinipes likely comprised nearly 80% of the total grasshopper catch. The remaining twelve grasshopper species comprised the remaining 4% of the total grasshopper catch. Four cricket species comprised 97% of the total catch of crickets. Gryllus sp. was the dominant species, comprising 80% of the total cricket catch. Ceuthophilus vicinus and Ceuthophilus agassizii together comprised 10% of the total cricket catch while Stenopelmatus 3 Dr. D. Weissman has identified these specimens as a possible new species. See Table 4 for further details. 4 See Table 4 for authorities. 33 sp. comprised 7% of the total cricket catch. The remaining cricket species comprised the remaining 3% of the total cricket catch. Seven species of grasshoppers were represented by less than five specimens: M. cinereus cinereus, Spharagemon equale, Aulocara ellioti, M. sp. 2, Trimerotropis sp. 1, Trimerotropis sp. 2, and Xanthippus sp. 1. Four species of crickets have yet to be named and described by experts: Gryllus sp., n. sp. dasy, Stenopelmatus sp., and Steiroxys sp. Several of the study sites were visited in August and September of 1998. Where the vegetation permitted (i.e. where antelope-brush was not abundant), the grasshopper assemblages were sampled with sweep nets. M. sanguinipes sanguinipes was by far the most abundantly captured grasshopper and Amphitornus coloradus ornatus was the second most abundantly captured grasshopper. Many Arphia pseudonietana pseudonietana were observed flying and were identified by the distinguishing red patch on their hind wings. However, none were captured via sweep netting because they were very active in the hot, afternoon sun. These results corroborate the pitfall trap collection data. No orthopteran species were captured that had not been caught in the pitfall traps. As a result of these findings, it is assumed that the pitfall trap method of collecting Orthoptera adequately sampled the orthopteran assemblage of the study area. 34 Table 4. List of orthopteran species found at ten study sites in the South Okanagan, BC. Order Orthoptera Suborder Caelifera (shorthorned grasshoppers, locusts, and relatives) Family Acrididae Subfamily Gomphocerinae (slant-faced grasshoppers) Tribe Aulocarini Genus Ageneotettix deorum (Scudder) Genus Aulocara elliotti (Thomas) Genus Psoloessa delicatula (Scudder) Tribe Eritettigini Genus Amphitornus coloradus ornatus McNeill Tribe Mermiriini Genus Pseudopomala brachyptera (Scudder) Subfamily Melanoplinae (spur-throated grasshoppers) Tribe Melanoplini Genus Melanoplus cinereus cinereus Scudder foedus foedus Scudder sanguinipes sanguinipes (Fabricius) Melanoplus sp. I5 Melanoplus sp. 2 6 Genus Phoetaliotes nebrascensis (Thomas) Subfamily Oedipodinae (Locustinae) (band-winged grasshoppers) Tribe Arphia Genus Arphia pseudonietana (Thomas) Tribe Hippiscini Genus Xanthippus Xanthippus sp. Tribe Sphingonotini Genus Conozoa sulcifrons (Scudder) 5 Only female specimens were collected. Many Melanoplus females are difficult to impossible to identify without corresponding males. Two Canadian experts, Dr. D. Johnson and Dr. V . R. Vickery, were unable to identify these females with certainty. 6 Only female specimens were collected. Dr. D. Johnson was unable to identify these females with certainty. 7 Only one female specimen was collected. Identification was made difficult by loss of colour (a key identifying feature for many Oedipodinae) during preservation. Specimen likely X. corallipes (Haldeman) or X. aquilonius (Otte). 35 Genus Spharagemon equate (Say) Genus Trimerotropis Trimerotropis sp. 18 Trimerotropis sp. 2 9 Suborder Ensifera (crickets, katydids, weta) Family Gryllidae Tribe Gryllini Genus Gryllus Gryllus sp.]0 Tribe Oecanthini Genus Oecanthus quadripunctatus (Beutenmuller) Family Rhaphidophoridae Tribe Ceuthophilini Genus Ceuthophilus agassizii (Scudder) vicinus (Hubbell) Tribe Pristoceuthophilini Genus Pristoceuthophilus n.sp. dasy11 Family Stenopelmatidae Genus Stenopelmatus Stenopelmatus sp.n Family Tettigoniidae Tribe Platycleidini Genus Steiroxys Steiroxys sp.U Only one female specimen was collected. Identification was made difficult by loss of colour (a key identifying feature for many Oedipodinae) during preservation. Specimen resembles Trimeroptropis pallidipennis. 9 Only one female specimen was collected. Further identification was not possible. 10 Gryllus sp. keys out to Gryllus veletis and Gryllus pennsylvanicus. These two species can be differentiated only by life history in northeastern North America; this method of differentiating the two species does not apply in southwestern Canada. Additionally, Dr. David Weissman (who is currently revising the genus Gryllus) listened to the songs of Gryllus sp. and has identified them as a possible new species (personal communication through Dr. Scudder, 2000). " Pristoceuthophilus n. sp. dasy keyed out to Pristoceuthophilus pacificus. However, specimens were sent to Dr. T. Cohn who is working on a revision of the subfamily Ceuthophilinae. He is currently in the process of naming this new species (personal communication, 1999). 12 Stenopelmatus sp. keys out to Stenopelmatus fuscus Haldeman. However, Dr. David Weissman has heard the male calls and has identified them as a new, undescribed species (personal communication through Dr. Scudder, 2000). , J The taxonomy of Steiroxys is currently being revised (Vickery and Kevan, 1985). 36 Table 5. Species occurrence at ten grassland study sites in the South Okanagan, BC. Non-grazed Moderately Grazed Heavily Grazed Site Name and Abbreviation Oroville Brights Winery Kennedy Bench Black Sage Road Inkameep Kennedy Flats Canadian Wildlife Bench Osoyoos Lake Watertower Near Mud Lake Number of Sites Found Site Name and Abbreviation 0 V Z S T X Y P U W Species Grasshoppers Gomphocerinae Ageneotettix deorum • • • • • • • • 8 Aulocara elliotti • 1 Psoloessa delicatula delicatula • • • 3 Amphitornus coloradus ornatus • • • • • • • • • 9 Pseudopomala brachyptera • • • • 4 Melanoplinae Melanoplus cinereus cinereus • • • 3 Melanoplus foedus foedus • • • • • • • 7 Melanoplus sanguinipes sanguinipes 10 Melanoplus sp. 1 • • • • • • • 7 Melanoplus sp. 2 • 1 Phoetaliotes nebrascensis • • • • • 5 Oedipodinae Arphia pseudonietana pseudonietana • • • • • • • • • 9 Xanthippus sp. • 1 Conozoa sulcifrons • • • 3 Spharagemon equale • • • 3 Trimerotropis sp. 1 • 1 Trimerotropis sp. 2 • 1 Number of grasshopper species per site 8 5 6 10 10 5 2 14 9 7 Number of grasshopper species per grazing intensity 8 14 14 37 Table 5 (cont). Species occurrence at ten grassland study sites in the South Okanagan, BC. Non-grazed Moderately Grazed Heavily Grazed Site Name and Abbreviation Oroville Brights Winery Kennedy Bench Black Sage Road Inkameep Kennedy Flats Canadian Wildlife Bench Osoyoos Lake Watertower Near Mud Lake Number o f Sites Found Site Name and Abbreviation 0 V Z S T X Y P U W Species Crickets Gryllus sp. 10 Oecanthus quadripunctatus • • • • • 5 Ceuthophilus agassizii • • • • 4 Ceuthophilus vicinus 9 Pristoceuthophilus n. sp. dasy • • • • • 5 Stenopelmatus sp. • • • • • • • • 8 Steiroxys sp. • • • • • • 6 Number o f cricket species 3 3 6 4 6 6 4 6 5 4 Number o f cricket species per grazing treatment 6 6 6 Number o f grasshopper and cricket species 11 8 12 14 16 11 6 20 14 11 Number o f grasshopper and cricket species per grazing treatment 14 20 20 38 Table 6. Total catch over the entire collection period from ten grassland study sites in the South Okanagan, BC (missing traps are not accounted for). Non-grazed Moderately Grazed Heavily Grazed Site Name and Abbreviation Oroville Brights Winery Kennedy Bench Black Sage Road Inkaneep Kennedy Flats Canadian Wildlife Bench Osoyoos Lake Watertower Near Mud Lake Total Site Name and Abbreviation O V Z S T X Y P U W Species Caelifera (Grasshoppers) Melanoplus sanguinipes sanguinipes 240 962 40 209 85 132 53 115 72 74 1982 Melanoplus nymph14 23 241 11 136 96 51 12 119 66 46 801 Amphitornus coloradus ornatus 10 43 1 26 19 2 21 4 42 168 Arphia pseudonietana pseudonietana 2 18 16 29 28 4 6 8 17 128 Melanoplus foedus foedus 43 1 9 21 1 34 14 123 Ageneotettix deorum 40 2 3 15 1 29 3 7 100 Melanoplus sp. 1 7 5 3 2 16 3 25 61 Phoetaliotes nebrascensis 16 15 16 1 6 54 Psoloessa delicatula delicatula 4 4 13 21 Conozoa sulcifrons 3 1 15 19 Pseudopomala brachyptera 3 4 1 1 9 Melanoplus cinereus cinereus 1 1 3 5 Gomphocerinae nymph 1 1 3 5 Spharagemon equate 1 2 1 4 Oedipodinae nymph 1 1 1 1 4 Aulocara elliotti 2 2 Melanoplus sp. 2 2 2 1 4 Early instar grasshopper nymphs are difficult to identify to species. Those that were unidentifiable are listed as nymphs to the subfamily level. 39 Table 6 (cont.). Total catch over the entire collection period from ten grassland study sites in the South Okanagan, BC (missing traps are not accounted for). Non-grazed Moderately Grazed Heavily Grazed Site Name and Abbreviation Oroville Brights Winery Kennedy Bench Black Sage Road Inkaneep Kennedy Flats Canadian Wildlife Bench Osoyoos Lake Watertower Near Mud Lake Total Site Name and Abbreviation O V Z S T X Y P U W Trimerotropis sp. 1 1 1 Trimerotropis sp. 2 1 1 Xanthippus sp. 1 1 Sum 384 1274 87 435 277 191 66 357 199 221 3491 Average number o f grasshoppers per treatment 582 242 259 Ensifera (Crickets) Gryllus sp. 221 2478 187 577 458 288 297 29 24 258 4817 Stenopelmatus sp. 112 60 75 5 56 6 85 51 450 Ceuthophilus females 1 5 7 2 30 55 242 22 16 22 48 444 Ceuthophilus vicinus 8 14 3 26 16 11 18 8 11 115 Pristoceuthophilus n. sp. dasy 21 7 59 2 8 97 Ceuthophilus agassizii 3 0 61 1 12 77 Steiroxys sp. 2 4 3 7 15 1 32 Oecanthus quadripunctatus 2 3 1 6 2 14 Sum 348 2484 319 658 555 729 338 170 108 337 6046 Average number o f crickets per treatment 1050 570 205 Total number o f grasshoppers and crickets 732 3758 406 1093 832 920 404 527 307 558 Average number o f grasshoppers and crickets per treatment 1632 812 464 Female Ceuthophilus agassizii and Ceuthophilus vicinus are difficult to distinguish. Upon examination of these specimens, Dr. Cohn was unable to distinguish the females by conventional means (ovipositor shape and colour patterns) (personal communication, 2000). 40 3.2 Orthopteran Abundance The total number of Orthoptera captured in pitfall traps (all sites) peaked in June of 1994 and August 1995 (Fig. 3). The total number of orthopteran species captured peaked in September of 1994 and August of 1995. Figure 4 presents the proportion of captured Orthoptera that are grasshoppers or crickets, and adults or nymphs. i i Number of S p e c i e s — • — Number of Orthoptera S p e c i m e n s Fig. 3. The total number of orthopteran species and the total number of orthopteran specimens collected per month from ten grassland study sites in the South Okanagan, BC. 41 1800 -I «o 1600 -® 1400 1 • G rasshoppe r adults ffl Cr icke t adults • G r a s s h o p p e r nymphs • Cr icke t nymphs Fig. 4. The total number of Orthopterans (divided into nymphs and adults, grasshoppers and crickets) summed across all ten grassland study sites per month, captured in the South Okanagan, BC. General trends were examined between the ten study sites. The average number of Orthoptera per pitfall trap was highest at non-grazed site V and lowest at heavily grazed site U (Fig. 5). Heavily grazed site P had the highest number of species (20) while moderately grazed site Y (6) had the lowest number of species (Fig. 5). The mean number of Orthoptera captured per grazing treatment did not vary across grazing treatments (Fig. 6). A non-significant trend of decreasing number of Orthopterans with increased grazing is observed. An analysis of variance test was performed to test for the null hypothesis that the number of Orthoptera captured did not vary between grazing regimes and could not be rejected (F 2 j9 = 1.15,/? = 0.371). 42 Q. 200 (0 k_ 180 0) *•> Q. 160 O SI 140 •e Q. 120 O (0 k. o 100 k. « jmbe pitf 80 60 c 40 <1> D) TO 20 0 > < i O V Non-grazed 14 i T X Moderately grazed Site I G r a s s h o p p e r s • C r i cke t s 14 11 h u w Heavily grazed Fig. 5. Average number of Orthoptera per pitfall trap per site (with standard error bars) captured from ten grassland study sites in the South Okanagan, BC, for all collection periods. The number of orthopteran species per site is listed above the standard error bars. 43 2400 -100 Non-grazed Moderately grazed Heavily grazed Fig. 6. Mean number of Orthoptera per grazing category, captured from ten grassland study sites in the South Okanagan, B C (with 95% confidence intervals). Non-grazed (O, V , Z) ; moderately grazed (S, T, X , Y ) ; and heavily grazed (P, U , W) . The number of orthopteran species per grazing category is listed above the confidence intervals. 3.3 Site Groupings Fourteen species of Orthoptera were found in all grazing categories (Table 7). A l l seven species of crickets were found in all grazing categories and seven species of grasshoppers were found in all grazing categories. Three of these seven grasshopper species, Melanoplus sanguinipes sanguinipes, Melanoplus foedus foedus, and Phoetaliotes nebrascensis (the first, fourth, and seventh most abundant grasshopper species caught, respectively) are mixed grass and forb feeders. Mixed grass feeders are often the most abundant species because they are best able to be successful in varying conditions as they can feed on both grasses and forbs. 44 Table 7. List of orthopteran species captured at all three grazing categories from the ten grassland study sites in the South Okanagan, BC. Suborder Caelifera (Grasshoppers) Suborder Ensifera (Crickets) Melanoplus sanguinipes sanguinipes Gryllus sp. Amphitornus coloradus ornatus Stenopelmatus sp. Arphia pseudonietana pseudonietana Ceuthophilus vicinus Melanoplus foedus foedus Ceuthophilus agassizii Ageneotettix deorum Pristoceuthophilus n. sp. dasy Melanoplus sp. 1 Steiroxys sp. Phoetaliotes nebrascensis Oecanthus quadripunctatus No orthopteran species were to found only on non-grazed sites (Fig. 7). Two species of grasshoppers were found only on moderately grazed sites: Trimerotropis sp. 1 and Trimerotropis sp. 2 (both at site T). Both are represented by only one specimen. Three species of grasshoppers were found only on heavily grazed sites: Xanthippus sp. (site P), represented only by one specimen; Melanoplus sp. 2 (site P), represented by 61 specimens; and Aulocara elliotti (site P) represented by 2 specimens (obligate grass feeder). Nine specimens of Pseudopomala brachyptera were found at both non-grazed and moderately grazed sites (sites O, V , S, X) . Four species of grasshoppers were found at both moderately grazed and heavily grazed sites: Psoloessa delicatula delicatula (sites T, P, U) (obligate grass feeder), represented by 21 specimens; Conozoa sulcifrons (sites T, P, U) (mixed grass and forb feeder), represented by 19 specimens; Melanoplus cinereus cinereus (sites S, P, W) (obligate forb feeder), represented by 5 specimens; and Spharagemon equale (sites S, P, U) (mixed grass and forb feeder), represented by 4 specimens. 45 Non-grazed ^seudopomala bpamyptera Moderately Grazed Trimerotropis sp. f Trimerotropis sp. 2 14 species Conozoa sulcifrons Melanoplus cinereus cinereus Psoloessa delicatula delicatula Spharagemon equate Heavily Grazed lAulochara elliotti Melanoplus sp. 2 Xanthippus sp. Fig. 7. Venn diagram representing grasshopper species presence and absence according to grazing category and the overlap between grazing categories from ten grassland study sites in the South Okanagan, B C 3.4 Trap Disturbance Twenty-nine traps were destroyed over the ten collection periods used in this study. The traps were destroyed or disturbed by cattle, wildlife (such as coyotes and deer), or inclement weather. In addition to the many arthropod species captured in the pitfall traps, a few small vertebrates were captured in the pitfall traps. These include mice, shrews, voles, and spade-foot toads. As these vertebrates were most likely to have died (i.e. by drowning or ingesting the trap liquid) soon after falling into the traps, they are assumed to have a negligible effect on the trap catch. These traps with dead vertebrates had an abundance of silphid (Family Silphidae) 46 beetles, which were assumed to be attracted to the odour of the decaying vertebrates. Silphid beetles have not been observed to be attracted to the dead arthropods in pitfall traps. No other insect group was observed to be attracted or repelled by the odour of these dead vertebrates. 3.5 Biodiversity Measures Table 8 shows the results of the calculated biodiversity indices. A single factor analysis of variance between the calculated biodiversity indices and the three grazing regimes did not reveal significant differences for any of the grazing levels. A l l A N O V A tests revealed low F values and high p values (assumptions were met) and are as follows: 1) Margalef s Species Richness: (F 2 i9 = 2.36, p = 0.165); 2) Pielou's Evenness Index: (F 2 ;9 = 0.67p = 0.541); 3) Shannon-Weiner Index: (F 2 j 9 = 1.63, p = 0.262); and 4) Simpson's Index: (F 2, 9 = 0.55,/? = 0.599). 47 Table 8. The calculated biodiversity indices for each of the ten grassland study sites and the biodiversity indices averaged for each grazing regime from the South Okanagan, BC. Grazing History on-grazed Moderate y Grazed Heavily Grazed Site V Z O S Y X T U W P Margalef s Species Richness 0.8 1.8 1.3 2.0 1.1 1.1 2.4 2.1 1.9 3.0 Average 1.3 1.6 2.3 Pielou's Evenness Index 0.4 0.7 0.8 0.5 0.6 0.7 0.7 0.7 0.8 0.7 Average 0.6 0.6 0.7 Shannon-Weiner Index 0.8 1.7 1.7 1.3 1.1 1.4 1.8 1.7 1.8 2.0 Average 1.4 1.4 1.8 Simpson's Index 0.5 0.8 0.8 0.6 0.6 0.7 0.8 0.7 0.8 0.8 Average 0.7 0.7 0.8 Taxonomic Distinctness Index 82.6 74.12 69.46 76.57 80.25 77.89 73.42 72.1 70.71 67.96 Variation in Taxonomic Distinctness 383 390.1 511.6 470.5 177.8 144.9 337.1 411.8 495.5 413.4 3.5.1 Taxonomic Distinctness Measures Figure 8 shows the 95% probability funnel of average taxonomic distinctness plotted against number of species for 1000 random simulations. Sites Z, Y , X , and T fall on or above the theoretical mean while the remaining sites fall below the theoretical mean. A l l sites fall within the 95% probability limits. The higher the average taxonomic distinctness value, the greater the taxonomic diversity (i.e. the distribution of species between suborder, family, subfamily, and genus). As seen in Figure 8, the range of average taxonomic distinctness values does not vary greatly between sites. Moderately grazed sites Y and X are somewhat more biodiverse than non-grazed sites V and O, moderately grazed site S, and heavily grazed site U , W, and P. 48 90-r § 40 I 1 1 1 < 5 10 15 20 Number of Species Fig. 8. Simulated means (dashed line), 95% probability funnels (continuous line), and measured average taxonomic distinctness ( A + ) values for each of the ten grassland study sites in the South Okanagan, B C , plotted against the number of species for 1000 random simulations ( • = non-grazed, A= moderately grazed, and • = heavily grazed). Figure 9 presents the 95% probability funnel of variation in taxonomic distinctness plotted against number of species for 1000 random simulations. The majority of sites have expected or below expected values of variance in taxonomic distinctness. Six sites, O, V , Z, S, U , and W, fall very near the theoretical mean while sites Y , T, and P fall within the lower probability limits. Site X falls just below the lower probability limit. 49 +< 1000-r Number of Species Fig. 9. Simulated means (dashed line), 95% probability funnels (continuous line), and measured variation in taxonomic distinctness (A + ) values for each of the ten grassland study sites in the South Okanagan, BC, plotted against the number of species for 1000 random simulations ( • = non-grazed, A= moderately grazed, and • = heavily grazed). An ellipse diagram (Fig. 10) shows the variation in taxonomic distinctness plotted against average taxonomic distinction. Each 'ellipse' is the 95% probability contour for the simulated distribution of species size M = 5, 10, or 15. A negative correlation is observed, which means the higher the average taxonomic distinctness, the lower the variation in taxonomic distinctness. 50 Average Taxonomic Distinctness (A+) Fig. 10. Fitted 95% probability contours of the joint distribution of A + and A + , from 1000 random simulations (for species sublist sizes = 5, 10, and 15), calculated for each of the ten grassland study sites in the South Okanagan, B C ( • = non-grazed, A= moderately grazed, and • = heavily grazed). 3.6 Cluster Analysis and non-metric Multidimensional Scaling The results of cluster analysis on the Bray-Curtis similarity matrix showed that the sites did not group based on grazing regime. Cluster analysis was performed on square root-transformed data and presence/absence data. The square root-transformed data showed the following four groupings (Fig. 11): V ; X , Z , and Y ; U and P; O, S, T, and W. The presence/absence data showed the following four groupings (Fig. 12): X ; V ; U , T, P, S, and O; W , Y , and Z . 51 -w © © u A , 1 , , , © 20 40 60 80 100 Similarity Fig. 11. Dendrogram for hierarchical clustering of the ten grassland study sites in the South Okanagan, BC, using group-average linking of Bray-Curtis similarities calculated on square root-transformed data (O = non-grazed, A = moderately grazed, and • = heavily grazed). 52 Similarity Fig. 12. Dendrogram for hierarchical clustering of the ten grassland study sites in the South Okanagan, BC, using group-average linking of Bray-Curtis similarities calculated on presence/absence data (O = non-grazed, A = moderately grazed, and • = heavily grazed). The resulting ordination plots from non-metric multidimensional scaling do not show sites grouping based on grazing regime. For the square root-transformed data (Fig. 13), site V is an outlier, sites S, O, and T loosely group, sites W, X , Z, and Y loosely group, and sites U and P loosely group. For the presence/absence data (Fig. 14), sites V and X are outliers, sites U , P, T, S, and O loosely group, and sites W, Z, and Y loosely group. 53 ^ S t ress : 0.11 A w ® # A U P Fig. 13. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC (O = non-grazed, A = moderately grazed, and • = heavily grazed). 54 A Stress: 0.1 ® w / \ © p u Fig. 14. MDS of Bray-Curtis similarities from presence/absence data from the ten grassland study sites in the South Okanagan, BC (O = non-grazed, A = moderately grazed, and • = heavily grazed). 55 Ordination plots are listed with stress levels by PRIMER v5 and these stress levels give a measure of how adequately the two-dimensional "map" represents the high-dimensional relationships among samples (Clarke and Gorley 2001). These stress levels refer to the distortion between the similarity rankings (from the similarity matrix generated by cluster analysis) and the corresponding distance rankings in the ordination plot. The stress levels of 0.11 (square root-transformed data) and 0.1 (presence/absence data) fall within or near reliable guidelines (Clarke and Warwick 1994). (Clarke and Warwick 1994) state that stress levels of less than 0.1 correspond to a good ordination that is extremely unlikely of giving a misinterpretation, ordination plots with stress levels between 0.2-0.3 should be treated with a great deal of skepticism, and ordination plots with stress levels greater than 0.3 should be discarded. 3.6.1 Vegetation Overlays Vegetation data (Table 9) were superimposed onto the MDS ordination plots. While these overlays are not statistical tests, they can reveal site groupings that are not obvious and link site groupings to a vegetation variable. The larger the superimposed "bubble", the larger the value of the superimposed variable (no scale for these bubbles is provided by PIMER v5). 56 Table 9. List of percent bare ground, percent plant cover, and plant species richness for the ten grassland study sites in the South Okanagan, BC (raw data provided by Dr. P. Krannitz). Grazing Intensity Study Site Percent Bare Ground (arcsine transformed) Percent Plant Cover (arcsine transformed) Plant Species Richness Non-grazed 0 21.64 29.6 41 Non-grazed V 3.53 24.73 40 Non-grazed z 11.39 43.68 67 Moderately grazed S 12.11 49.89 46 Moderately grazed T 23.26 41.96 53 Moderately grazed X 4.55 28.93 44 Moderately grazed Y 9.97 48.50 56 Heavily grazed P 43.97 17.95 35 Heavily grazed u 29.93 34.76 50 Heavily grazed w 13.79 18.72 37 Percent bare soil appears to influence heavily grazed sites (Fig. 15 and 16). No trends were found when percent plant cover (Fig. 17 and 18) and plant species richness (Fig. 19 and 20) were overlayed onto the MDS ordination plots of orthopteran species abundance data and orthopteran presence/absence data. 57 M Stress: 0.11 M H N N 8> Fig. 15. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent bare soil (arcsine transformed) (N = non-grazed, M = moderately grazed, H = heavily grazed). 58 „ S t r e s s : 0.1 M N B ^ N M ( H J ® M ® Fig. 16. MDS of Bray-Curtis similarities from species presence/absence data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent bare soil (arcsine transformed) (N = non-grazed, M = moderately grazed, H = heavily grazed). 59 e Fig. 17. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent plant cover (arcsine transformed) (N = non-grazed, M = moderately grazed, H = heavily grazed). 60 Fig. 18. MDS of Bray-Curtis similarities from species presence/absence data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing percent plant cover (arcsine transformed) (N = non-grazed, M = moderately grazed, H = heavily grazed). 61 Fig. 19. MDS of Bray-Curtis similarities from square root-transformed species abundance data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing plant species richness (N = non-grazed, M = moderately grazed, H = heavily grazed). 62 Fig. 20. MDS of Bray-Curtis similarities from species presence/absence data from the ten grassland study sites in the South Okanagan, BC, with superimposed circles of increasing size with increasing plant species richness (N = non-grazed, M = moderately grazed, H = heavily grazed). 3.7 Correlation Analysis Correlation analyses between orthopteran data and vegetation variables (Table 10) show that both orthopteran species richness and the calculated Shannon-Wiener Index are strongly correlated with percent bare soil. 63 Table 10. Correlations between vegetation data and orthopteran data from the ten grassland study sites in the South Okanagan, BC. Bolded correlations show a high, positive correlation. Orthopteran Data Percent Bare Soil Percent Plant Cover Plant Species (arcsine (arcsine Richness transformed) transformed) Average number of Orthoptera per pitfall trap per -0.419 -0.219 -0.317 site Total number of orthopteran species per site 0.812 0.561 0.554 Shannon-Wiener Index 0.737 -0.279 -0.060 Simpson's Index 0.537 -0.256 0.042 3.8 Regression Analysis Regression analysis was performed on orthopteran species richness versus percent bare soil (Fig. 21) and the Shannon-Wiener Index versus percent bare soil (Fig. 22) (the assumptions for normality and homoscedasticity are met). A significant, positive relationship is shown for both tests with F , ; 8 = 15.518,/? = 0.004, r 2 = 66% and F i > 8 = 9.51,p = 0.015, r 2 = 54.3%, respectively. Orthopteran species richness increases with increasing percent bare soil and the calculated Shannon-Wiener Index increases with increasing percent bare soil. 64 Number of Orthoptera Species = 7.84 + 0.258 Percent Bare Soil 25 O O 10 15 20 25 Percent Bare Soil Regress i on 95% CI Fig. 21. Fitted line plot of orthopteran species richness and percent bare soil (arcsine transformed) (Fi ; g = 15.518,/? = 0.004, r 2 = 66%) from the ten grassland study sites in the South Okanagan, BC. 65 Shannon Wiener Index = 1.52 + 0.022 Percent Bare Soil Regress ion 95% CI T i r~ T i i \ i i r O 5 10 15 20 25 30 35 40 45 Percent Bare Soil Fig. 22. Fitted line plot of Shannon-Wiener Index values and percent bare soil (arcsine transformed) (F^ g -9.5l,p- 0.015, r 2 = 54.3%) from the ten grassland study sites in the South Okanagan, BC. 66 4 DISCUSSION 4.1 Trapping Method Pitfall trapping is one of the most commonly used and effective methods of collecting surface-active arthropods (Lavallee 1999; Luff 1975; Melbourne et al. 1997). This collection method has become the preferred trapping method in biodiversity surveys (Melbourne et al. 1997). As the arthropod component of the South Okanagan Grassland Conservation Research Project was designed to investigate the biodiversity of all ground-dwelling arthropods, pitfall trapping was the collection method chosen. This collection method is not widely used for collecting grasshoppers, but has turned out to be particularly well suited for the collection of grasshoppers in the antelope-brush ecosystem of the South Okanagan, for reasons which are discussed below. 4.1.1 Sweep Netting Sweep netting is the most common method of capturing grasshoppers in grasshopper community studies on grasslands (Evans and Bailey 1993; Evans et al. 1983; Hewitt and Onsager 1988; Knutson and Campbell 1974; Miller and Onsager 1991). Other grasshopper capture techniques used less frequently include night cages (Anderson and Wright 1952), pan traps (Evans et al. 1983), vacuums (Dietrick 1961), drop cages (Smalley 1960), and pitfall traps (Dr. D. Johnson, personnel communication, 2000) (Parmenter et al. 1991). Grasshopper density measures are usually conducted by the standardized ring sampling technique (Fielding et al. 2001; Fielding and Brusven 1995b; Joern 2004; Onsager 1977; Onsager and Henry 1977). This technique requires the visual estimation of the number of grasshoppers jumping out of a ring, which was placed in the sample site several hours prior to the count, when flushed with a wand. 67 Grasshoppers are not captured by this density measurement technique. Therefore, ring sampling is usually followed with sweep netting to capture grasshoppers if identification or relative abundance measurements are desired. As with all arthropod collection methods, shortcomings exist with the use of sweep netting. The limitations of sweeping include variation of sampling effectiveness due to vegetation structure and weather influences (Southwood 1978), variation in grasshoppers' jump reaction (and thus their ability to jump out of the net) with daytime temperature changes (Delong 1932), the lower strata of vegetation is not sampled (Southwood 1978), and individual grasshopper species have different abilities to avoid capture (Capinera and Sechrist 1982; Southwood 1978). Despite these limitations, sweeping is considered one of the most practical means of sampling a large number of grassland sites in a short period of time. This method is considered reliable for estimating relative abundances of grasshopper species and for measuring diversity indices (Evans 1988; Prendini et al. 1996). While sweep netting may be an effective technique for collecting grasshoppers, difficulties occur with using this method in the antelope-brush ecosystem. Most grasshopper community studies in North America sample grasshoppers with sweep nets in short-grass, mixed-grass, or tall-grass prairies. These grasslands are not dotted with large shrubs, such as antelope-brush, which varies in sizes from 2.5-6.1 m 2 (Krannitz 1994) in the study sites. Antelope-brush interferes with standard sweeping methods as one cannot sweep through this shrub as one does through grass. The standard sweep arc is 180° (Evans et al. 1983). While the number of sweeps and the number of sets of sweeps along a transect varies for each study, they 68 remain the same within each study. As the density of antelope-brush varied among the ten study sites, maintaining a consistent sweep net protocol between study sites would have been difficult to impossible. For example, the antelope-brush stands at site O were so dense that sweep netting would not have even been possible at that site. Additionally, the collection of sufficient number of crickets would not have been possible without pitfall traps, as they are specifically ground-dwelling (most species do not perch on grasses nor do they fly) and most captured species were nocturnal. Similar to this study, Evans (1983) describes how the ring sampling method of estimating grasshopper density was not conducive to tall-grass prairie. The tall, dense vegetation of tall-grass prairie in eastern Kansas did not allow for easy visual estimation of numbers of jumping grasshoppers. Therefore, the selection of sampling method is also dependent upon the habitat structure in a study area. 4.1.2 Pitfall Trapping The advantages of pitfall trapping include that it is inexpensive (Topping and Sunderland 1992), set-up and maintenance are straightforward (Luff 1975), sampling is continuous, thus allowing capture of both diurnal and nocturnal arthropods (Luff 1975; Topping and Sunderland 1992), sufficient specimens are caught for rigorous statistical analyses (Spence and Niemala 1994; Topping and Sunderland 1992), and disturbance to habitat is minimal (Spence and Niemala 1994). As with all trapping methods, there are certain limitations, including differential rate of sexes falling into the traps, differential rate of different species falling into the traps, temporal variation in trap efficiency, and the influence of vegetation structure on capture rates (Greenslade 1964; Luff 1975; Southwood 1978; Topping 1993). 69 Greenslade (1964) found that habitat structure did bias the capture rate of arthropods, but these results were only significantly affected when groundcover was very dense (fewer specimens were captured under very dense groundcover). The non-grazed sites of this study were not nearly as dense as those grasslands in the study by Greenslade (1964), which had a litter layer of about 20 cm. Despite these limitations, pitfall trapping was the best method of collecting Orthoptera for this study. The orthopteran total catch and number of species caught were similar to other studies. Sweep netting would not have been a feasible alternative in the antelope-brush dominated study sites and would not have captured cricket species. Pitfall trapping was continuous for two seasons (no significant differences were found between the two years, hence the data from the two seasons were combined), thus eliminating temporal differences in capture rates while sufficiently sampling the orthopteran assemblage in the study sites. As trapping effort was equal among all ten study sites, it is assumed that trap efficiency for each species and sex was equal among the study sites. Therefore, this sampling regime enables accurate intersite comparisons of orthopteran communities. A study similar to this took place on revegetated coal mines in Wyoming (Parmenter et al. 1991). Pitfall trapping was the primary collection method used in this shrub-steppe community, which was also dominated by large shrubs (i.e., big sagebrush, low sagebrush (Artemisia arbuscula Nutt), antelope-brush). They too examined the cricket community in addition to grasshoppers as pitfall trapping collected crickets in sufficient numbers. They checked for potential bias of using pitfall traps by visually noting relative abundances of each species while 70 servicing pitfall traps, the results of which were similar to those from pitfall trapping. I also checked for potential bias of pitfall trapping with sweep netting (at areas without antelope-brush) and visual observations and found them to coincide with the pitfall trap data. 4.2 Orthopteran Study in the South Okanagan Minimal research into the grasshopper assemblage of the South Okanagan has been done since the early collections of Orthoptera (before 1958) by E.R. Buckell, G.J. Spencer, and R.C. Treherne (Scudder 1996). An overview of British Columbia's orthopteran ecology took place in the early 1970's (Vickery and Nagy 1973) and a survey of arthropods in the Osoyoos-Mt. Kobau area occurred in 1991 (Blades and Maier 1996). 4.2.1 Orthopteran Species Descriptions The seventeen grasshopper species collected in this study are all known to occur in British Columbia's South Okanagan (Vickery and Kevan 1985). Seventeen species of grasshoppers were also collected in the arthropod survey in the Osoyoos- Mt. Kobau area, although only seven were identical species (Blades and Maier 1996). This result is not surprising, as only a small portion of the collecting was done in the same ecosystem as this study (antelope-brush). Two grasshopper species collected in this study are listed as potentially rare and endangered invertebrates, Phoetaliotes nebrascensis and Aulocara elliotti (Scudder 1984). In British Columbia, P. nebrascensis is known only to occur in Okanagan Falls and Oliver, although it has a much greater range into the United States (Scudder 1984). Similar to P. nebrascensis, A. elliotti is known only to occur in the South Okanagan Valley in British Columbia, while its range extends far greater in the southern United States. It can be a serious 71 pest on grasses and can contribute significantly to grasshopper outbreaks on rangelands (Pfadt 2002). In contrast with the grasshoppers, the taxonomy of several of the captured cricket species are in a state of flux and currently being revised. Both the genus Gryllus and the genus Stenopelmatus are currently being revised by Dr. David Weissman (California Academy of Science). The two species identified in this study, Gryllus sp. and Stenopelmatus sp., may be new, undescribed species (personal communication through Dr. Scudder, 2000). The taxonomy of the genus Steiroxys is also currently being revised (Vickery and Kevan 1985). The family Rhaphidophoridae is currently being revised by Dr. T. Cohn (University of Michigan). N . sp. dasy is a relative of Pristoceuthophilus pacificus (n. sp. dasy incorrectly keys out to be P. pacificus). P. pacificus is not found north of central of Washington. New species dasy has the same peculiarly curved tibia and two elongate subdistal teeth on the hind femur as P. pacificus, but has a distinct phallus which requires a special preparation to distinguish. New species dasy is found from the Southern Sierra Nevada, up to the middle of California, where it then swings eastward and up the desert mountains north to Eastern B.C. and east to Glacier Park in Montana. 4.3 Descriptive Statistics The number of grasshopper species captured (17 species) in this study were similar to other grasshopper studies in North American grasslands (Fielding and Brusven 1995b; Holmes et al. 1979; Parmenter et al. 1991; Pfadt 1984). Five grasshopper species comprised 97% of the total grasshopper catch, with M. sanguinipes sanguinipes as the dominant species (nearly 80% of total grasshopper catch). These results are also similar to related work, where a few 72 grasshopper species make up the majority of the total catch (Fielding and Brusven 1993b; Holmes et al. 1979; Miller and Onsager 1991) and M. sanguinipes sanguinipes is the dominant grasshopper species (Porter 1995) (Fielding and Brusven 1993b; Fielding and Brusven 1995b; Parmenter et al. 1991; Pfadt 1982). It is not surprising that M. sanguinipes sanguinipes is so often found to be a dominant species as it is a generalist that is successful in a variety of habitats (e.g. grasslands, overgrazed rangelands, meadows), is widely distributed in North America (occurs in every province in Canada and every state in the United States), can thrive at high densities on heavily disturbed areas, and is a mixed feeder of both forbs and grasses with a wide range of host plants (Pfadt 2002). Additionally, M. sanguinipes sanguinipes is very mobile, hence its common name, the migratory grasshopper (Pfadt 1982). As expected, the most abundant grasshopper species, M. sanguinipes sanguinipes, and the most abundant cricket species, Gryllus sp., were found in all ten study sites. Orthoptera abundance peaked in June 1994 and August 1995 (Fig. 3). As expected, grasshopper numbers increased during late spring and early summer (as nymphs hatched) and reached their peak during late summer, when most nymphs have hatched and matured into adults (Fig. 4). The number of species of Orthoptera also peaked in late summer, coinciding with the peak in adult grasshoppers. The high number of Orthoptera in June and July of 1994 (Fig. 3) were due to very high numbers of crickets, specifically Gryllus sp., at site V. Besides the high cricket numbers in June and July of 1994, cricket abundance did not vary greatly throughout the summer. This is similar to the study on revegetated strip mines, when cricket numbers were seen to slightly increase throughout the summer season (Parmenter et al. 1991). 73 The average number of Orthoptera caught per grazing regime did not differ significantly across grazing regimes. Site V appears to be an outlier, for unknown reasons. The extremely high number of Gryllus sp. in June and July of 1994 is not seen in 1995. Site V still has the greatest number of Gryllus sp. captured compared to the other sites, but not overwhelmingly so. Additionally, M. sanguinipes sanguinipes was found to be most abundant at site V in both years, more than triple the number of individuals at the next most abundant site. Site V has a large number of closely packed boulders, which may provide good shelter for both these species. Additionally, a water source is nearby site V . However, the relative abundance of other grasshopper and cricket species is not elevated at site V. There does not appear to be simple explanation for the high numbers of the two most common Orthopterans captured during this study at site V . The null hypothesis that grasshopper abundance is unaffected by grazing regime cannot be rejected. 4.4 Biodiversity Measures The results of the calculated biodiversity indices do not show trends across grazing regime. Species richness, evenness, the Shannon-Wiener Index, and Simpson's Index do not differ significantly across the different grazing regimes. While not statistically significant, species richness does appear to slightly increase with increased grazing levels (Table 8). The 95% probability funnels calculated for average taxonomic distinctness and variation in taxonomic distinctness also show no significant differences between study sites. For average taxonomic distinctness, all sites fall within the 95% probability funnel. For variation in 74 taxonomic distinctness, all sites except site X fall within the 95% probability funnel. Site X falls just outside the lower limit of the probability funnel. This means that a lower than expected variation in taxonomic distinctness of species pairs occurs at site X . However, no trend of a particular grazing regime falling in the lower limits of this probability funnel occurs. The ellipse plot (Fig. 10) shows no significant departure from the null hypothesis that the average taxonomic distinctness and variation in taxonomic distinctness pairs have a taxonomic structure representative from a random sample, as no study sites fall outside their respective 95% probability envelope. These biodiversity indices show that the null hypothesis that grasshopper species composition is unaffected by grazing regime cannot be rejected. 4.5 Cluster Analysis and non-metric Multidimensional Scaling Cluster analysis and non-metric multi-dimensional scaling also show that grazing levels do not affect orthopteran species composition, as no groupings of sites by grazing level is seen. Site V appears to be an outlier in both dendrograms and ordination plots, but this is explained above from the high numbers of Gryllus sp. in late spring of 1994. 4.6 Vegetation Overlays, Correlation Analyses and Regression Analyses The results of the vegetation overlays do not show a relationship between percent plant cover and plant species richness with site groupings. Additionally, correlation analyses are non-significant between orthopteran data and percent plant cover or plant species richness. Many studies have shown a positive relationship between grasshopper species richness and plant species richness (Evans 1988; Fielding and Brusven 1993a; Kemp 1992; Quinn and 75 Walgenbach 1990) and some studies have shown a positive correlation between grasshopper diversity and plant species diversity (Kemp 1992; Quinn and Walgenbach 1990). The lack of observed relationship in this particular study can be explained by the fact that plant species richness did not differ significantly between the grazing regimes (Krannitz 1999). Additionally, introduced alien species and native plant species did not differ significantly between the grazing regimes. In the antelope-brush ecosystem, the different grazing intensities of this study did not affect plant species richness. Heavily grazed sites were found to have more significantly more bare soil than non-grazed or moderately grazed sites (Krannitz 1999). The null hypothesis that changes in orthopteran abundance and species diversity will be related to changes in the plant community cannot be rejected. The vegetation overlays do suggest that percent bare soil is greater on heavily grazed sites than the other sites. Correlation and regression analyses then reveal a significant, positive relationship between the number of orthopteran species and the Shannon-Wiener Index results with percent bare soil. For this study, percent bare soil is not simply an inverse of percent plant cover, owing to the cryptogam crust cover, which was calculated separately from percent plant cover (Dr. P. Krannitz, personal communication, 2000). A study in Montana revealed that grasshoppers preferred areas with less than 40% plant cover and that the number of grasshopper species increased with increasing bare soil (Anderson 1964). The observed increase in orthopteran species number with increasing bare soil of this study is likely due to the preference of the majority of grasshopper and cricket species of ovipositing on bare or exposed soil (Kevan 1989; Quinn and Walgenbach 1990; Rentz 1996; Smith 1940). 76 4.7 Effects of Grazing on Grasshoppers in Different Grassland Types Different studies of grasshopper responses to different grazing levels have yielded contradictory results. As discussed in the introduction, some studies have shown grasshopper abundances increase with grazing, others have shown grasshopper abundances decrease with grazing, and still others have shown grazing has no effect on grasshopper abundances. Little discussion about the influence of grassland type on grasshopper responses has been reviewed in literature. A l l seven studies that examined different grazing levels and grasshopper densities on mixed-grass prairie or tall-grass prairie reported that grasshopper populations were greater on heavily grazed areas than non-grazed or lightly grazed area (Coyner 1938; Smith 1940; Weese 1939) (Campbell et al. 1974; Joern 2004; Knutson and Campbell 1974; Quinn and Walgenbach 1990). These studies all identified profound changes to the plant community caused by grazing, including decreases in tall grass species, increases in forbs and short grass species, and decreases in percentage of plant cover, which led to changes in the grasshopper densities. In Alberta's fescue prairie, which is moister than mixed-grass prairie, grasshopper density was also greater on heavily grazed sites than lightly grazed sites (Holmes et al. 1979). This study did not, however, examine the corresponding plant community. In contrast, three studies on short-grass prairie found that grasshopper densities were greater on non-grazed or lightly grazed pastures than on heavily grazed pastures (Capinera and Sechrist 1982; Pfadt 1982; Welch et al. 1991). Two of these studies identified the prairie flora as the source of variation in grasshopper densities, while the third did not examine the plant 77 community (Welch et al. 1991). Unlike the tall-grass prairies, the short-grass prairie flora of northeastern Colorado was not extensively altered by grazing (Capinera and Sechrist 1982). The greater densities of grasshoppers on ungrazed or lightly grazed pastures were due to the higher plant biomass of these pastures (Capinera and Sechrist 1982). The other study on short-grass prairie identified plant species composition as the cause of differences in grasshopper densities (Pfadt 1982). Heavily grazed sites could not support high grasshopper densities because they had a greater proportion of annual vegetation that matured and dried by the middle of May. In contrast, lightly grazed or ungrazed sites had a greater proportion of perennial grasses that served as green host plants for grasshoppers through the summer and fall months. An additional study on short-grass prairie showed no effects of grazing on grasshopper populations, nor were the populations influenced by reductions in plant cover or increases in bare ground (Miller and Onsager 1991). This is similar to the results of my study, except for the effects of bare ground. A likely explanation for the different responses of grasshoppers on different grasslands is that the effects of heavy grazing are more severe on tall-grass and mixed-grass prairie than short-grass prairie, due to the greater moisture levels and greater vegetation structure heterogeneity found on tall-grass prairies. Essentially, the difference between a heavily grazed and non-grazed tall-grass prairie would be much greater than the difference between a heavily grazed and non-grazed short-grass prairie, which already has lower plant productivity levels (Joern 2004). It is likely that grasshopper increase after heavy grazing on tall-grass and mixed-grass prairies due to the increased availability of favourable habitat conditions. In contrast, on the short-grass prairies, the heavy grazing likely creates an environment so extreme that it no 78 longer favours grasshopper populations (e.g. significantly lower plant biomass or lack of green host plants). As the plant community was not severely disturbed in my study, these above findings would explain why the orthopteran species composition did not differ significantly across the different grazing regimes. It would appear that severe grazing levels are required to cause significant changes to the grasshopper community in the South Okanagan. Studies in other countries have also shown varied results of grasshopper responses to grazing. Of the two studies that took place in Hungary, one showed grasshopper density was greatest on the most heavily grazed site while species richness and diversity were greatest on the ungrazed site (Baldi and Kisbenedek 1997) while in another study grasshopper density and species composition did not differ between ungrazed and heavily grazed sites (Kisbenedek 1995). In a South African savanna, no difference in grasshopper abundance was observed between grazed and heavily grazed areas (Prendini et al. 1996). In the sagebrush-grass region of the intermountain west, which is dominated by sagebrush and bunchgrass, two studies found higher densities of grasshoppers on heavily disturbed sites than on undisturbed sites (grazing was examined in one study, revegetated surface coal mines in the other) (Fielding and Brusven 1993a; Parmenter et al. 1991). Another study showed grasshopper densities to be higher on ungrazed sites versus heavily grazed sites under drought conditions (Fielding and Brusven 1995b). While a clear pattern of grasshopper response to disturbance or grazing is not apparent from these studies in other countries and the intermountain west, it is likely due to differences in 79 the effect of disturbance or grazing on the plant community, and hence the grasshopper community, as discussed earlier. For example, the South African study found that grasshopper abundance, species richness, and diversity were lower on a mowed area compared to the grazed areas, even though no differences were observed between the different grazing levels (Prendini et al. 1996). The habitat and microclimate change from mowing is much greater than that of the two grazing levels (Prendini et al. 1996), which would explain the significant affect mowing had on grasshoppers. Similarly, the study on reclaimed surface mines showed higher orthopteran species richness and diversity on undisturbed shrub-steppe compared to the disturbed shrub-steppe (Parmenter et al. 1991). The plant community was significantly different between the undisturbed and disturbed areas of this study. In a study of the grasshopper community of a California native grassland, currently managed for the preservation of perennial bunchgrasses, only five grasshopper species were found (Porter et al. 1996). This species richness number is much lower than my results and those of several other studies listed earlier. Porter (1996) suggested that the grasshopper community of this grassland was particularly depauperate because the grasslands were isolated and have been highly disturbed (from heavy grazing and invasion of exotic annual grasses and forbs) from their climax condition. This is another example of low grasshopper species richness on extremely disturbed grasslands. Only some of these other studies investigated grasshopper species richness and diversity. Some found that grasshopper species richness and diversity were greater on the undisturbed sites while grasshopper densities were greater on the disturbed sites (Baldi and Kisbenedek 80 1997; Parmenter et al. 1991; Prendini et al. 1996). Parementer (1991) explains that the plant community of their highly disturbed sites were relatively homogeneous (had lower richness and diversity values), allowing the populations of certain grasshopper and camel cricket species, which preferred the plant community of these disturbed areas, to proliferate. This explanation appears to well-apply for the other two studies. In the short-grass prairie study where grasshopper densities were greater on non-grazed or lightly grazed sites due to the availability of green plants, grasshopper species richness and diversity were also greater at these sites (Pfadt 1982). From the work of many researchers, it appears that grasshoppers respond to the effects of grazing based on the severity of the grazing effects on the plant community. Each study defines their grazing parameters differently from others. Even if further studies were to use standard grazing parameters, the effects of these standard parameters would differ on different grassland types, as different grasslands have different vulnerabilities. The findings of other researchers taking part in the South Okanagan Grassland Conservation Research Project were varied. Ant abundance was found to be greatest on heavily grazed sites and increased with percent bare soil (Heron 2001). Mice abundance (Dr. G.G.E. Scudder, personal communication, 2005) and antelope-brush dieback were not associated with grazing intensity (Krannitz 1994). Seed weight variability of antelope-brush did not vary with grazing (Krannitz 1997). Density, percent frequency and percent cover of two invasive weeds, cheatgrass and knapweed (Centaurea diffusa Lam.), were not associated with grazing history (Krannitz 1994). Krannitz (1994) described that the organization of the antelope-brush 81 ecosystem appeared to be most affected by grazing intensity, not biodiversity. Two variables that describe the structure of each site, percent exposed soil and individual bunchgrass species sizes, did vary with grazing. Percent exposed soil increased with grazing while the size of individual bunchgrass species decreased with grazing. In my study, grazing likely did not affect orthopteran species abundance or species diversity because the grazing intensities were not extreme enough to cause changes to the orthopteran species assemblage. This is supported by a lack of effect of grazing on the plant community in the areas of plant species richness, introduced alien species richness, native plant species richness, and percent cover of grasses (Krannitz 1999). As discussed above, the results of many studies support the idea that only severe disturbances or grazing which cause a significant change to the plant community then create significant changes to the grasshopper community. 82 5 CONCLUSION Seventeen grasshopper species and seven cricket species were captured in this study. The number of orthopteran species increased with increasing percent bare soil. Orthopteran abundance arid diversity did not vary across the different grazing regimes. The lack of effect of grazing on the orthopteran species assemblage was probably due to the lack of effect of grazing on the plant community species assemblage. More extreme levels of grazing are likely needed to elicit a response from the orthopteran species assemblage. 83 LITERATURE CITED Alldritt-McDowell, J., and R. Coupe. 1998. The ecology of the bunchgrass zone, Biogeoclimatic Zones of British Columbia Series, B.C. Ministry of Forests, Research Branch, Victoria. Anderson, D. C , K. T. Harper, and S. R. Rushforth. 1982. Recovery of cryptogamic soil crusts in Utah deserts. Journal of Range Management 35: 355-359. Anderson, D. C , and S. R. Rushforth. 1977. The cryptogamic flora of desert soil crusts in southern Utah. Nova Hedwigia 28: 691-729. Anderson, N . L . 1964. Some relationships between grasshoppers and vegetation. Annals of the Entomological Society of America 57: 736-742. Anderson, N . L. , and J. C. Wright. 1952. Grasshopper investigations on Montana grasslands. Montana Agricultural Experiment Station, Technical Bulletin 486: 46 pp. Baldi, A . , and T. Kisbenedek. 1997. Orthopteran assemblages as indicators of grassland naturalness in Hungary. Agriculture Ecosystems & Environment 66: 121-129. Blades, D. C. A. , and C. W. Maier. 1996. A survey of grassland and montane arthropods collected in the southern Okanagan region of British Columbia. Journal of the Entomological Society of British Columbia 93: 49-73. Breymeyer, A . I., and G. M . van Dyne. 1980. Grasslands, Systems Analysis, and Management. Cambridge University Press, New York, New York. Brink, V . C. 1982. What is Grassland? pp.21-26 in A . C. Nicholson, A. McLean, and T. E. Baker, (eds). Grassland Ecology and Classification Symposium Proceedings. June 2-4, 1982. Kamloops, BC. Brooks, A . R. 1958. Acridoidea of Southern Alberta, Saskatchewan, and Manitoba (Orthoptera). The Canadian Entomologist Supplement 9 25: 3-92. Brotherson, J. D., and S. R. Rushforth. 1983. Influence of cryptogamic crusts on moisture relationships in Navajo National Monument, Arizona. Great Basin Naturalist 43: 73-78. Cameron, R. E., and W. H. Fuller. 1960. Nitrogen fixation by some algae in Arizona soils. Proceedings - Soil Science: Society of America 24: 353-356. Campbell, J. B., W. H. Arnett, J. D. Lambley, O. K. Jantz, and H. Knutson. 1974. Grasshoppers (Acrididae) of the Flint Hills native tall grass prairie in Kansas. Kansas Agricultural Experiment Station Research Paper 19: 147 pp. 84 Capinera, J. L. , and T. S. Sechrist. 1982. Grasshopper (Acrididae) - host plant associations: response of grasshopper populations to cattle grazing intensity. The Canadian Entomologist 114: 1055-1062. Carder, A . C. 1970. Climate and rangelands of Canada. Journal of Range Management 23: 263-267. Clarke, K. R., and R. N . Gorley. 2001. PRIMER v5: User Manual/Tutorial. PRIMER-E Ltd, Plymouth. Clarke, K. R., and R. M . Warwick. 1994. Change in Marine Communities: A n Approach to Statistical Analysis and Interpretation. Plymouth Marine Laboratory, Plymouth. Clarke, K. R., and R. M . Warwick. 1998. A taxonomic distinctness index and its statistical properties. Journal of Applied Ecology 35: 523-531. Clarke, K . R., and R. M . Warwick. 2001. A further biodiversity index applicable to species lists: variation in taxonomic distinctness. Marine Ecology Progress Series: 265-278. Coupland, R. T. 1992. Mixed Prairie, pp. 151-182 in R. T. Coupland, (ed). Ecosystems of the World 8A: Natural Grasslands: Introduction and Western Hemisphere. Elsevier, Amsterdam. Coyner, W. R. 1938. Insect distribution and seasonal succession in overgrazed and normal grasslands. M.S. Thesis, University of Oklahoma, Norman: 78 pp. Crumpacker, D. W. 1984. Regional Riparian Research and a Multi-University Approach to the Special Problem of Livestock Grazing in the Rocky Mountains and Great Plains, pp. 413-422 in R. E. Warner, and K. Hendrix, (eds.) Californian Riparian systems: Ecology, Conservation, and Productive Management. University of California Press, Berkeley. Daubenmire, R. F. 1978. Plant Geography: with Special Reference to North America. Academic Press, New York. Delong, D. M . 1932. Some problems encountered in the estimation of insect populations by the sweeping method. Annals Entomological Society of America 25: 13-17. Dietrick, E. J. 1961. An improved back pack motor fan for suction sampling of insect populations. Journal of Economic Entomology 54: 394-395. Dyer, O., and E. C. Lea. 2003. Status and importance of the antelope-brush - needle-and-thread grass plant community in the South Okanagan Valley, British Columbia, pp. 13-18 in R. Seaton, (ed). Proceedings, Ecosystem at Risk: Antelope Brush Restoration. March 28-30, 2003. Osoyoos, BC. 85 Evans, E. W. 1987. Grasshopper (Insecta: Orthoptera: Acrididae) assemblages of tallgrass prairie: influences of fire frequency, topography, and vegetation. Canadian Journal of Zoology 66: 1495-1501. Evans, E. W. 1988. Community dynamics of prairie grasshoppers subjected to periodic fire: predictable trajectories or random walks in time? Oikos 52: 283-292. Evans, E. W., and K . W. Bailey. 1993. Sampling grasshoppers (Orthoptera: Acrididae) in Utah grasslands: pan trapping versus sweep sampling. Journal of the Kansas Entomological Society 66: 214-222. Evans, E. W., R. A. Rogers, and D. J. Opfermann. 1983. Sampling grasshoppers (Orthoptera: Acrididae) on burned and unburned tallgrass prairie: night trapping vs. sweeping. Environmental Entomology 12: 1449-1454. Fielding, D. J., M . Brusven, B. Shafii, and W. Price. 2001. Spatial heterogeneity of low-density populations of Melanoplus sanguinipes (Orthoptera: Acrididae) associated with grazing and vegetation treatments. The Canadian Entomologist 133: 843-855. Fielding, D. J., and M . A. Brusven. 1993a. Grasshopper (Orthoptera: Acrididae) community composition and ecological disturbance on Southern Idaho rangeland. Environmental Entomology 22: 71-81. Fielding, D. J., and M . A . Brusven. 1993b. Spatial analysis of grasshopper density and ecological disturbance on southern Idaho rangeland. Agriculture, Ecosystems and Environment 43: 31-47. Fielding, D. J., and M . A. Brusven. 1995a. Ecological correlates between rangeland grasshopper (Orthoptera: Acrididae) and plant communities of southern Idaho. Environmental Entomology 24: 1432-1441. Fielding, D. J., and M . A. Brusven. 1995b. Grasshopper densities on grazed and ungrazed rangeland under drought conditions in southern Idaho. Great Basin Naturalist 55: 352-358. Fleischner, T. L. 1994. Ecological costs of livestock grazing in Western North America. Conservation Biology 8: 629-644. Greenslade, P. J. M . 1964. Pitfall trapping as a method for studying populations of Carabidae (Coleoptera). Journal of Animal Ecology 33: 301-310. Heron, J. 2001. The effect of grazing on ant diversity in the south Okanagan grasslands, British Columbia. M.Sc. Thesis, University of British Columbia, Vancouver: 98 pp. Hewitt, G. B. 1977. Review of forage losses caused by rangeland grasshoppers. U S D A Miscellaneous Publication 1348: 24 pp. 86 Hewitt, G. B., and J. A . Onsager. 1988. Effects of sagebrush removal and legume interseeding on rangeland grasshopper populations (Orthoptera: Acrididae). The Canadian Entomologist 120: 753-758. Hewitt, G. B., andN. E. Rees. 1974. Abundance and distribution of grasshoppers (Acrididae: Orthoptera) in relation to rangeland renovation practices. Journal of Range Management 27: 156-160. Holmes, N . D., D. S. Smith, and A. Johnston. 1979. Effect of grazing by cattle on the abundance of grasshoppers on fescue grassland. Journal of Range Management 32: 310-311. Hubbell, T. H. 1936. A monographic revision of the genus Ceuthophilus (Orthoptera: Gryllacrididae: Rhapiophorinae). University of Florida Biological Series 2: 1-551. Hutchinson, K. , and K . L . King. 1970. Sheep numbers and soil arthropods. Search 1: 42-43. Jepson-Innes, K. , and C. E. Bock. 1989. Response of grasshoppers (Orthoptera: Acrididae) to livestock grazing in southeastern Arizona: differences between seasons and subfamilies. Oecologia78: 430-431. Joern, A . 1979. Resource utilization and community structure in assemblages of arid grassland grasshoppers (Orthoptera: Acrididae). Transactions of the American Entomological Society 105: 253-300. Joern, A . 1982. Distributions, densities and relative abundances of grasshoppers (Orthoptera: Acrididae) in a Nebraska sandhills prairie. Prairie Naturalist 14: 37-45. Joern, A . 2004. Variation in grasshopper (Acrididae) densities in response to fire frequency and bison grazing in tallgrass prairie. Environmental Entomology 33: 1617-1625. Kemp, W. P. 1992. Rangeland grasshopper (Orthoptera: Acrididae) community structure: a working hypothesis. Environmental Entomology 21: 461-470. Kemp, W. P., S. J. Harvey, and K. M . O'Neill. 1990. Patterns of vegetation and grasshopper community composition. Oecologia 83: 299-303. Kevan, D. K. M . 1989. Grigs that dig and grasshoppers that grovel. Revue d'Ecologie et de Biologie du Sol 26: 267-289. Kisbenedek, T. 1995. The effect of sheep grazing on the community structure of grasshoppers (Orthoptera). Folia Entomologica Hungarica 56: 45-56. Knopf, J. A . E. 1957. Relation of grasshopper populations to grazing intensity at the Eastern Colorado Range Station. M.S. Thesis, Colorado State University, Fort Collins: 106 pp. 87 Knutson, H. , and J. B. Campbell. 1974. Relationships of grasshoppers (Acrididae) to burning, grazing, and range sites of native tallgrass prairie in Kansas. Proceedings of the Tall Timbers Conference on Ecological Animal Control by Habitat Management 6: 107-120. Krannitz, P. 1994. South Okanagan Grassland Conservations Research: Progress Report. Pacific Wildlife Research Centre, Canadian Wildlife Service: 40 pp. Krannitz, P. 1997. Seed weight variability of Antelope Bitterbrush (Purshia tridentata: Rosaceae). American Midland Naturalist 138: 306-321. Krannitz, P. 1999. Effects of cattle and horse grazing on the herbaceous plant community associated with antelope-brush ecosystem, unpublished manuscript. Krebs, C. J. 1999. Ecological Methodology. Addison-Wesley Educational Publishers, Inc., New York. Kruskal, J. B. 1964. Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika 29: 1-27. Lavallee, S. L. 1999. Changes in the carabid beetle community of the Sicamous Creek Research Site in response to prescribed logging practices. M.Sc. Thesis, University of British Columbia, Vancouver: 72 pp. Luff, M . L. 1975. Some features influencing the efficiency of pitfall traps. Oecologia 19: 435-357. Mack, R. N . , and J. N . Thompson. 1982. Evolution in steppe with few large, hoofed animals. American Naturalist 119: 757-773. McLean, A. , and L. Marchand. 1968. Grassland Ranges in the southern interior of British Columbia, Canadian Department of Agriculture Publication, Ottawa. McLean, A. , and E. W. Tisdale. 1972. Recovery rate of depleted range sites under protection from grazing. Journal of Range Management 25: 178-184. Melbourne, B. A. , P. J. Gullan, and Y . N . Su. 1997. Interpreting data from pitfall-trap surveys: crickets and slugs in exotic and native grasslands of the Australian Capital Territory. Memoirs of the Museum of Victoria 56: 361-367. Miller, R. H. , and J. A . Onsager. 1991. Grasshopper (Orthoptera: Acrididae) and plant relationships under different grazing intensities. Environmental Entomology 20: 807-814. Mitchell, J. E., and R. E. Pfadt. 1974. The role of grasshoppers in a short-grass prairie ecosystem. Environmental Entomology 3: 358-360. 88 Mochida, K. , and M . Gomyoda. 1987. Toxicity of ethylene glycol, diethylene glycol, and propylene glycol to human cells in culture. Bulletin of Environmental Contaminant Toxicology 38: 151-153. Morris, M . G. 1967. Differences between invertebrate faunas of grazed and ungrazed chalk grassland, 1. Journal of Applied Ecology 5: 601-611. Nerney, N . J. 1958. Grasshopper infestations in relation to range condition. Journal of Range Management 11: 247-252. Nicholson, A. , E. Hamilton, W. L. Harper, and B. M . Wikeem. 1991. Bunchgrass Zone, pp. 125-137 in D. Meidinger, and J. Pojar, (eds.) Ecosystems of British Columbia. B.C. Ministry of Forests Special Report Series 6, Victoria, B.C. Onsager, J. A . 1977. Comparison of five methods for estimating density of rangeland grasshoppers. Journal of Economic Entomology 70: 187-190. Onsager, J. A. , and J. E. Henry. 1977. A method for estimating the density of rangeland grasshoppers (Orthoptera, Acrididae) in experimental plots. Acrida (Paris) 6: 231-237. Otte, D. 1976. Species richness patterns of New World desert grasshoppers in relation to plant diversity. Journal of Biogeography 3: 197-209. Otte, D. 1981. The North American Grasshoppers. Harvard University Press, Cambridge. Parmenter, R. R., J. A . Macmahon, and C. A. B. Gilbert. 1991. Early successional patterns of arthropod recolonization on reclaimed Wyoming strip mines: the grasshoppers (Orthoptera: Acrididae) and allied faunas (Orthoptera: Gryllacrididae, Tettigoniidae). Environmental Entomology 20: 135-142. Pepper, J. H . 1955. The ecological approach to management of insect populations. Journal of Economic Entomology 48: 451-456. Pfadt, R. E. 1982. Density and diversity of grasshoppers (Orthoptera: Acrididae) in an outbreak on Arizona rangeland. Environmental Entomologist 11: 690-694. Pfadt, R. E. 1984. Species richness, density, and diversity of grasshoppers (Orthoptera: Acrididae) in a habitat of the mixed grass prairie. The Canadian Entomologist 116: 703-709. Pfadt, R. E. 2002. Field Guide to Common Western Grasshoppers. Wyoming Agricultural Experiment Station, University of Wyoming, Laramie. 89 Pitt, M . , and T. D. Hooper. 1994. Threats to Biodiversity of Grasslands in British Columbia, pp. 279-292 in L. E. Harding, and E. McCullum, (eds.) Biodiversity in British Columbia: Our Changing Environment. Environment Canada, Canadian Wildlife Service, Vancouver. Porter, E. E. 1995. The grasshoppers of a California native grassland: a description of the community and its ecological importance. M.S. Thesis, University of California, Riverside: 105 pp. Porter, E. E., R. A . Redak, and H. E. Braker. 1996. Density, biomass, and diversity of grasshoppers (Orthoptera: Acrididae) in a California native grassland. Great Basin Naturalist 56: 172-176. Prendini, L., L.-J . Theron, K. Van Der Merwe, and N . Owen-Smith. 1996. Abundance and guild structure of grasshoppers (Orthoptera: Acridoidea) in communally grazed and protected savanna. South African Journal of Zoology 31: 120-130. Quinn, M . A. , R. L. Kepner, D. D. Walgenbach, R. A . Bohls, and P. D. Pooler. 1991. Habitat characteristics and grasshopper community dynamics on mixed-grass rangeland. The Canadian Entomologist 123: 89-105. Quinn, M . A. , and D. D. Walgenbach. 1990. Influence of grazing history on the community structure of grasshoppers of a mixed grass prairie. Environmental Entomology 19: 1756-1766. Rentz, D. C. F. 1996. Grasshopper Country: The Abundant Orthopteroid Insects of Australia. University of New South Wales Press, Sydney. Schluter, A. , T. Lea, S. Cannings, and P. Krannitz. 1995. Antelope-brush ecosystems, Ecosystems at Risk in British Columbia Series, Province of British Columbia Ministry of Environment, Lands, and Parks, Victoria. Scoggan, A . C , and M . A . Brusven. 1973. Grasshopper-plant community associations in Idaho in relation to the natural and altered environment. Melanderia 12: 22-33. Scudder, G. G. E. 1984. An annotated systematic list of the potentially rare and endangered freshwater and terrestrial invertebrates in British Columbia. Entomological Society of British Columbia, Occasional Paper 2: 1-92. Scudder, G. G. E. 1996. Terrestrial and freshwater invertebrates of British Columbia: priorities for inventory and descriptive research, Research Branch, B.C. Ministry of Forests, and Wildlife Branch, B.C. Ministry of Environment, Lands, and Parks, Victoria. Scudder, G. G. E. 2000. Osoyoos Desert Society: Experimental studies on ecological restoration of the shrub-steppe habitat, Internal Report. 90 Smalley, A . E. 1960. Energy flow of a salt marsh grasshopper population. Ecology 41: 672-677. Smith, C. C. 1940. The effect of overgrazing and erosion upon the biota of the mixed-grass prairie of Oklahoma. Ecology 21: 381-397. Southwood, T. R. E. 1978. Ecological Methods with Particular Reference to the Study of Insect Populations. Chapman and Hall, London. Spence, J. R., and J. K . Niemala. 1994. Sampling carabid assemblages with pitfall traps: the madness and the method. The Canadian Entomologist 126: 881-894. Tisdale, E. W. 1947. The grasslands of the Southern Interior of British Columbia. Ecology 28: 346-365. Titlyanova, A . A. , R. I. Zlotin, and N . R. French. 1990. Changes in Structure and Function of Temperate-Zone Grasslands Under the Influence of Man, pp. 301-334 in A. I. Breymeyer, (ed). Ecosystems of the World 17A: Managed Grasslands Regional Studies. Elsevier, Amsterdam. Topping, C. J. 1993. Behavioural responses of three linyphiid spiders to pitfall traps. Entomologia Experimentalis et Applicata 68: 287-293. Topping, C. J., and K. D. Sunderland. 1992. Limitations to the use of pitfall traps in ecological studies exemplified by a study of spiders in a field of winter wheat. Journal of Applied Ecology 29: 485-491. •Urbanek, R. P. 1982. Arthropod community structure on strip-mine lands in Ohio. Ph.D. Thesis, Ohio State University, Columbus: 187 pp. Van Jaarsveld, A. S., J. W. H. Ferguson, and G. J. Bredenkamp. 1998. The Groenvaly grassland fragmentation experiment: design and initiation. Agriculture, Ecosystems and Environment 68: 139-150. Vickery, V . R. 1987. The Orthopteroid insects in Northern North America: updating and rationalizing recent work. The Canadian Entomologist 119: 1043-1054. Vickery, V . R., and D. K . M . Kevan. 1985. The Grasshoppers, Crickets, and Related Insects of Canada and Adjacent Regions: Insects and Arachnids of Canada; Part 14. Research Branch Agriculture Canada, Ottawa. Vickery, V . R., and B. Nagy. 1973. Ecological notes on Orthoptera (s.str.) in British Columbia. Journal of the Entomological Society of British Columbia 70: 27-33. Wagner, F. H. 1978. Livestock Grazing and the Livestock Industry, pp. 121-145 in H . P. Brokaw, (ed). Wildlife and America. Council on Environmental Quality, Washington, D.C. 91 Warwick, R. M . , and K. R. Clarke. 1995. New "biodiversity" measures reveal a decrease in taxonomic distinctness with increasing stress. Marine Ecology Progress Series 129: 301-305. Warwick, R. M . , and K . R. Clarke. 2001. Practical measures of marine biodiversity based on relatedness of species. Oceanography and Marine Biology. Annual Review 39: 207-231. Weese, A . O. 1939. The effect of overgrazing on insect populations. Proceedings of the Oklahoma Academy of Science 19: 95-99. Welch, J. L. , R. Redak, and B. C. Kondratieff. 1991. Effect of cattle grazing on the density and species of grasshoppers (Orthoptera: Acrididae) of the Central Plains Experimental Range, Colorado: a reassessment after two decades. Journal of the Kansas Entomological Society 64: 337-343. West, N . E. 1983. Overview of North American Temperate Deserts and Semi-Deserts, pp. 321-330 in N . E. West, (ed). Ecosystems of the World 5: Temperate Deserts and Semi-Deserts. Elsevier Scientific Publishing Company, Amsterdam. Willms, W. D., S. Smolak, and J. F. Dormaar. 1985. Effects of stocking rates on rough fescue grassland vegetation. Journal of Range Management 38: 220-225. Wilson, E. O. 1992. The Diversity of Life. Harvard University Press, Cambridge, Massachusetts. Zar, J. H . 1984. Biostatistical Analysis. Prentice-Hall, Inc, Englewood Cliffs, New Jersey. 92 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092083/manifest

Comment

Related Items