UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Breeding bird communities and habitat associations in the grasslands of the Chilcotin Region, British… Hooper, Tracey D 1994

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

Item Metadata


831-ubc_1994-0522.pdf [ 1.83MB ]
JSON: 831-1.0087683.json
JSON-LD: 831-1.0087683-ld.json
RDF/XML (Pretty): 831-1.0087683-rdf.xml
RDF/JSON: 831-1.0087683-rdf.json
Turtle: 831-1.0087683-turtle.txt
N-Triples: 831-1.0087683-rdf-ntriples.txt
Original Record: 831-1.0087683-source.json
Full Text

Full Text

BREEDING BIRD COMMUNITIES and HABITAT ASSOCIATIONSin the GRASSLANDS of the CHILCOTIN REGION, BRITISH COLUMBIAbyTRACEY DENINE HOOPERB.Sc., The University of Victoria, 1985A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Plant Science)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAOctober 1994© Tracey Define Hooper, 1994Signature(s) removed to protect privacyIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)__________________________Department of ‘v-The University of British ColumbiaVancouver, CanadaDate 1E3, ;11L]DE-6 (2/88)Signature(s) removed to protect privacyabstractThe objectives of this study were (1) to characterizebreeding bird communities in the grasslands around Riske Creek,in the Chilcotin Region of British Columbia, (2) to identifyassociations between grassland habitat characteristics andbreeding bird diversity and species abundances, (3) to determinethe relationship between food resource availability and grasslandbird diversity and species abundances, and (4) to elucidate thepotential impacts of livestock grazing on breeding birds andtheir grassland habitats.Point—counts and spot—mapping methods were used to censusbirds. Habitat characteristics measured were vegetation height,vertical cover, and patchiness, horizontal cover of differentphysiognomic features, arthropod abundances, and site slope,aspect, elevation. The season, intensity, and grazing bylivestock and California bighorn sheep within th study area werealso determined. Principal component, multiple correlation, andcluster analyses were used to analyze bird/habitat associations.Eleven species of birds were common throughout the studyarea. They were, in order of decreasing abundance: horned lark(Eremophila alestris), vesper sparrow (Pooecetes gramineus),European starling (Sturnus vulgaris), western meadowlark(Sturnella neglecta), long—billed curlew (Numenius americanus),Brewer’s blackbird (Euihagus cyanocephalus), tree swallow(Tachycineta bicolor), mountain bluebird (Sialia currucoides),American crow (Corvus brachyrhynchos), savannah sparrowi_i(Passerculus sandwichensis), and red crossbill (Loxiacurvirostra). Three red—listed (i.e., threatened or endangered)species (upland sandpiper (Bartramia lonciicauda), Sprague’s pipit(Anthus spragueii), Brewer’s sparrow (Spizella breweri)) and twoblue-listed (vulnerable) species (sharp-tailed grouse(Tympanuchus phasianellus) and short—eared owl (Asio flammeus)),were also recorded. Over 13 study sites, bird density rangedfrom 0.82 to 1.24 pairs/ha and Shannon’s species diversityindices ranged from 0.71 to 1.38.Vegetation structure and patchiness, horizontal cover,topographic features, and grazing characteristics combined,explained 15-96% of the variability in bird diversity and speciesabundances. Topographic features were most often associated withbird diversity and species abundances, possibly due tocorrelations between topography and vegetation structure.Vegetation structural components were the second most commonfeatures associated with birds.Although arthropod abundances explained 15—72% ofvariability in bird diversity and species abundances,associations between birds and food resources wereoften unclear. Associations between birds and grazing were alsounclear.111Table of Contents1. Introduction . . . . .1.1 Literature Review . . . . . . .1.1.1 Grassland Bird Communitiesand Habitats . . . . . .1.1.2 Influences of Livestock Grazingon Grassland Bird Communitiesand Habitats1.2 Objectives . .1.3 Research Hypotheses . . .2 . Study Area . . . . 1014141414171717171819192021222324272729323234393939394040A1 stract . . . . . . . .Tableof Contents . .ListofTables . . . .Listof Figures. . . . . . .Acknowledgements . . . . . .• . . . . :1_i• . . . . iv• . • . . vii• . . . . ix• . . . . . x• . . . . . 1• 335• 883. Methods . . •3.lStudySites. . . . . . . . . .3.2 Grassland Breeding Bird Censuses3.2.1 Species Censuses .3.2.2 Long—billed Curlew Censuses3.3 Habitat Characteristics .3.3.1 Vegetation Structure . • .Height and Vertical Cover .Horizontal Cover . . . .Vegetation Patchiness . •3.3.2 Vegetation Species Composition3.3.3 Food Resources3.3.4 Grazing Information . . . .3.3.5 Topographic Features . . . .3.4 Habitat Structure Use by Birds .3.5 Statistical Analyses . . . . . .4 Results .4.1 Bird Censuses . . . . .4.2 Long-billed Curlew Breeding Densities4.3 Bird/habitat Associations4.3.1 Principal Component Analysis . .4.3.2 General Bird/habitat Correlations4.3.3 Specific Bird/habitat CorrelationsSpecies Diversity . . . . . • •HornedLark. • • . •Vesper Sparrow . . . .Long—billed CurlewWestern Meadowlark . . . . . .iv5 Discussion . .5.1 Bird Censuses . . . . . . . .5.2 Long—billed Curlew Censuses .5.3 General Bird/habitat Associations5.4 Specific Bird/habitat Associations5.4.1 Species diversity .5.4.2 Horned Lark5.4.3 Vesper Sparrow . .5.4.4 Long-billed Curlew5.4.5 Western Meadowlark5.4.6 Brewer’s Blackbird5.4.7 Tree Swallow . . . .5.4.8 Mountain Bluebird . .5.4.9 Savannah Sparrow . .5.4.10 Upland Sandpiper . .5.4.11 Sprague’s Pipit . .5.4.12 Brewer’s Sparrow . .5.4.13’Sharp—tailed Grouse5.4.14 Short—eared Owl . .5.5 Habitat Structure Use by Birds6 Conclusions . . . . . . . . . . .7 Management Recommendations . . .• • • . . 58• . • . • • 586061• . . . • . 6666676971727474. 7575. 76• • . . • . 78• 78• • • • • 79• 80• . . . • . 8081• • 838 Literature Cited . . . . . . • •9Appendices • • . . • . • . • . .10 Appendix 1: Bird species censused in theChilcotin study area, British Colubia, 1991/9211 Appendix 2: Codes for habitat variables . •v879999102Brewer’s Blackbird • . • 41Tree Swallow • • . • . . . • • 41Mountain Bluebird • • • • 41Savannah Sparrow 42Upland Sandpiper • . • 42Sprague’sPipit 42Brewer’s Sparrow . . • • • • • • • 43Sharp—tailed Grouse . • • • • . • • 43Short—earedOwl. . . . • • . • 434.3.4 Species Management Guidelines • • • • . 43Species Diversity • • • 43HornedLark. • • • • • • . • • • 44VesperSparrow • • • • • 44Long—billedCurlew • • . • 50Western Meadowlark • • • • . . • . 52UplandSandpiper • • • • • • • • • • 54Sprague’sPipit. . . . • • 544.3.5 Habitat Management Guidelines . • • • 544.3.6 Habitat Structure Use by Birds . • • 5512 Appendix 3: Plant species recorded in theChilcotin study area, British Columbia, 1992 . . . 10313 Appendix 4: Chi-square analysis of habitatstructural use among horned larks,vesper sparrows, western meadowlarks, andmountain bluebirds in the Chilcotin study area,British Columbia, 1991/92 . . . . 10514 Appendix 5: Chi-square analysis of habitatstructural use by horned larks, vesper sparrows,western meadowlarks, and mountain bluebirdsin the Chilcotin study area, British Columbia,1991/92 . . . . . . . . . . . . 10615 Appendix 6: Multiple correlations between vegetationstructure and topographic features for theChilcotin study area, British Columbia,1991/92 . 10716 Appendix 7: Multiple correlations between vegetationstructure and seasonal grazing (AUMs) for theChilcotin study area, British Columbia,1991/92 . . 108viList of Tables1. Most common, and red- and blue-listed birdspecies censused in the Chilcotin study area,British Columbia, 1991/92 . . 282. Bird densities and Shannon’s diversityindices by study site, for the Chilcotinstudy area, British Columbia, 1991/92 . . . 293. Long-billed curlew breeding densities withinthe Chilcotin study area, British Columbia,1990, 1991, and 1992 . . . . . . . . . . . . 304. Long-billed curlew breeding densities withinthe Chilcotin study area, British Columbia,1987, 1990, 1991, and 1992 . . . . . . . . 315. Principal components and component loadingsfor habitat variables in the Chilcotinstudy area, British Columbia, 1991/92 . . . 336. Mean numbers of bird/habitat correlationsby general habitat characteristic for theChilcotin study area, British Columbia,1991/92 . . . . . . . . . 347. Pearson correlation coefficients forbird/habitat correlations for the Chilcotinstudy area, British Columbia, 1991/92 . . . 368. Habitat variables associated with low,medium, and high diversity of bird speciesin the Chilcotin study area,British Columbia, 1991/92 . . . . . . . . . . . 469. Habitat variables associated with lowand high abundances of horned larks inthe Chilcotin study area, British Columbia,1991/92 . . . . . . 4810. Habitat variables associated with lowand high abundances of vesper sparrowsin the Chilcotin study area,British Columbia, 1991/92 . . . 5011. Habitat variables associated with low,medium, and high breeding densities oflong-billed curlews in the Chilcotinstudy area, British Columbia, 1991/92 . . . . . 52vii12. Habitat variables associated withlow, medium, and high abundances of westernmeadowlarks in the Chilcotin study area,British Columbia, 1991/92 . . . . . 5413. Mean habitat variables associated withhighest bird species diversity and speciesabundances in the Chilcotin study area,British Columbia, 1991/92 . . . . . . . . . . . 56viiiList of Figures1. Location of the study area in British Columbia . 112. Location of the grassland study sites andarthropod collection sites within theChilcotin study area, British Columbia,1991/92 . . . . . . . . . . . . . . . 153. Cluster analysis of Shannon’s diversityindices by study site for the Chilcotinstudy area, British Columbia, 1991/92 . . . 454. Cluster analysis of horned lark abundancesby study site for the Chilcotin study area,British Columbia, 1991/92 . . . . . . . . . 475. Cluster analysis of vesper sparrowabundances by study site for the Chilcotinstudy area, British Columbia, 1991/92 . . . 496. Cluster analysis of long-billed curlewbreeding densities by study site for theChilcotin study area, British Columbia,1991/92 • • 517. Cluster analysis of western meadowlarkabundances by study site for theChilcotin study area, British Columbia,1991/92 . 538. Habitat structural availability and useby horned larks, vesper sparrows,western iueadowlarks, and mountain bluebirdsin the Chilcotin study area, British Columbia,1991/92 . . . . . . . . . . . 57ixAcknowledgementsI thank my research supervisor, Dr. Michael Pitt forenthusiastically supporting this project when others did not.am also grateful to Dr. Pitt for addressing my doubts andanxieties with sympathy and humour, and for teaching me the valueof patience and diplomacy. Above all though, I thank Mike foroffering the rare gift of friendship between supervisor andstudent.Special thanks also go to my committee members. Dr. Jean—Pierre Savard provided the initial idea and support for thisproject. Dr. Val Lemay gave generously of her time andexpertise, and showed infinite patience during the data analysis.Dr. Judy Myers and Dr. Fred Bunnell provided insight anddirection throughout the project.I would also like to thank Harold Armieder, Herb Langin, BrianNyberg, Jim Young, Michaela Waterhouse, Fred Knezevich, RossFredell, and Derek White for their advice, and generous technicaland logistic support throughout the project, and Dr. PeterMarshall for his help with the data analysis. Special thanksalso go to Anna Roberts, Gina Roberts, Wayne Campbell, Ed Houeck,Peter Dryce, and the Williams Lake Naturalists fbr their wealthof knowledge about the natural history of the study area, andtheir willingness to share it.Steve McConnell, Ruth Van den Driessche, Colleen Bryden, andAlice Cassidy were professional and dedicated field assistants.Barry Forer offered invaluable advice and moral supportthroughout the project.And finally, my sincere thanks go to Chief Francis Laceese ofthe Toosey Indian Band, and local ranchers — Lynn Bonner, GrantHuffman, Neil and Kerry McDonald, Ron and Stephanie Thomson, andArt Graves for permission to do my research on their lands.This research was funded jointly by the Ministry of ForestsOld Growth/Biodiversity Fund, the Forest Resource DevelopmentAgreement (FRDA), the Ministry of Environment, Lands, and ParksHabitat Conservation Fund, and the Canadian Wildlife Service.Forestry Canada provided accommodation for the field crew.Personal funding was provided by the Natural Sciences andEngineering Research Council of Canada (NSERC).x1. IntroductionGraul (1980) suggested that the grasslands of North Americahave been more severely altered by man than any other majorecosystem on the continent. In the American Great Plains and theCanadian prairies, 45-85% and 76-99%, respectively, of the nativegrassland vegetation communities have been converted to intensiveagricultural practises (Kiopatek 1979; World Wildlife FundCanada 1989). In comparison, for western forest types, only 1 to5% of the native vegetation communities have been replaced(Kiopatek 1979). In British Columbia (B.C.), probablyless than 1% of grasslands have been officially protected (T.Void, pers. comm.). Only 9% of the historical grasslands in theOkanagan Region remain in a relatively natural state (Redpath1990), while less than 2% of the Chilcotin-Cariboo nativegrasslands have been preserved (J. Youds, pers. comm.). Thefailure to preserve representative North American grasslands mayhave resulted from an ignorance of their biological significanceof these areas (Hooper and Pitt 1994). For example, in Canada,more than 1/3 of the birds and mammals on the 1988 COSEWIC(Committee on the Status of Endangered Wildlife in Canada) listare from the prairie provinces, and most of these are associatedwith prairie grasslands and parklands (World Wildlife Fund Canada1989). Approximately 27% of the species on B.C. Environment’sRed (threatened and endangered) and Blue (vulnerable) Lists(1993), are associated with the Chilcotin-Cariboo grasslands(Hooper and Pitt 1994).1British Columbia’s grasslands are unique in Canada becausethey are dominated primarily by the bunchgrass, bluebunchwheatgrass (Agrovron sDicatum), which only rarely occurs east ofthe Canadian Rocky Mountains (Hooper and Pitt 1994). Moreover,B.C. ‘s grasslands represent the northern limit of contiguousbunchgrass vegetation in North America, and are distinguishedfrom their ecological counterparts in Oregon and Washington by agreater proportion of boreal rather than austral plant species(Daubenmire 1978, Hooper and Pitt 1994). B.C.’s grasslands arealso rather exceptional because a smaller proportion than manyother grasslands in Canada has been converted to crop cultivation(Hooper and Pitt 1994).The B.C. Ministry of Environment, Lands, and Parks (B.C.Environment) (1991) considers bunchgrass and big sagebrush(Artemisia tridentata) habitats in the province to be majorwildlife habitats of concern because:a) their area and distribution is limited;b) they are being lost to urban and agricultural development; andC) they are being altered by livestock grazing and forestencroachment.Despite these concerns, and the fact that B.C. Environment andthe B.C. Ministry of Forests have a common goal of preservingbiodiversity, there are few detailed studies of grassland faunafor British Columbia.Birds are a conspicuous and important component of grasslandecosystems (Wiens 1973b), and some species are often restricted2to grassland habitats (Cody 1985). Disturbances to grasslands,then, could affect grassland bird populations and communities.In their problem analysis on biodiversity of the Chilcotin—Cariboo grasslands, Hooper and Pitt (1994) found that livestockgrazing was considered by grassland resource—users to be theimpact of greatest concern to that region’s grasslands.Detailed studies of grassland bird communities or theirhabitat associations for B.C., and data that document thepotential impacts of livestock grazing on British Columbia’sgrassland bird populations are lacking. In the interest ofgrassland biodiversity conservation, this research was designedto address these information needs.1.1 Literature Review1.1.1 Grassland Bird Communities and HabitatsBird diversity and abundance are often positively correlatedwith vegetation complexity (Roth 1977, Cody 1985). Becausegrassland vegetation structure is fairly homogeneous, grasslandbird communities are relatively simple. In general, grasslandsare characterized by two to six passerine species, andoccasionally, as many nonpasserine species (Cody 1985). Incomparison, the mean number of bird species in coniferous forestscan be two to four times greater than that of grasslands (Wiens1975).Vegetation structure is considered to be the most important3factor affecting grassland bird distribution (Tester and Marshall1961, Hilden 1965, Wiens 1969, Whittaker and Woodwell 1972, Cody1985). The most important components of vegetation structure forgrassland birds are grass height and/or density (Cody 1966,Creighton 1974, Ohanjanian 1985), litter and vegetationpatchiness (Wiens 1969, 1973b, 1974a, 1974b, 1976, Wiens and Dyer1975), and amount of ground and shrub cover (Bock 1984).Wiens and Rotenberry (1981) suggested, however, that birds ofshrub—steppe environments also respond to vegetation floristics.A more complete understanding of grassland avian habitatassociations may, therefore, be obtained by studying bothvegetation physiognomy and floristics (Wiens and Rotenberry1981).Bird diversity can also be correlated with food availability(Wiens 1974a). Grassland birds are omnivorous, but during thebreeding season, arthropods form the bulk of the diet (Wiens1973b, Terres 1980, Cody 1985). Grasshoppers, ants, beetles,bugs, butterfly and moth larvae, and spiders are the most commonprey items (Wiens 1973b, Rotenberry and Wiens 1978, Cody 1985,Redmond and Jenni 1985), with grasshoppers probably being themost important arthropod in grassland bird diets (Baldwin 1971,1973, Maher 1979).. Because food can, at times, be a limitingresource, food supplies may exert strong influences on birdcommunities (Rotenberry 1980). Limitations in arthropodabundances then, could affect the composition of grasslandbreeding bird communities.41.1.2 Influences of Livestock Grazing on Grassland BirdConimunities and HabitatsMost B.C. grasslands are managed, either privately or by theMinistry of Forests, for sustained forage and livestockproduction (Hooper and Pitt 1994). Livestock grazing can changegrassland vegetation structure by altering vegetation density,plant vigour and growth, and plant community species composition(Ryder 1980). Not surprisingly, these changes often correspondto increases or decreases of many grassland bird species(Townsend and Smith 1977).Possibly the most important component of grassland birdhabitat affected by grazing is availability of cover for nesting,rearing, hiding, and thermoregulation. Birds that prefer to nestin dense or tall vegetation, and/or require elevated singingperches may benefit from levels of grazing and browsing thatstimulate bushier shrub growth, or that leave large amounts ofstanding vegetation (Ryder 1980). For example, sharp—tailedgrouse are more likely to use dancing grounds and adjacent nestsites in areas where grazing levels maintain, rather than removenesting and brooding cover (Pepper 1972). Conversely, grazinglevels that reduce vegetation height and density could favourspecies such as the long—billed curlew which may require goodvisibility for communal detection of predators, and forunhampered movements of their precocial chicks when feeding(Allen 1980). Adult horned larks also forage less efficiently intall, dense vegetation which can result in chicks starving to5death in the nest (Cody 1985).Changes in vegetation structure due to grazing can alsoaffect the availability of cover for hiding or thermoreguiation.For example, long-billed curlews may require patchy plantdistribution for camouflage since they nest in open areas withoutprotection of overhanging vegetation (Ohanjanian 1985). Otherspecies, however, need sufficient vegetative cover during warmweather to prevent water loss in eggs and heat stress to chicks.Even short—grass nesters like horned larks seem to require sometaller grasses around their nests for thermal protection(Cannings 1981).Aside from grazing action, the mere presence of grazers canaffect habitat structure. For example, trampling can changevegetation structure by killing or compacting plants and reducingcompetitive cover, but it can also facilitate seed dispersal(Springfield 1976, Little 1977). Trampling can also cause soildisturbances or compaction which can be beneficial or detrimentalto ground-nesting birds which build their nests in soft soil(Harrison 1984). Additionally, manure pies left by livestock canincrease structural variability of the habitat by providing windbreaks or camouflage for ground—nesting birds (Cameron 1907).In addition to influencing habitat structure, livestockgrazing may affect grassland birds by altering food resources.Grassland arthropods can be affected by changes in microclimate,living space, and food availability as a result of grazing(Breymeyer and van Dyne 1980). Arthropod numbers may increase,6but species numbers generally decline with heavy grazing (Smith1940). Like birds, different arthropod types respond differentlyto grazing pressures. For example, grasshoppers prefer opengrasslands with sparse vegetative cover (Hewitt . 1974), andthus, apparently benefit from grazing (Shotwell 1958, Skinner1975), as they are more common in heavily grazed than lightlygrazed areas (Smith 1940, Kelly and Middlekauff 1961). Incontrast, higher densities of harvester ants (Pogonomvrmex spp.)occur in moderately rather than heavily grazed areas (Rogers1972).These effects of grazing on arthropod distribution andabundances could be beneficial or detrimental to grassland birdsdepending on their dietary needs and preferences. Additionalbenefits to grassland bird food resources may be derived fromlivestock grazing. Manure piles left by grazers can provide foodin the form of insects and partially digested plant matter.Long-billed curlews will probe, and horned larks will tear apart“cow pies,” presumably to retrieve seeds and insect parts, orwhole insects feeding in the “pies” (Ryder 1980).Although the species of grassland birds have been recordedfor B.C., few quantitative data are available on breeding birddiversity and density, species abundances, and habitatassociations. Few studies in the current literature haveexamined the relationships among grassland birds, habitatstructure, food availability, and livestock grazing. Thisresearch was designed to examine those relationships.71.2 ObjectivesThe objectives of this research were:1. to characterize grassland breeding bird communities of thestudy area in terms of bird density, species diversity, andspecies abundances;2. to identify associations between grassland habitatcharacteristics and breeding bird diversity and speciesabundances;3. to determine the relationship between food resourceavailability and breeding bird diversity anI speciesabundances; and,4. to elucidate potential influences of livestock grazing ongrassland breeding bird communities and their habitats.Only breeding birds were considered since British Columbia’sgrasslands, in general, offer too little food and shelter tosupport a winter bird community (Cannings j,. 1987).1.3 Research HypothesesThe research hypotheses and predictions tested in this studywere:1. Vegetation structure is the primary determinant of grasslandbreeding bird diversity and species abundances.a) Bird diversity will be greater in more structurallycomplex habitats, and will therefore, increase withvegetation height, vertical cover, and vertical and8horizontal patchiness.b) Individual bird species will be associated withparticular vegetation structural types. Abundances ofsome species will be most strongly associated withstructurally complex habitats while other species willbe associated with less complex habitats (e.g., withreduced vegetation height, vertical cover, andpatchiness).2. Grassland breeding bird diversity and species abundances arealso associated with food availability.a) Bird diversity will be greater in areas with greaterarthropod biomasses.b) Individual bird species will be associated withparticular food resources. Abundances of individualspecies will be positively correlated with biomasses ofparticular arthropod types.3. Grassland breeding bird diversity and species abundances areaffected by livestock grazing.a) The greatest influence of livestock grazing on grasslandbirds will be due to alterations in habitat structure.Habitat complexity will decrease with increased grazingpressure due to reduced vegetation height, verticalcover, and vertical and horizontal patchiness.i) Bird diversity will decline with increased grazingpressure due to reduced habitat complexity.ii) Individual bird species will be affected differently9by livestock grazing. Species associated withstructurally simple habitats will increase inabundance with increased grazing pressure due toreduced vegetation height, vertical cover, andvertical and horizontal patchiness. Speciesassociated with structurally complex habitats willdecrease in abundance with increased grazingpressure.b) Livestock grazing will also affect individual species ofbreeding grassland birds due to alterations in foodresources. Individual arthropod types will be affectedby livestock grazing due to alterations in habitatstructure.i) Arthropod types associated with structurally simplehabitats will increase in abundance with grazingpressure due to reduced vegetation height, verticalcover, and vertical and horizontal p&tchiness.Increases in abundances of certain arthropod typeswill be associated with increases in certain birdspecies.2. Study AreaThe study area was the grasslands around Riske Creek (51°52’ N, 122° 21’ W) in the Chilcotin Region of British Columbia(Fig. 1). These grasslands are within the Fraser River BasinEcosection, and the Bunchgrass and Interior Douglas—fir10Fig. 1. Location of the study area in British Columbia.11biogeoclimatic zones (Demarchi 1988).The study area included grassland and shrub—steppe sitesranging from 585 to 1000 in elevation within the Bunchgrass (BG)and Interior Douglas-fir (IDF) biogeoclimatic zones. Bothbiogeoclimatic zones are typified by warm to hot, dry summers,and cool to cold winters with relatively little snowfall(Nicholson j. 1991, Hope . 1991). Precipitation tendsto be bimodal with the main peak occurring in December andJanuary, and a secondary peak occurring in June (Nicholson g1991, Hope , . 1991). Throughout the study area, theBunchgrass biogeoclimatic zone often intergrades into the IDFbiogeoclimatic zone.Grasslands of the BG zone occur on Brown, Dark Brown, Black,and Dark Gray Chernozemic soils, and are characterized by widely—spaced bunchgrasses and a well—developed cryptogam layer(Nicholson . 1991). Two subzones - the Very Dry Hot (BGxh)and the Very Dry Warm (BGxw) occur from valley bottoms to about700 iu elevation, and from about 700 to 1000 in elevation,respectively (Nicholson 1991).Dominant vegetation on mesic and submesic sites within theBGxh includes bluebunch wheatgrass, big sagebrush, and Sandberg’sbluegrass (g sandbergii), while needle-and-thread grass (Stipacomata) and sand dropseed (Siorobolus crvtandrus) dominate xericsites with coarse-textured soils. Within the BGw, dominantvegetation includes bluebunch wheatgrass, pasture sage (Artemisiafrigida), and junegrass (Koeleria macrantha). Rabbit-brush12(Chrysothamnus nauseosus) is also often present, and porcupinegrass (Stipa curtiseta) and Kentucky bluegrass ( pratensis)occur on moist sites and on the bases of some slopes. Mesic tosubmesic sites at higher elevations of the BGxw are dominated bya porcupine grass/Rocky Mountain fescue (Festuca. saximontana)association. Plant communities found on wetter sites above 700 inwithin the BG zone include small groves of trembling aspen(PoDulus tremuloides) with understories of western snowberry(Svnrnhoricarpos occidentalis), Kentucky bluegrass, northernbedstraw (Galium boreale), American vetch (Vicia americana),quackgrass (AgroDyron repens), and star—flowered false Solomon’s—seal (Smilacina stellata) (Roberts 1992).Grasslands of the IDF zone occur on Orthic Dark Brown,Orthic Black, and Dark Grey Chernozems. Dominant plantcommunities include bluebunch wheatgrass, porcupine grass,spreading needlegrass (Stipa richardsonii), and Rocky Mountainfescue, with pasture sage, junegrass, Kentucky bluegrass, andwoolly cinquefoil (Potentilla hipiana) also being common.Timber milk-vetch (Astragalus miser), yarrow (Achilleamillefolium), cheatgrass (Bromus tectorum), cut—leaved daisy(Ericeron coinpositus), yellow-rattle (Rhinanthus crista-aalli),and goats—beard (Tragopocron spp.) can also become common withincreased livestock grazing pressure (Hope j,. 1991).133 Methods3.1. Study SitesThirteen grassland study sites were established within anarea bounded by the Junction Wildlife Management Area (WMA),Becher’ s Prairie, and the Fraser River (Fig. 2). Study siteswere selected to represent regional variability in topographicand vegetation characteristics. Sites were chosen then, based ondifferences in elevation, slope, aspect, and early spring (April)vegetation height and vertical cover. Differences in vegetationheight and cover were used as indicators of levels of grazingpressure. Sites were also selected for their accessibility byvehicle, and for their size (i.e., sites had to be large enoughto accommodate at least four bird census points).3.2 Grassland Breeding Bird Censuses3.2.2. Species CensusesBirds were censused using the point—count method (Verner1985). Points had a radius of 100 m and were placed at least 300m apart, as measured from the centre of each point. Althoughsome edge habitat was included in some points, areas with largeor dense stands of trees, or large wetlands were avoided. Fourto 15 points (determined by grassland area) were established ineach of the 13 grassland sites (145 points total) (Hooper andSavard 1991).All points were established as semi—circles along roadwaysthrough the sites. The observer drove to the centre of14Bedie?s PraldeItE17———-——.——-..• u ‘8..E.‘Hvy2OE’. - I\ .BW tQ/%. ii%____M_NE\%, I— — — —— ç•M_NWM_SE I••T_N MdonaIds RandiToosey ênckn Reseve )•T_S.0 /Deer Paik Ranch———pavedmads_s . . -. . - gravel .oed••S..SW dkt roadJunclon WMANI —I0 2.5Fig. 2. Location of the grassland study sites (.) and arthropodcollection sites (U) within the Chilcotin study area,British Columbia, 1991/92.15a point, turned off the vehicle, got out and quietly closed thedoor, then waited two minutes before beginning the count. Thismethod was used to avoid disturbing birds by walking through thepoint—count area (Hooper and Savard 1991). Counts were made forthree consecutive four—minute periods at each point (12 minutestotal per point). All birds detected within the point-countcircle, and up to 50 m beyond the circle were recorded. Eachpoint was surveyed between 05:30 a.m. and 10:00 a.m., five or sixtimes between early May and early July (the height of the birdbreeding season), in each of 1991 and 1992.Bird species diversity per grassland site was calculatedusing Shannon’s diversity index: H’= -E p1 log p1 (where p. is theproportion of all observations of the ith species and i=1,2,...S(where S is the total number of species)) (Magurran 1988). Birddensity per site was calculated as the total number of birdsdivided by the total area censused. The area of the 100-rn radiuspoint count semi—circle, plus the 50—rn radius border beyond thecircle was 3.92 ha (Hooper and Savard 1991). To compare densityestimates with other grassland bird studies, the number of birdsper hectare was divided by 1.5 to give an estimate of number ofpairs of birds per hectare.Bird common and taxonomic names were taken from the AmericanOrnithological Union Checklist pj North American Birds (1983) andrecent supplements (1984; 1985; 1987; 1989).163.2.2 Long—billed Curlew CensusesBecause 100—rn radius point—counts may not adequately censuslong—billed curlews (Hooper and Savard 1991), a modified spot-mapping technique was also used. Observers drove along roadwayswithin the 13 study sites and recorded the number and sex oflong—billed curlews observed. Each site was surveyed 3—5 timesduring April and May in each of 1991 and 1992, when curlews wereestablishing and defending territories. The repeated surveysdocumented individual breeding territories. Curlew breedingdensities were determined as the number of territories divided bytotal grassland area for that site.3.3 Habitat Characteristics3.3.1 Vegetation StructureHeight and Vertical CoverHeight of herbaceous vegetation was measured as the height(cm) at which the vegetation was most dense, based on anobserver’s eye height of 30 cm, at a distance of 10 m from themeter stick (Hooper and Savard 1991). Vertical cover (%) ofherbaceous vegetation was measured using a 30— x 50—cm visionboard, with an observer eye height of 30 cm, at a distance of 10m from the board (Bicak j,. 1982, Hooper and Savard 1991).Five samples of height and vertical cover were taken withinthe circumference of each point—count circle, in early May andagain in late June, in each of 1991 and 1992 (n = 40—150/site/year depending on site area). A random numbers table was17used to locate individual sample points at a specific distanceand direction from the centre of the point—count circle.Vegetation height and vertical cover were measured in both Mayand June to determine if species site associations were based onhabitat requirements for breeding (i.e., in May) or for rearing(i.e., in June).Horizontal CoverThe frequency and horizontal cover (%) of nine physiognomicfeatures were measured using a 20— x 50—cm Daubenmire frame.Physiognomic features assessed were grasses, forbs, shrubs,trees, cryptogams, rocks, bare soil, litter, and animal feces.Percent cover was recorded according to cover class: class 0: 0%cover, 1: 1—5%, 2: 5—25%, 3: 25—50%, 4: 50—75%, 5: 75—95%, and 6:95—100%. Mid—points of each cover class were used as covervalues in the data analyses. An area considered representativeof the site’s dominant plant community was used for sampling. Ifmore than one type of plant community dominated a site, samplingwas done in each community. In each plant community, three 50—mtransects were laid parallel with the land contours, with 30 mbetween transects. Ten samples were taken along each transect.All transects were at least 50 m from any roads to avoiddisturbance effects. A random numbers table was used to locatesampling points along the transects. Measurements were made inmid-June, in each of 1991 and 1992. These sampling methods wereconsistent with those used by the B.C. Ministry of Forests (H.18Armieder, F. Knezevich, pers. comm.).Vegetation PatchinessVegetation patchiness was assessed using heterogeneityindices for vegetation height and vertical cover, and horizontalcover of bare ground. Indices were calculated by dividing therange of measurements of the attribute, by the mean, for a givenpoint (Rotenberry and Wiens 1980). These indices were chosen soas to be consistent with the those used by Rotenberry and Wiens(1980).3.3.2 Vegetation Species CompositionThe frequency and horizontal cover (%) of plant species wasmeasured using a 20— x 50—cm Daubenmire frame. Percent cover wasrecorded according to cover class: class 0: 0% cover, 1: 1—5%, 2:5—25%, 3: 25—50%, 4: 50—75%, 5: 75—95%, and 6: 95—100%. Midpoints of each cover class were used as cover values in the dataanalyses. An area considered representative of the site’sdominant plant community was used for sampling. If more than onetype of plant community dominated a site, sampling was done ineach community. In each plant community, three 50—rn transectswere laid parallel to the land contours, with 30 m betweentransects. Ten samples were taken along each transect. Alltransects were at least 50 m from any roads to avoid disturbanceeffects. A random numbers table was used to locate samplingpoints along the transects. Measurements were rn.ade in mid—June,19in each of 1991 and 1992. These sampling methods were consistentwith those used by the B.C. Ministry of Forests (H. Armleder, F.Knezevich, pers. comm.).Plant common and taxonomic names were taken from Taylor andMacBryde (1977), Meidinger (1987), and Douglas (1989,1990, 1991).3.3.3 Pood ResourcesFood resources were measured in terms of arthropodabundances. Pan traps were used to collect both low—flying andground-crawling arthropods (Martin 1977, Hooper and Savard 1991).Five traps were set in each of six grassland study sites (30traps total) (Fig. 2). Three sites were ungrazed or lightlygrazed, while the other three sites were more heavily grazed.Each trap was set at a point—count location, and traps werelocated equi-distant throughout the site. Traps were set frommid-April to late June, in each of 1991 and 1992. In the dataanalysis, 36 rather than 30 sites had arthropod data, because onesite and one sampling point had to be changed between years dueto livestock disturbances.Traps were made from 23 x 23 x 4—cm cake pans buried to therim and filled about 2/3 full with water and dishwashingdetergent. A 30.5 x 30.5-cm board was placed at approximately450 over the pan. Arthropods were removed from the traps byfiltering the trap contents through a small aquarium net. Liquidcontents were returned to the traps and replenished, if20necessary. Arthropods were collected weekly and stored in 10%isopropyl alcohol until identified. Samples were discarded ifthe traps had been disturbed by livestock.Arthropods were counted and identified by broad groupings:ants: large (6 mm) or small (5 mm), bees: large (lO mm) orsmall (9 mm), beetles: large (16 mm), medium (10—15 ram), orsmall (9 mm), bugs: large (6 mm) or small (5 mm), flies: large(6 mm) or small (5 mm), grasshoppers: large (15 mm), medium(10—15 mm), or small (9 mm), larvae: large (15 mm) or small(14 mm), butterflies/moths: large (ll mm) or small (l0 mm),spiders: large (l0 mm), medium (6—10 mm), or small (5 mm), andothers. These groupings were chosen because ants, beetles, bugs,butterfly and moth larvae, grasshoppers, and spiders are the mostcommon prey items of breeding grassland birds (Rotenberry andWiens 1978, Cody 1985, Redmond and Jenni 1985). Flies and beeswere included in this analysis because they were commonlycollected.Arthropods were oven—dried to constant weight at 60° C, andweighed on a Mettler AC 100 electronic balance to 0.0001 gin(Southwood 1978). Biomass estimates were then made for ants,bees, beetles, bugs, flies, grasshoppers, larvae, moths andbutterflies, spiders, and total arthropods.3.3.4 Grazing InformationTo assess the potential impact of grazing on grassland birdcommunities on the 13 study sites, information on the season,21intensity, and duration of grazing by livestock (cattle andhorses) and California bighorn sheep (Ovis canadensiscaliforniana) was gathered from ranchers, Ministry of Forestsagrologists, and B.C. Environment wildlife biologists. Number ofAnimal Unit Months (AUM5) was then calculated by season for eachsite. The definition of AUM used in this study was the amount offorage required to feed a mature cow, with or without sucklingcalf, for one month. AUMS were calculated based on the metabolicweight of a 454 kg cow. Grazing by one horse was equivalent to1.3 AUMs. Using a weight of 68 kg for a bighorn sheep ewe (Burtand Grossenheider 1976), grazing by one bighorn sheep wasequivalent to 0.25 AUMS. Animal Unit Months were used toquantify grazing levels so as to be consistent with Ministry ofForests grazing assessment methods.3.3.5 topographic FeaturesSlope, aspect, elevation, and soil bulk density wererecorded as additional site descriptors for each of the 13 sites.Slope was measured using an Abney level; aspect was determinedusing a hand—held compass. Elevation was taken from Departmentof Energy, Mines, and Resources 1:50,000 topographic maps with 50m contour intervals. Soil bulk density was measured to assessdegree of soil compaction, which may be important to ground—nesting birds that build nest scrapes. Soil cores were removedwith a 2” Dutch auger. The hole’s volume was measured by liningthe hole with a plastic bag and filling it with a measured amount22of water. Soil cores were dried to constant weight at 1000 C andweighed on a Mettler PC 4400 electronic balance to 0.01 gm. Soilbulk density was estimated as the ratio of soil weight to watervolume. One soil sample was collected at each point—countlocation (n=145 samples). A random nuLlbers table was used tolocate the distance and direction of sampling points from thecentre of the point—count circle.3.4 Habitat Structure Use by BirdsTo clarify the importance of habitat structural features tograssland birds, the use of various structures by mountainbluebirds, western meadowlarks, vesper sparrows, and horned larksfor displaying, resting, and feeding was recorded during thepoint—count censuses. These species were selected because theywere among the most common grassland bird species (Hooper andSavard 1991), and because they were more commonly observed usingstructural features than other species. The measurements ofphysiognomic features covered only a small portion of each studysite, and thus, may have underestimated the numbers of somestructural features important to grassland birds; consequently,additional counts were made of rocks, shrubs, trees, fences,logs, and roads (i.e •, narrow gravel roads and dirt tracks), bothon, and 50 m beyond each point—count circle.233.5 Statistical AnalysesAnalyses were initially done separately for each of thefield seasons. Because results were similar between years,however, the data were combined and re—analyzed to give trends inbird habitat associations across the two field seasons. Datafrom the 1991 and 1992 field seasons then, were combined, andmeans of the bird and habitat variables were calculated for eachof the 145 points used to census birds. The bird diversityindices, and measures of the physiognomic classes which were madeper site rather than per point, were simply repeated for each ofthe points within the site. Missing values were given for sitesand points without arthropod variables.Data reduction of the habitat variables was done usingprincipal components analysis (PCA) on SYSTAT (Version 5.02,1991). Because the data set was too large to analyze with thepersonal computer version of SYSTAT, the habitat variables weresplit into two files. One file included the vegetationstructure, topographic, and grazing data, while the second fileincluded the vegetation species composition data. The arthropodvariables were not included because data were available for only36 of the 145 points. Because there were only 10 arthropodvariables, data reduction by PCA was unnecessary.Principal components were selected based on eigenvalues ofthe correlation matrix that were greater than one. Componentloadings ‘t0.6O0I were chosen to select variables for furtheranalysis.24Following PeA, multiple correlations identified associationsbetween the habitat variables and Shannon’s diversity indices,the most common bird species (i.e., those specie comprising >1%of all observations), and those species on B.C. Environment’sRed— and Blue—lists (1993). Red crossbills, European starlings,and American crows were among the most common species recorded,but were excluded from the correlation analyses. Crossbills wereexcluded because they are a forest—dwelling species and were onlyrecorded flying over the plots. Starlings and crows wereexcluded because they are broad habitat generalists, and not ofcurrent management interest.Multiple correlations were done using SAS (Version 6.07,1989), since the necessary SYSTAT module for performing theanalysis was not available. A disadvantage of using SAS,however, was that it did not make pairwise comparisons for caseswith missing variables. This was a concern, as only 36 of the145 cases had arthropod biomass data. Separate multiplecorrelations then, were done for bird and arthropod data.Although possible associations between habitat variables andarthropods may have been overlooked, it was believed that greaterinformation regarding bird and habitat associations would beretained if the arthropod data were analyzed separately.Although SAS provided Pearson correlation coefficients, notests of significance were made because the data were notmultivariate normally distributed, and because this study wasstructured to be descriptive rather than predictive. Pearson25correlation coefficients >0.200 were selected to identify thosehabitat variables most strongly associated with the birdvariables.Following the correlation analyses, cluster analyses wereused to create general management guidelines for grassland birdhabitats. Cluster analyses were done using Euclidean distancemeasures. The single linkage method was used for clustering thespecies diversity index, while the complete linkage method wasused for the most common ground—nesting species (i.e., long—billed curlews, horned larks, vesper sparrows, and westernmeadowlarks). Only ground—nesting species were used to developthe management guidelines, as these birds were considered mostvulnerable to potential grassland disturbances, Study sites wereclustered based on similarities in curlew breeding densities andin abundances of the other species. Means of the habitatvariables associated with the bird variables (as identified bythe correlation analyses) were then calculated for each clusterof study sites. Heterogeneity indices and arthropod variableswere not included in the analyses because of the perceiveddifficulty in trying to manage for these characteristics. SYSTATwas used to perform the cluster analyses.Numbers of upland sandpipers and Sprague’s pipits were toolow to be used effectively in cluster analyses, so overall meansof the habitat variables associated with these species werecalculated for the sites in which the species were recorded.Guidelines were not determined for sharp—tailed grouse or short—26eared owls, even though they are ground—nesters. Sharp—tailedgrouse were not strongly associated with any of the habitatvariables measured in this study, and short—eared owl numberswere very low and their populations so variable (pers. obs.) thatmeaningful guidelines could not be formulated.Additional statistical analyses of the data were done usingchi—square tests and paired-sample, 2—tailed t—tests. T-testswere done with SYSTAT and were used to test for differences incurlew population means between years. A chi—square analysis ofa 4 x 6 contingency table was used to test the nill hypothesisthat there was no difference in habitat structural use among birdspecies. Chi—square analyses were also used for individual birdspecies to test the null hypotheses that there was no differencebetween availability of structural types and use by a givenspecies. A significance level of a = 0.10 was used for the t—tests and chi—square tests.4 Results4.1 Bird CensusesNinety-nine species and 13,584 individual birds wererecorded during the point—count censuses. Two species were verycommon — horned lark and vesper sparrow. Another nine specieswere relatively common (i.e., >1% of all observations): Europeanstarling, western meadowlark, long—billed curlew, Brewer’sblackbird, tree swallow, mountain bluebird, American crow,savannah sparrow, and red crossbill (Table 1). Three red—listed27species (upland sandpiper, Sprague’s pipit, and Brewer’ssparrow), and two blue-listed species in addition to the long-billed curlew were recorded (sharp—tailed grouse and short—earedowl) (Table 1). A complete list of all bird species censused isgiven in Appendix 1.Table 1. Most commona, and red— and blue—listed bird speciescensused in the Chilcotin study area, British Columbia,1991/92.Common name Total % totalobservations observationsMost common speciesHorned lark 4581 33.7Vesper sparrow 3465 25.5European starling 933 6.9Western meadowlark 860 6.3Long—billed curlewb 650 4.8Brewer’s blackbird 365 2.7Tree swallow 308 2.3Mountain bluebird 298 2.2American crow 227 1.7Savannah sparrow 157 1.2Red crossbill 156 1.1Red-listed speciesUpland sandpiper 12 0.1Sprague’s pipit 16 0.1Brewer’s sparrow 7 0.1Blue—listed speciesSharp-tailed grouse 69 0.5Short-eared owl 47 0.3Total 12,151 89.5d >1% of total observations.b Blue—listed species.28Bird densities for the 13 study sites, ranged from 1.09 to1.65 pairs/ha, while species diversity indices ranged from 0.71to 1.38 (Table 2).Table 2. Bird densities and Shannon’s diversity indices by studysite, for the Chilcotin study area, British Columbia,1991/92.Study site Density Shannon’s diversity# pairs/ha indexB_E 1.65 0.87B_W 1.55 1.38C 1.09 1.00D 1.37 0.92J 1.31 0.97M_NE 1.60 0.76M_NW 1.33 0.75M_SE 1.53 0.75S_N 1.47 0.78S_SE 1.51 0.91S_SW 1.41 0.71T_N 1.24 1.08T_S 1.64 0.784.2 Long—billed Curlew Breeding DensitiesDensities of breeding curlews were consistent from 1990 to1992 on most sites, but were higher in 1992 than the previousyears on McDonald’s Ranch-NW (P = 0.05), and the north and southsites of the Toosey Reserve (P = 0.06, P = 0.10)) (Table 3).Compared to those areas surveyed by Ohanjanian (1987), curlewbreeding densities were lower in 1990 (P = 0.05), 1991 (P <0.08),and 1992 (P < 0.09) than in 1987 (Table 4).29Table 3. Long-billed curlew breeding densities within theChilcotin study area, British Columbia, 1990a, 1991,and 1992bTotal 5003 1990 23—32 0.5—0.6 156—2181991 30—35 0.—0.7 143—1671992 37—40 0.7—0.8 125—135Hooper and Savard (1991).b No curlews were recorded on B E, B W, C, or S_SE 1990—1992, orM_SE 1990.C Sites were combined since the pasture was continuous betweenthe two sites.Site area Year Breeding I pairs/ I ha/(ha) pairs 100 ha pairD 1666JMNEMNWM SES_N/S_SWCTNTS1990 3 0.2 5551991 2—3 0.1—0.2 555—8331992 3 0.2 555395 1990 1—2 0.3—0.5 198—3951991 1—2 0.3—0.5 198—3951992 2 0.5 198288 1990 4 1.4 721991 5—6 1.7—2.1 48—581992 3—4 1.0—1.4 72—96292 1990 5—8 1.7—2.7 37—581991 5 1.7 581992 10 3.4 29133 1991 1 0.8 1331992 1 0.8 1331141 1990 6—11 0.5—1.0 104—1901991 10—11 0.9—1.0 104—1141992 7—9 0.4—0.8 127—163687 1990 3 0.4 2291991 4—5 0.6—0.7 137—1721992 6—7 0.9—1.0 98—115401 1990 1 0.3 4011991 2 0.5 2011992 4 1.0 10030Table 4. Long-billed curlew breeding densities within theChilcotin study area, British Columbia,. 1987a, 1990b,1991, and 1992.Site Area Year Breeding # Pairs! I ha!(ha)c pairs 100 ha pairJunction 410 1987 3 0.7 1371990 1—2 0.2—0.5 205—4101991 1—2 0.2—0.5 205—4101992 2 0.5 205Pass Pasture 474 1987 7—8 1.5—1.7 59—681990 2—4 0.4—0.8 119—2371991 5 1.1 951992 4—5 0.8—1.1 95—119South Fraser 470 1987 10 2.1 47Field 1990 1 0.2 4701991 0 0 01992 0 0 0McDonald’s 575 1987 20 3.4 29Ranch—1990 9—12 1.6—2.1 48—64NE & NW 1991 11—12 1.9—2.1 48—521992 13—14 2.3—2.4 41—44Total 1929 1987 40—41 2.1 47—481990 13—19 0.7—1.0 102—1481991 17—19 0.9—1.0 102—1141992 19—21 1.0—1.1 92—102a from Ohanjanian (1987)b from looper and Savard (1991)C area estimates from Ohanjanian (1987)314.3 Bird/Habitat Associations4.3.1 Principal Component AnalysisNine principal components had eigenvalues >1.0, and theseaccounted for 80% of the variability in the habitat variables(Table 5). The first three, and the ninth components hadvariables with component loadings > 0 • 600:. Twenty of the 32habitat variables were retained from these four components. Thefirst component represented vegetation structure as measured byheight and vertical cover, horizontal cover of differentphysiognomic features, and horizontal patchiness. The secondcomponent represented topographic features and vegetationstructure in terms of vertical patchiness and horizontal cover ofshrubs. The third component included one topographic feature,and the grazing variables, while the ninth component wasassociated vegetation structure as measured by the number oftrees in the point—count area.32Table 5. Principal components and component loadings for habitatvariables in the Chilcotin study area, British Columbia,1991/92.pc 1 2 3 9Eigenvalue 6.81 5.69 3.36 1.06% variance 21.3 17.8 10.5 3.3% totalvariance 21.3 39.1 49.6 80.0HabitatvariablebVMt4 nri structureHeight/vertical coverMAYVGCOV -0.763MAYVGHT -0.752JUNVGCOV -0.718JUNVGHT -0.610Horizontal coverBAREGRND 0.807GRASS —0.611LITTER -0.787TREES —0.703SHRUBS 0.805Horizontal patchinessBGRNDHI -0.837Vertical patchinessMAYCOVHI 0.648MAYHTHI 0.666JUNHTHI 0.712Additional structuralfeaturesTREES_O 0.723Topographic featuresELEV —0.773SLOPE 0.866ASPECT 0.659Grazing characteristicsSPRNGAUM 0.692SMMERAUM 0.761FALLAUM 0.813Principal componentb Habitat variable codes given in Appendix 2.334.3.2 General bird/habitat correlationsThe vegetation species composition data were excluded fromthe correlation analyses because a high degree of correlation andlinear relationship among plant species greatly confounded theanalysis. A list of plant species recorded on the study sites,however, is provided in Appendix 3.Topographic features and arthropods had greater mean numbersof correlations with bird diversity and species than didvegetation structure or grazing levels (Table 6).Table 6. Mean number of bird/habitat correlations by generalhabitat characteristic, for the Chilcotin study area,British Columbia, 1991/92.8No. variables/,b n characteristic3.9 54 14HabitatcharacteristicVegetationstructureTopography 6.3 19 3Grazing 2.7 8 3Arthropods 5.3 48 98 results from long—billed curlew point—counts were excluded.b total number of correlations divided by the number of variablesrepresenting that habitat characteristic.Multiple correlations revealed that vegetation structure,topographic features, and grazing characteristics combined,explained 15-96% of the variability in bird diversity and speciesabundances (Table 7). Greatest variability was accounted for inlong—billed curlew breeding densities, followed by species34diversity, and abundances of western meadowlarks, vespersparrows, Brewer’s sparrows, Sprague’s pipits, Brewer’sblackbirds, horned larks, long—billed curlew numbers from point—counts, savannah sparrows, short—eared owls, mountain bluebirds,upland sandpipers, tree swallows, and sharp—tailed grouse (Table7).Arthropod biomasses accounted for 15—72% of the variability inthe bird diversity and species abundances (Table 7). Greatestvariability was accounted for in upland sandpipers, followed bymountain bluebirds, long-billed curlew breeding densities,savannah sparrows, species diversity, vesper sparrows, long—billed curlew numbers from point—counts, sharp—tailed grouse,Sprague’s pipits, western meadowlarks and horned larks, short—eared owls, tree swallows, and Brewer’s blackbirds. Five ofthese bird variables, however, had only, or mainly, negativeassociations with arthropod biomasses — long—billed curlewnumbers, Sprague’s pipits, western meadowlarks, horned larks, andBrewer’s blackbirds, while tree swallows showed no correlationswith arthropods (Table 7). Only three species were positivelyassociated with total arthropod biomass — vesper sparrows,mountain bluebirds, and upland sandpipers.35Table 7. Pearson correlation coefficients for bird/habitatcorrelations, for the Chilcotin study area, BritishColumbia, 1991/92.Bird diversity (H) and speciesHabitat H HOLA VESP LBCU° LBCU_denccharacteristicaVegetationStructureHeight/Vert. coverMAYVGCOV 0.35 —0.32 0.50 —0.31 —0.47MAYVGHT 0.32 —0.31 0.41 —0.27 —0.43JUNVGCOV 0.31 —0.41 0.46 —0.33. —0.48JUNVGHT 0.22 —0.40 0.34 —0.22 —0.37PatchinessJUNHTHI -0.23BGRNDHI -0.26HorizontalcoverGRASS —0.22 —0.31 0.23SHRUBS —0.33 —0.20TopographyELEV 0.29 0.21ASPECT 0.45 —0.40 0.53 —0.32 —0.70SLOPE —0.37 —0.38GrazingSPRNGAUM —0.23 —0.22 —0.24SMMERAUM -0.39FALLAUM 0.21 —0.23 —0.41variability 85% 39% 58% 38% 96%explainedArthropodsANTS 0.20BEES —0.35 0.26BEETLES —0.21 0.31 —0.24BUGS 0.21 —0.22 0.26GRASSHOPPERS -0.20LARVAE —0.35 —0.33 0.31 0.49MOTHS —0.21 —0.25SPIDERS —0.37 —0.36 —0.29TOTAL BIOMASS 0.21 -0.21variability 49% 27% 44% 36% 56%explainedBird species, habitat characteristic codes given in Appendix 2.b based on point—count censuses.C based on spot—mapping censuses (i.e., breeding densities).36Table 7. cont.Bird speciesHabitat WEME BRBI TRES MOBL SAVBcharacteristicVeetat ionStructureHeight/Vert. coverMAYVGCOV 0.34MAYVGHT 0.34JUNVGCOV 0.43JUNVGHT 0.42PatchinessMAYCOVHI 0.32 0.26 —0.23MAYHTHI 0.30JUNHTHI 0.31 —0.24HorizontalcoverBAREGRND 0.22GRASS —0.24LITTER 0.27SHRUBS 0.66 0.50TREES 0.20 0.27TREES_O 0.20ToorahvELEV —0.63 —0.45 0.27ASPECT 0.21SLOPE 0.68 0.43 —0.27GrazingSNNERAUM 0.30FALLAUM —0.20 0.26variability 74% 40% 17% 26% 35%explainedArthropodsANTS —0.23BEES —0.25 —0.32BEETLES -0.21 0.34FLIES —0.27 —0.31GRASSHOPPERS —0.26 0.2 —0.30LARVAE —0.33 —0.35MOTHS —0.28 0.58 —0.28SPIDERS 0.37TOTAL BIOMASS -0.21 0.33variability 27% 15% 19% 64% 50%explained37Table 7. cont.Bird speciesHabitat UPSA SPPI BRSP STGR SEOWcharacteristicVegetationStructurePatchinessMAYCOVHI 0.22MAYHTHI 0.31BGRNDHI 0.35HorizontalcoverBAREGRND —0.38 —0.44GRASS 0.24 0.40LITTER 0.37 0.21SHRUBS 0.21 0.26TREES 0.32 0.64TopographyELEV -0.34ASPECT -0.24SLOPE 0.34 0.28variability 20% 49% 57% 15% 29%explainedArthropodsANTS —0.24BEES 0.29 0.28BEETLES 0.35FLIES 0.57 0.24 0.20GRASSHOPPERS 0.66LARVAE 0.26MOTHS 0.30 0.32SPIDERS 0.44TOTALBIOMASS 0.41variability 72% 30% —c 20%explainedwere not made on the sites where BRSPC arthropod collectionswere recorded.384.3.3 Specific bird/habitat correlationsSpecies DiversityHabitat and arthropod variables accounted for 85% and 49% ofthe variability in species diversity, respectively. Diversityincreased with aspect and was greatest on sites with tall, densevegetation, low grass cover, and low levels of spring grazing.Diversity was also greatest on sites with high ant and bugbiomass, and low bee, grasshopper, larva, and moth biomass (Table7).Horned LarkThirty-nine percent and 27% of the variability in hornedlark numbers was accounted for by the habitat and arthropodvariables, respectively. Horned larks were most abundant ongently sloping, high-elevation sites with low aspect. Larks werealso most common on sites with short, open vegetation, low shrubcover, and reduced patchiness of June vegetation height. Hornedlarks were not positively associated with any of the arthropodtypes (Table 7).Vesper SparrowHabitat variables explained 58% of the variability in vespersparrow abundance, Vesper sparrows were most common on siteswith high aspect, tall, dense vegetation, and low grass cover.Sparrow numbers also increased with fall grazing levels.Arthropod variables accounted for 44% of the variability in39vesper sparrow abundance. Sparrow numbers increased with beetle,bug and total arthropod biomass, but declined on sites with highlarva biomass (Table 7).Long-billed CurlewHabitat variables accounted for 38% of the variability inlong-billed curlew numbers based on point-counts, but 96% of thevariability of curlews based on breeding densities. Curlews weremost common on gently sloping, high-elevation sites with lowaspect. Curlews were also most abundant on sites with short,open vegetation, low shrub cover, high grass cover, and reducedpatchiness of bare ground. Curlew numbers declined as grazinglevels increased.Arthropod variables accounted for 36% and 56% of curlewnumbers and breeding densities, respectively. Curlews were mostabundant in areas with high bee and larva biomass, and leastabundant on sites with high spider and total arthropod biomass(Table 7).Western MeadowlarkSeventy-four percent and 27% of the variability in westernmeadowlark numbers was explained by the habitat and arthropodvariables, respectively. Meadowlarks were most common on steep,low elevation sites with high aspect. Meadowlarks were also mostabundant on sites with tall, dense, patchy vegetation, and highshrub cover. Meadowlark numbers declined with declines in fall40grazing levels and larva biomass (Table 7).Brewer’s BlackbirdHabitat and arthropod variables accounted for 40% and 15% ofvariability in Brewer’s blackbird numbers, respectively.Blackbirds were most abundant on steep, low elevation sites withhigh shrub cover and patchy vertical cover of May vegetation.Brewer’s blackbirds were not positively associated with anyarthropod groups (Table 7).Tree SwallowOnly 17% of the variability in tree swallow abundance wasattributed to the habitat variables. Tree swallows were mostcommon on sites with high tree cover and low patchiness ofvegetation vertical cover in May. Although the arthropodvariables, as a whole, explained 19% of the variability in treeswallow numbers, no strong associations between swallows andindividual arthropod types were identified (Table 7).Mountain BluebirdHabitat variables explained 26% of the variability inmountain bluebird abundance. Bluebird numbers were greatest ongently sloping, high-elevation sites, with high tree and littercover, and low patchiness of June vegetation height. Arthropodsaccounted for 64% of variability in bluebird numbers. Mountainbluebird abundance increased with beetle, grasshopper, moth,41spider, and total arthropod biomass, but declined as ant biomassincreased (Table 7).Savannah SparrowHabitat and arthropod variables accounted for 35% and 50% ofthe variability in savannah sparrow numbers, respectively.Sparrows were most common on sites with low gra&s cover and highbare ground cover, and numbers increased with summer and fallgrazing levels. All correlations between sparrows and arthropodswere negative (Table 7).Upland SandpiperTwenty per cent and 72% of the variability in uplandsandpiper abundance was attributed to the habitat and arthropodvariables, respectively. Sandpipers were found on sites withhigh grass, litter, and tree cover, but low and patchy coverageof bare ground. Sandpiper numbers also increased with allarthropod types except larvae (Table 7).Sprague’s PipitHabitat and arthropod variables explained 49% and 30% of thevariability in Sprague’s pipit abundance. Pipits were found onsites with low aspect, high grass and litter cover, and low bareground cover. Sprague’s pipits were not positively associatedwith any arthropod group (Table 7).42Brewer’s SparrowFifty—seven percent of the variability in Brewer’s sparrowabundance was explained by the habitat variables. These sparrowswere common on steep sites with patchy vertical vegetation, andhigh shrub and tree cover (Table 7).Sharp-tailed GrouseThe habitat variables as a whole, accounted for 15% of thevariability in sharp-tailed grouse numbers, but no strongassociations with individual habitat characteristics were found.Arthropods explained 33% of the variability in grouse abundance,with grouse being most common on sites with high bee, fly, andmoth biomass (Table 7).Short-eared OwlHabitat and arthropod variables accounted for 29% and 20% ofthe variability in short—eared owl numbers, respectively. Owlswere abundant on steep, low—elevation sites with high shrubcover. Short—eared owl numbers also increased with fly and mothbiomass (Table 7).4.3.4 Species Management GuidelinesSpecies DiversityThe cluster analysis separated the species diversity indicesinto three groups associated with low, medium, and high species43diversity (Fig. 3). Trends in habitat associations identified bycluster analysis (Table 8) were similar to those of thecorrelation analysis (Table 7). Highest species diversity was onsites with south-facing aspects and no spring grazing, and withaverage May/June vegetation height of 9-13 cm, vegetationvertical cover of 36—44%, and grass cover of 29%.Horned LarkThe cluster analysis separated horned larks into two groupsassociated with low and high abundance (Fig. 4). Trends inhabitat associations identified by cluster analysis (Table 9)were similar to those of the correlation analysis (Table 7).Highest numbers of horned larks were on east—facing sitesaveraging 939 m elevation, with slope of 40 May/June vegetationheight of 6-10 cm, vegetation vertical cover of 21-30%, and shrubcover of 0.10%.Vesper SparrowThe cluster analysis separated vesper sparrows into twogroups associated with low and high abundance (Fig. 5). Trendsin habitat associations identified by cluster analysis (Table 10)were similar to those of the correlation analysis (Table 7).Highest numbers of vesper sparrows were on south—west—facingsites with average May/June vegetation height of 8-12 cm,vegetation vertical cover of 32—41%, grass cover• of 33%, and fallgrazing levels averaging 548 AUMs.44SiteBW ()HighTN1 MediumDS_SE‘1 IB_ET_SSNMNETtIi LowMNWM_SESSWI I0.1 0.2 0.3 0.4 0.5Euclidean DistanceFig. 3. Cluster analysis of Shannon’s diversity indices bystudy site for the Chilcotin study area, BritishColumbia, 1991/92.45Table 8. Mean habitat variables associated with low, medium, andhigh diversity of bird species in the Chilcotin studyarea, British Columbia, 1991/92.Species diversityLow Medium HighHabitatvariableb (0.71—0.78) (0.91—1.08) (1.38)MAYVGCOV (%) 20 30 36MAYVGHT (cm) 6 8 9JUNVGCOV (%) 30 42 44JUNVGHT (cm) 10 13 13GRASS (%) 39 35 29ASPECT (°) 103 210 192SPRNGAUM 507 541 0a Shannon’s diversity index (H).b Habitat variable codes given in Appendix 2.46SiteD’JIC ___LI IT_NJs_sw I[__B_E IHighMNW_______________I IT_SM_SE}MNE \I I •i1 2 3 4Euclidean DistanceFig. 4. Cluster analysis of horned lark abundances by studysite for the Chilcotin study area, British Colimthia,1991/92.47Table 9. Mean habitat variables associated with low and highabundances of horned larks in the Chilcotin study area,British Columbia, 1991/92.Horned lark abundanceLow HighHabitat (0.15—1.71) (3.13—5.06)variableaMAYVGCOV (%) 33 21MAYVGHT (cm) 9 6JUNVGCOV (%) 45 30JUNVGHT (cm) 12 10SHRUBS (%) 0.60 0.10ELEV (in) 818 939ASPECT (°) 211 119SLOPE (°) 9 4d Habitat variable codes given in Appendix 2.48SiteSSE (NBWII IS_N 1•B_B 1—iI h-IJ I I I IS_SWDTN II !LowM_SE1-iMNW1 II IT_S II IMNE1 2 3 4 5Euclidean DistanceFig. 5. Cluster analysis of vesper sparrow abundances by studysite for the Chilcotin study area, British Columbia,1991/92.49Table 10. Mean habitat variables associated with low and highabundances of vesper sparrows in the Chilcotin studyarea, British Columbia, 1991/92.Vesper sparrow abundanceLow HighHabitat (1.20—2.35) (2.60—3.66)van ableaMAYVGCOV (%) 21 32MAYVGHT (cm) 6 8JUNVGCOV (%) 32 41JUNVGHT (cm) 11 12GRASS (%) 40 33ASPECT (°) 113 208FALLAUM 146 548a Habitat variable codes given in Appendix 2.Long-billed CurlewThe cluster analysis separated long—billed curlews intothree groups associated with low, medium, and high breedingdensities (Fig. 6). Trends in habitat associations identified bycluster analysis (Table 11) were similar to those of thecorrelation analysis (Table 7). Highest curlew breedingdensities were on north—facing sites averaging 940 in elevation,with slope of 30 May/June vegetation height of 5—8 cm,vegetation vertical cover of 14—22%, grass cover of 41%, andshrub cover of 0.3%. Curlew densities were also highest on siteswith no summer or fall grazing, and with spring grazing levelsaveraging 420 AUM5.50SiteMNWTNMNEJH_D MediumI IMSE’’__SSE I- I ILowBW____B_EI IC2 4 6 8 .10Euclidean DistanceFig. 6. Cluster analysis of long—billed curlew breedingdensities by study site for the Chilcotin study area,British Columbia, 1991/92.51HabitatvariableaMAYVGCOV (%) 35MAYVGHT (cm) 9JUNVGCOV (%) 46JUNVGHT (cm) 13GRASS (%) 42SHRUBS (%) 0.06ELEV (m) 953ASPECT (°) 192SLOPE (°) 5SPRNGAUM 410SMMERAUM 265FALLàUM 496a Habitat variable codesWestern MeadowlarkThe cluster analysis separated western meadowlarks intothree groups associated with low, medium, and high abundances(Fig. 7). Trends in habitat associations identified by clusteranalysis (Table 12) were somewhat similar to those of thecorrelation analysis (Table 7). Highest meadowlark numbers wereon south—facing sites averaging 585 m elevation, with slope of17°, May/June vegetation height of 8-15 cm, vegetation verticalcover of 26—43%, and shrub cover of 2%. Meadowlarks were alsomost abundant on sites with fall grazing levels averaging 75AUMs.Table 11. Mean habitat variables associated with low, medium, andhigh breeding densities of long-billed curlews in theChilcotin study area, British Columbia, 1991/92.Long-billed curlew breeding densityLow Medium High(0—2) (3—5) (7)2173211330.58421587569175223given in145228410.39408342000Appendix 2.52SiteSSEedium_MSEI II IB_W IS_N i_____c’H1LowT_NI Is_sw II — IB_E II IM_NEI II IMNWI II I1 2 3 4 .5Euclidean DistanceFig. 7. Cluster analysis of western meadowlark abundances bystudy site for the Chilcotin study area, BritishColumbia, 1991/92.53Table 12. Mean habitat variables associated with low, medium, andhigh abundances of western meadowlarks in the Chilcotinstudy area, British Columbia, 1991/92.Western meadowlark abundancesMedium HighHabitat (0.84—1.40) (2.35).variableaMAYVGCOV (%) 22 45MAYVGHT (cm) 6 12JUNVGCOV (%) 31 57JUNVGHT (cm) 10 16SHRUBS (%) 0.14 0.03ELEV (in) 929 919ASPECT (0) 142 200SLOPE (0) 4 8FALLAUM 331 463a Habitat variable codes given in Appendix 2.Low(0. 02—0. 45)268431525851951775Upland SandpiperUpland sandpipers were found on sites with mean grass coverof 59%, bare ground cover of 13%, litter cover of 24%, and treecover of 0.04%.Sprague’s PipitSprague’s pipits were recorded on sites with mean .aspect of24°, grass cover of 62%, bare ground cover of 27%, and littercover of 12%.4.3.5 Habitat Management GuidelinesThe range of variables associated with highest speciesdiversity and greatest nunthers of individual species was narrower54than that measured throughout the entire study area (Table 13).The range of variables associated with high species diversity andspecies abundances was: 14—44% vegetation vertical cover (May andJune combined), 5-15 cm vegetation height (May and Junecombined), 13—31% bare ground cover, 29—59% grass cover, 17-24%litter cover, 0.10—2.0% shrub cover, 0.04% tree cover, 8—208°aspect, 585—940 in elevation, 3-17° slope, 0—420 spring AUMs, 0summer AUMS, and 0—548 fall AUMs.4.3.6 Habitat Structure Use by BirdsMountain bluebirds were most commonly observed using habitatstructures (71% of all bluebird observations), followed bywestern meadowlarks (50%), vesper sparrows (28%), and hornedlarks (18%). These species did not have similar distributionsamong habitat structures ( < 0 • 001; Appendix 4). Horned larksused rocks and roads more than any other species (Fig. 8).Vesper sparrows used shrubs, western meadowlarks used trees, andmountain bluebirds used fences more than any other species.There were also differences between the availability of habitatstructures and their use by each species ( < 0.001 for eachspecies; Appendix 5). Horned larks used rocks and roads in muchgreater proportions than were available. Vesper sparrows usedtrees, fences, and roads, while western meadowlarks used trees ingreater proportions than were available. Mountain bluebirds usedtrees and fences in much greater proportions than were available.55Table 13. Mean habitat variables associated with highestbird species diversity and species abundances inthe Chilcotin study area, British Columbia,1991/92.Study H HOLA VESP LBCU WEME UPSA SPPIsitesHabitatvariablesaMAYVGCOV (%) 26 36 21 32 14 26(7_66)bMAYVGHT(cm) 7 9 6 8 5 8(3—21)JUNVGCOV (%) 36 44 30 41 22 43(13—75)JUNVGHT (cm) 11 13 10 12 8 15(5—38)BAREGRND (%) 44 13 31(9—58)GRASS (%) 37 29 33 41 59 49(22—63)LITTER (%) 9 24 17(5—39)SHRUBS (%) 0.3 0.1 0.3 2.0(0—2)TREES (%) 0.001 0.04(0—0.04)ASPECT (°) 155 192 119 208 8 195 79(8—348)ELEV (m) 892 939 940 585(585—1000)SLOPE (°) 6 4 3 17(1—21)SPRNGAUM 595 0 420(0—2050)SMMERAUM 261 0(0—1367)FALLAUM 372 548 0 75(0—2050)a Habitat variable codes given in Appendix 2.b Range in variable measured across the 13 study sites.5670605040.‘oo’3020MOBLWEMEEJVESPHOLAS Availability 00S 0S S00SSStructureFig. 8. Habitat structural availability and use by hornedlarks, vesper sparrows, western meadowlarks, andmountain bluebirds in the Chilcotjn study area, BritishColumbia, 1991/92.575. Discussion5.1 Bird CensusesBird census results were typical for grassland habitats.Grassland bird communities are characteristically relativelysimple. Both intra— and inter—continental surveys have shownthat in general, grasslands provide habitat for two to sixpasserine species, and occasionally, as many nonpasserine species(Cody 1966, Wiens 1973b, Wiens 1974a, 1974b, Wiens and Dyer 1975,Cody 1985). In this study, two passerine species were verycommon (horned lark and vesper sparrow), while seven passerinespecies were relatively common (tree swallow, American crow,mountain bluebird, European starling, savannah sparrow, westernmeadowlark, and Brewer’s blackbird). One nonpasserine species —the long—billed curlew, was also common.Within grassland bird communities, one or two widespreadspecies tend to dominate (Graul 1980). Wiens and Dyer (1975)found that throughout North American grasslands, almost 50% ofthe birds recorded were of one species, while the two mostabundant species comprised 75-88% of all observations. In thisstudy, two species also dominated the bird community, but to alesser extent than that found by Wiens and Dyer (1975). Hornedlarks and vesper sparrows, the dominant species in this study,comprised almost 34% and 26% of all observations, respectively,or almost 60% of all observations in total. Differences inproportions of the dominant bird species in this and Wiens andDyer’s (1975) study may be due to differences in bird censusing58techniques. Wiens and Dyer (1975) used spot-mapping or territorymapping techniques, whereas point—count censuses were used inthis study. In a comparison of spot-mapping and point-countcensus techniques, Hooper and Savard (1991) found that morespecies, but fewer numbers of birds were detected with point—counts than with spot-mapping.Additional attributes of the bird communities in this studythat were typical for grassland habitats included bird densitiesand species diversity. Throughout North American grasslands,bird densities typically average 0.5-2.0 pairs/ha (Wiens 1973b,Cody 1985) and species diversity indices range from 0.44 to 1.43(Wiens 1973b, l974a, 1974b). In this study, densities rangedfrom 1.1 to 1.7 pairs/ha and species diversity indices rangedfrom 0.71 to 1.38.Characterization of the breeding bird communities in thestudy area also included records of the first occurrence andbreeding of Sprague’s pipit, and the first breeding of uplandsandpipers in British Columbia (McConnell aj,. 1994, Van denDriessche . in press). More recently, Roberts (1994)recorded first occurrences for the Chilcotin grasslands ofanother three bird species — the American golden plover(Pluvialis dominica), yellow-breasted chat (Icteria virens), andlark sparrow (Chondestes arammacus). This suggests that the birdcommunities of the Chilcotin grasslands may be more diverse thanpreviously thought. These communities, therefore, shouldcontinue to be studied and monitored.595.2 Long—billed Curlew CensusesBreeding densities of curlews from 1990-1992 were fairlyconstant, but were higher on McDonald’s Ranch and the TooseyIndian Reserve in 1992 than in the previous two years. Thissuggests that either curlew numbers throughout the study areaincreased slightly in 1992, or that some curlews changed breedingsites between years. More detailed and long—term breedingdensity censuses throughout the study area are needed to clarifythese trends.Most curlew breeding densities in this study were within therange found in other North American studies. Densities in thisstudy were one pair/29—833 ha: those in other studies were onepair/12—40 ha in Idaho (Jenni et j. 1982), one pair/24 ha atSkookumchuck Prairie, British Columbia (Ohanjanian 1985), onepari/66—136 ha in Washington (Allen 1980), and one pair/600—700ha in Saskatchewan (Sadler and Maher 1976).The relative stability in curlew numbers and breedingdensities from 1990-1992 suggests that inconsistencies in surveymethods, rather than a population decline, may have beenresponsible for the decrease in curlew breeding densities betweenthis study and Ohanjanian’s (1987). Ohanjanian (1987) did notindicate when, or how many surveys she did. If the 1987 surveyswere done during the pre-laying period in early April, counts ofsingle males may have overestimated the number of breeding pairs,since not all males attract a female (Ohanjanian 1985, 1987).Similarly, if counts of males were made during the brood—rearing60period (mid-June through July), breeding pair nuibers may alsohave been overestimated since single males will fly more thanhalf a kilometre to help with cooperative mobbing of predators(Redmond . 1981, Ohanjanian 1987).5.3 General Bird/habitat AssociationsVegetation structure is considered to be th most importantfactor affecting grassland bird distribution (Tester and Marshall1961, Hilden 1965, Wiens 1969, Whittaker and Woodwell 1972, Cody1985). In this study, the principal component analysis indicatedthat vegetation structure accounted for most of the variabilityin the habitat characteristics. The multiple correlationsindicated however, that topographic features and arthropodbiomasses were more often associated with birds than wasvegetation structure.The associations between birds and topographic features mayhave been partially due to correlations between topography andvegetation structure. For example, multiple correlations showedthat in general, vegetation patchiness, shrub cover and bareground cover decreased, and grass, litter, and tree coverincreased as elevation increased (Appendix 6). Steeper slopeswere correlated with taller, denser, patchier vegetation, morebare ground and shrub cover and less grass and litter cover thangently sloping sites. South— and west—facing sites wereassociated with taller, denser, patchier vegetation, less bareground and grass cover, and more shrub and tree cover than north—61and east-facing sites (Appendix 6).Despite these associations, topographic features explainedonly 9-12% of the variability in vegetation structure. Theassociations of birds with topographic features then, remainssomewhat unclear. No other studies were found in which grasslandtopography was characterized and correlated with birdcommunities. It may be that birds cue—in on micrositesassociated with different topographic features. For example,microsite characteristics such as rate of spring snow melt,amount of radiant light, air temperature intensity andfluctuation, and evapotranspiration rates could affect thequality of nesting sites and the reproductive success of breedingbirds. More research is needed to clarify the associationsbetween grassland birds and topography.Although arthropod biomasses in this study were associatedwith birds more than was vegetation structure, many of thebird/arthropod correlations (25 of 48) were negative. Althoughcause and effect can not be concluded from correlation analysis,it is unlikely that grassland birds were so selective in theirfood preferences that certain species avoided using sites due tothe presence of certain arthropod types. If the negativecorrelations were excluded, then vegetation structure would bemore often associated with birds than were arthropod biomasses.Other studies have concluded that guild structures of grasslandbirds were more likely associated with habitat structure thanwith food supply (Folse 1981).62The general lack of positive associations between birdspecies and total arthropod biomass suggests that strong habitatselection pressure based on overall food resourcs, did not occuramong the birds in this study. Although food resources forbreeding grassland birds may, at times, be extremely important,they are probably not a consistent limiting factor on an annualbasis (Wiens 1974a). Because primary productivity in grasslandsoccurs within a short season, and because breeding grasslandbirds have low energy demands, it is likely that food resourcesare often superabundant (Wiens 1974a, 1977, Wiens and Rotenberry1979). Short—term studies like this one, however, could miss anyyears when food may be limiting (Newton 1980). The diets ofgrassland birds can be highly variable between individuals,sites, and years (Wiens and Rotenberry 1979). This too, suggeststhat food resources are rarely limiting, and that individuals andpopulations exploit food resources opportunistically (Wiens andRotenberry 1979).If grassland birds are limited at all by food resources, itmay be due more to temporal or spatial availability than toabundance of arthropods (Wiens l974a). The method used forcollecting arthropods in this study may also have contributed tothe lack of clearly defined relationships between grassland birdsand food resources. For example, widely—dispersed arthropods maynot be adequately sampled with small pan traps, while otherarthropods may be more easily caught by traps than by foragingbirds (Wiens 1977).63The components of vegetation structure most often associatedwith grassland birds include grass height and density (Cody 1966,Creighton 1974, Ohanjanian 1985), litter and vegetationpatchiness (Wiens 1969, 1973b, l974a, l974b, 1976, Wiens and Dyer1975), and amount of ground and shrub cover (Bock i. 1984).Similar results were obtained in this study, but associationsbetween birds and grass and tree cover were also identified.Wiens and Rotenberry (1981) suggested that completeunderstanding of bird/habitat relationships requires knowledge ofvegetation floristics as well as structure — in some cases,floristic data may contribute more to the ability to predictbird/habitat associations than does structural data.Unfortunately, the methods of floristic sampling used in thisstudy, although consistent with those of the Ministry of Forests,did not provide data that produced easily interpreted ormeaningful correlation analyses with the bird data. More precisemeasurements than the mid—points of the six cover classes mayhave provided more useful data, especially for the more uncommonplant species. It should be noted though, that Wiens andRotenberry’s (1981) conclusions about the association of birdsand vegetation floristics were based on studies of shrub—steppebird communities. In these habitats, bird species werecorrelated with different shrub species, which clearly, havestrong structural characteristics. Cody (1968) concluded frominter—continental studies of grasslands (i.e., non—shrub—steppehabitats) that although some bird species have specialized food64and habitat requirements, it is unlikely that most speciesrecognize and exploit differences between species within a plantgenus. Consequently, because only one site in this study hadsome characteristics of a shrub—steppe habitat, thecharacterization of the floristic component of bird habitats wasprobably not as important as that of the vegetation structuralcomponent.Of all the habitat characteristics, livestock grazing hadthe fewest associations with birds. Correlations between birdsand grazing may have been due to the effects of brazing onvegetation structure (Wiens 1973b, Wiens and Dyer 1975). Ingeneral, cover of bare ground increased, but grass, litter,shrub, and tree cover, and vegetation height, cover, andpatchiness decreased as AUMS increased (Appendix 7). The amountof variability in vegetation structure attributed to grazinghowever, was only 5-7% (Appendix 7). This suggests that if birdsdid respond to grazing levels, it may have been due to factorsother than, or in addition to changes in vegetation structure.It is difficult to assess, however, to which factors birds mayhave responded. For example, birds could have been affected bytrampling and disturbance due to grazers. Significant nestlosses can occur at stocking densities greater than 2.5 AU/ha(Jensen 1990), and species such as the long-billed curlewcan experience serious nest losses and abandonments due tolivestock trampling and harassment (Sugden 1933, Jenni1982, Redmond and Jenni 1986). The behavioural responses of65grassland birds to grazers however, were not investigated in thisstudy.The impacts of grazing on birds can also be confounded bydifferences in climatic patterns, soil characteristics, andvegetation floristics and phenology between study sites.Additionally, the use of AUMs, although consistent with theMinistry of Forests methods, may not have adequately assessedgrazing intensity. Experiments on livestock trampling ofsimulated ground nests have shown that more meaningfulassessments of trampling damage are made when numbers of animalsper hectare, rather than livestock weights and forage demands,were used as measures of grazing intensity (Jensen j,. 1990).Comparison of vegetation phenology between study sites at time ofgrazing, and calculations of AUMS on a per unit area basis maythen, have given a more meaningful assessment of grazinginfluences throughout the study area. Conclusions on the impactsof grazing on grassland birds in this study then, should becautious, especially since the effects of soil, site conditions,and precipitation regimes may be more important than grazing inaffecting food, cover, and water for grassland birds (Ryder1980).5.4 Specific Bird/habitat Associations5.4.1 Species DiversityThe habitat variables measured in this study, explained muchof the variability (85%) in bird species diversity throughout the66study area. Bird diversity is often positively correlated withvegetation complexity (Roth 1977, Wiens and Rotenberry 1981, Cody1985). Similar results were found in this study, as diversitywas greatest on sites with the tallest, densest vegetation. Thenegative correlation between diversity and grass cover alsosuggests that diversity was greater on sites with more complexvegetation structure than where grasses dominated. Diversity mayhave been positively correlated with aspect since south—facingsites had greater vegetation height and vertical cover and lessgrass cover than east—facing sites. Species diversity was alsonegatively correlated with spring grazing, possibly becausegrazing reduced vegetation height and cover during the nestingseason. Species richness, a component of species diversity,generally declines as grazing pressure increases (Owens and Myres1973, Wiens l973b, Wiens and Dyer 1975, Kantrud 1981).5.4.2 Horned LarkHorned larks typically nest in bare, sandy, or stony groundwith sparse grass cover (Harrison 1984); consequently, moststudies on horned larks found negative associations with tall,dense vegetation, and forb and shrub cover, and positiveassociations with bare ground (Wiens l973b, Bock and Webb 1984,Wiens and Rotenberry 1985, Larson and Bock 1986). Similarresults were found in this study. Horned larks were leastabundant on sites with tall, dense vegetation, shrub cover, andpatchy vegetation height. Larks may also have been least common67on steep, south—facing sites because vegetation cover and height,and shrub cover increased with slope and aspect. Positiveassociations between larks and elevation may have been due toreduced shrub cover at higher elevations.Horned larks’ affinity for short, open vegetation may bedue, in part, to this species’ means of foraging. Larks arerapid feeders which pursue, rather than search for prey (Cody1968). Pursuing behaviour is often most efficient inhomogeneous, short—statured grasslands (Cody 1968). Hornedlarks’ inability to forage effectively in tall grass can resultin chicks starving to death in the nest (Cody 1985).The diets of horned larks tend to be highly variable,seasonally and geographically, and in composition (Wiens andRotenberry 1979, Rotenberry 1980). Although horned larks areomnivorous, during the breeding season, seeds can comprise up to73% of the diet (Wiens and Rotenberry 1979). This then, mayexplain the lack of positive associations between horned larksand arthropods in this study.Although no association of horned larks with livestockgrazing was found in this study, horned larks generally, are morecommon in more heavily than lightly grazed areas, presumably dueto this species’ association with low, open vegetation, and bareground (Maher 1973, Owens and Myres 1973, Wiens l973b, Karasiuk. 1977, Ryder 1980, Kantrud 1981, Kantrud and Kologiski1983, Bock j,. 1984, Renken and Dinsmore 1987).Observations of habitat structural use revealed the68importance of rocks and roads to horned larks. Rocks were usedmore than other structures for singing or perching posts.Although this study did not find correlations between hornedlarks and bare ground, larks did use dirt roads for feeding anddust-bathing sites. The importance of roadways to this specieswas also noted at a site in Newfoundland, where horned larkterritories were linearly distributed along a road that malesused for dust-bathing, roosting, and singing (Cannings andThrefall 1981). In this study, large numbers of horned larkswere also seen using dirt roads as evening roost sites. Hornedlarks are known to dig roost sites, often behind protectivevegetation, to reduce radiative and convective heat losses atnight (Trost 1972).5.4.3 vesper SparrowVesper sparrows were most common on sites with tall, densevegetation and low grass cover. Sparrow numbers were alsogreater on south— than east—facing sites, possibly because oftaller, denser vegetation and less grass cover in those areas.The negative correlation between sparrows and grass cover in thisstudy, suggests that sparrow abundance may have been greater onsites with more complex vegetation structure than where grassesdominated. Other studies have found that vesper sparrows werepositively correlated with vegetation density, ahd litter andground cover, but negatively correlated with cover of bare groundaround nest sites (Wray and Whitmore 1979, Reed 1986).69Given the habitat associations for this species, vespersparrows might be expected to be most common in ungrazed orlightly grazed grasslands. In some studies this was true (Maher1973, Kantrud and Kologiski 1983), in others, it was not (Owensand Myres 1973, Kantrud 1981). Vesper sparrows use a range ofhabitat types including grasslands, shrub-steppes, andgrassland/shrub-steppe ecotones (Johnsgard 1979, Kantrud andKologiski 1983, McNicholl 1988). Vesper sparrows also nest on orabove the ground, and frequently defend territories as large as0.8 ha (Johnsgard 1979, Kantrud and Kologiski 1983). Theserather broad habitat associations could explain any differencesin habitat descriptions and grazing results between studies. Itis difficult then, to determine what, if any, impacts livestockgrazing has on vesper sparrows. In this study, the positivecorrelations of sparrows with grazing is unclear, but could be aresult of reduced grass cover in spring due to fall grazing.The data on habitat structure use indicated that vespersparrows used shrubs for perching and displaying, more than anyother species. Similarly, Hooper and Savard (1991) foundpositive correlations between vesper sparrows and shrub cover.Vesper sparrows also used trees and fences in greater proportionsthan were available. Vesper sparrows may be associated withshrubs, trees, and fences, because of their use of elevatedsinging perches for territorial defense (Johnsgard 1979, Terres1980). Unlike horned larks, which sing from the ground or inflight, vesper sparrows sing primarily from elevated perches70(Castrale 1983). In the Okanagan Region of B.C., vesper sparrowshave also been found using shrubs (i.e., sagebrush bushes) fornesting cover (Cannings et al. 1987). In such areas, livestockgrazing may have benefitted vesper sparrows by increasing theamount of available sagebrush cover (Cannings i. 1987).5.4.4 Long-billed CurlewMore habitat associations were found using curlew breedingdensities than curlew numbers from point—counts, and thoseassociations explained much greater variability in curlewdensities (96%) than in curlew numbers (38%). This suggests thatstudies which use breeding territory censuses may identify moredetailed bird/habitat associations than were generally found inthis study.Long—billed curlews were most common in short, open, grassyvegetation with low shrub cover. Curlews were also most commonon gently sloping, high elevation sites with low aspects. Theseassociations with topography may be due to low vegetation heightand vertical cover, low shrub cover, and high horizontal cover ofgrass on those sites. Long-billed curlews typiclly nest inshort, open grasslands, and may require these sites for communalpredator detection, effective communication between nestingbirds, and ease of movement of chicks when feeding (McCallum. 1977, Allen 1980, Renaud 1980, Bicak 1982, Jenni. 1982, Ohanjanian 1986, 1987). Curlews may also requirepatchy grass cover for camouflage and thermal cover for chicks71(Allen 1980, Pampush 1980).The negative associations between curlews and grazing inthis study were somewhat confusing, since grazing generallyreduces vegetation height and density. Grazing studies haveconsistently found that curlews are more abundant in heavier—than lighter-grazed grasslands (King 1978, Bicak 1982,Kantrud and Kologiski 1982, Ohanjanian 1987). If curlews in thisstudy were responding negatively to grazing, it may be due todisturbance effects from grazing animals. In southwestern Idaho,10% of long-billed curlew egg clutches were abandoned whennesting birds were harassed by grazing livestock (Jenni g1982). Curlew nest abandonments have been shown, to be influencedby stocking rates, duration and frequency of grazing, and timingof grazing during the incubation period (Jenni . 1982).Greater insight into the potential influence of livestock grazingon long-billed curlews in the Chilcotin grasslands may be gainedfrom more research on the behavioural responses of curlews tograzing animals, and on the reproductive success of curlews onsites with different grazing regimes.5.4.5 Western MeadowlarkIn this study, western meadowlarks were commonly associatedwith shrub cover, and tall, dense, patchy vegetation cover. Anincrease in meadowlark numbers with slope and aspect may havebeen due to greater shrub cover and vegetation height, cover, andpatchiness, on steeper, south—facing sites. Meadowlarks may also72have been more common at lower elevations due to greater shrubcover and vegetation patchiness on those sites. Observations ofhabitat structural use indicated that trees, which were used forperching and displaying, were also important habitat componentsfor meadowlarks.Western meadowlarks build their nests in tall grass. Nestsare usually well hidden in grass clumps, and consist of a scrapedbowl covered with a domed canopy of grasses and forbs (Johnsgard1979, Harrison 1984). Meadowlarks also use elevated singingposts (Terres 1980, Cannnings j,. 1987). This use of tallgrass for nesting, and elevated structures for singing posts mayexplain why meadowlarks in this study were associated with talldense vegetation, shrub cover, and trees. The general lack ofassociation of meadowlarks with arthropods in this study, may bea result of the broad composition and geographically variablediets typical of western meadowlarks (Wiens and Rotenberry 1979).Although other studies have found positive correlationsbetween western meadowlarks and shrub cover and vegetationvertical structure (Wiens and Rotenberry 1981), meadowlarks areconsidered habitat generalists that occupy a range of habitatsfrom tall—grass prairies, wet meadows, hayfields, and weedyborders of croplands to short—grass and sage prairies (Johnsgard1979, Rotenberry and Wiens 1981, Larson and Bock 1986). Becausewestern meadowlarks tend to be habitat generalists then, grazingstudies have shown that meadowlarks are associated both withgrazed and ungrazed grasslands (Johnson 1972, 1973, 1974, Owens73and Myres 1973, Hopkins 1980, Kantrud 1981, Kantrud and Kologiski1983, Renken and Dinsmore 1987). In this study, westernmeadowlarks generally decreased with increased fall grazinglevels, possibly because fall grazing resulted in reduced springvegetation height, cover, and patchiness on breeding sites.5.4.6 Brewer’s BlackbirdBrewer’s blackbirds nest in a variety of habitats, but aremost often associated with shrubs and trees near moist meadowsand fields, which are used as feeding sites (Harrison 1984,Cannings 1987). This may explain why blackbirds in thisstudy were most strongly associated with shrub cover. Brewer’sblackbirds were also most common on steeply sloping, low—elevation sites, possibly due to greater shrub cover on thosesites. The variability in nesting habitats of Brewer’sblackbirds may explain why habitat associations for this specieswere otherwise, not well defined in this study.5.4.7 Tree SwallowAlthough tree swallows are found in a variety of habitats inthe Okanagan Region of B.C., they are most common around lakesand ponds that have open spaces for insect hunting, andstructures with nest holes unobstructed by vegetation (Cannings1987). In this study, tree swallow numbers increased withtree cover, presumably because of the potential nest sitesafforded by trees. The variability in nesting habitats of tree74swallows however, and the fact that point—counts were generallylocated away from water bodies, may explain why habitatassociations for tree swallows were generally not well defined inthis study.5.4.8 Mountain BluebirdMountain bluebirds were most common on gently sloping, high—elevation sites with trees. Observations of habitat structuraluse indicated that fences were also important habitat featuresfor bluebirds. Although mountain bluebirds are associated withopen grasslands, scrublands, and treeless meadows, they requirecavities in which to nest (Cannings 1987). This explainsthe association between bluebirds and trees in this study. Italso explains the use of fences by bluebirds, since bluebird nestboxes have commonly been placed on fences by local ranchers andthe Williams Lake Naturalists.5.4.9 Savannah SparrowSavannah sparrows numbers increased with amount of bareground, and declined as grass cover increased. These resultscontradicted other studies of savannah sparrow habitatassociations. Savannah sparrow breeding territories arecharacterized by vertically dense vegetation, grass cover, forbdensity and cover, and litter depth and patchiness (Wiens 1973a).Nests are usually built in dense ground cover where they arewell-concealed by overhanging vegetation (Linsdale 1938, Tester75and Marshall 1961, Lein 1968, Wiens 1969, Potter 1972, Terres1980).Savannah sparrows also tend to occupy wet meadow zones ofmid— and tall—grass prairies (Johnsgard 1979). Generalobservations suggested that savannah sparrows in this study werealso most common in wet areas. Because point—counts weregenerally located away from large wet areas, it is more likelythat savannah sparrow habitat associations were not adequatelyassessed in this study, rather than that sparrows had differenthabitat requirements in the study area.Savannah sparrows are most common in areas with little or nolivestock grazing (Lincoln 1925, Rand 1948, Maher 1973, Owens andMyres 1973, Karasiuk . 1977, Page 1978, Maher 1979,Kantrud 1981, Kantrud and Kologiski 1983). In this study,savannah sparrows were positively associated with spring and fallgrazing levels. Again, the contradictions between this and otherstudies may be because savannah sparrow habitat associations werenot adequately assessed in this study.5.4.10 Upland SandpiperIn B.C., upland sandpipers have been recorded using open,grasslands, overgrown fallow fields, bogs, burns, wet pastures,golf courses, lawns, meadows, dirt roads, and mudflats (Campbell. 1990). No details on nesting habitat in B.C. wereavailable, however, before this study. The nest site in thisstudy was a grassy alcove surrounded on three siçles by trees and76shrubs. Grasses comprised 75—95%, while bare soil comprisedonly 5—25% of the horizontal ground cover (Van den Driesschej., in press). This would explain why upland sandpipers in thisstudy were negatively associated with bare ground, and positivelyassociated with grass and tree cover. Conclusions about uplandsandpiper habitat requirements should be cautious though, becausenumbers of breeding sandpipers in the study area were so low(e.g., 12 observations representing an estimated two breedingpairs).Other studies have recorded the upland sandpiper’sassociation with mixed-grass and tall-grass habitats (Wiens1973b, Rotenberry and Wiens 1980); consequently, this speciestends to be most common in ungrazed to moderately grazed areas(Kantrud 1981, McNicholl 1988). No associations of sandpiperswith grazing were found in this study, possibly due to the lownumbers of this species within the study area.Arthropods accounted for more variability in uplandsandpiper numbers than any other species. Sandpipers also hadmore positive correlations with different arthropod types thanany other species. This suggests that food resources may beimportant factors in habitat selection for upland sandpipers inthe study area. Given the rather vague results of relationshipsbetween the other bird species and food resources in this study,however, conclusions on the importance of arthropods to uplandsandpiper habitat selection within the study area, should not bemade without more detailed studies on upland sandpiper feeding77ecology.5.4.11. Sprague’s PipitSprague’s pipits tend to be associated with extensive areasof grasslands dominated by grasses of medium height (Johnsgard1979). Pipits nest on the ground, in growing herbage, and nestsare well—concealed by overhanging vegetation (Harrison 1984).Similar nesting characteristics were found in this study(McConnell et al. 1994), and explain why pipits were positivelyassociated with grass and litter cover, and negatively correlatedwith cover of bare ground. Conclusions about Sprague’s pipithabitat requirements based on this study should be cautioushowever, since pipit numbers were so low (e.g., 16 observationsrepresenting one breeding pair).Given this species’ association with tall grass habitats,Sprague’s pipits are generally most common in ungrazed tomoderately areas (Maher 1973, Owens and Myres 1973, Kantrud andKologiski 1983). The lack of association between pipits andgrazing in this study may have been due to the low numbers ofpipits recorded.5.4.12 Brewer’s SparrowBrewer’s sparrows are characteristic of sagebrush grasslandsin British Columbia (Cannings j,. 1987). This may explain whyBrewer’s sparrows in this study, were most common on steeplysloping sites with patchy vegetation, and high shrub and tree78cover. More detailed habitat associations for Brewer’s sparrowsmay have been obtained if more sagebrush sites had been censused.These sites were under—represented in this study because theywere often inaccessible by vehicle, and/or too small toaccommodate a sufficient number of point—counts. Thedistribution of Brewer’s sparrow throughout the Chilcotin remainsunclear however, as the species was not recorded in the areauntil 1992 (A. Roberts, pers. comm., Roberts 1994). More surveysare needed then, to characterize Brewer’s sparrow habitatassociations throughout the study area.5.4.13 Sharp—tailed GrouseSharp-tailed grouse occupy a variety of habitats, butcommon features of these habitats include open grasslandsadjacent to brushy or scattered open woodlands (Campbell g1990). In the Chilcotin, open parklands adjacent to spruce,Douglas—fir, or trembling aspen stands are characteristic sharp—tailed grouse habitat (Campbell . 1990). Although somepoint—counts were located near the grassland/woodland ecotone,habitat measurements were generally made in open grassland. Thismay explain why strong habitat associations were not found forsharp—tailed grouse in this study.Additionally, Moyles (1981) found that sharp—tailed grouseused different sites within the grassland/woodland border duringdifferent seasons and different times of the day. In the earlymornings of spring, (i.e., when point—counts were conducted),79grouse feed in trees and shrubby borders, as well as in opengrasslands (Moyles 1981); consequently, censuses of the opengrasslands may have underestimated sharp—tailed numbers in thestudy area.Sharp-tailed grouse habitat requirements would probably bestbe determined then, by sampling habitat characteristics at knownlekking, nesting, and feeding sites. In the interim, maintenanceof a mosaic of open grasslands associated with extensiveshrub/tree ecotones, may provide optimal habitat for sharp-tailedgrouse in the study area (Moyles 1981).5.4.14 Short-eared OwlShort—eared owls were most comnmon in shrubby areas, and onsteeply sloping sites at low elevations. Because mostobservations were of hunting birds, however, little informationabout habitat requirements for this species was obtained fromthis study. Behavioural observations of habitat use, anddetailed habitat measurements around nest sites are needed toidentify true habitat requirements for short-eared owls in thestudy area.5.5 Habitat Structural Use by BirdsObservations of habitat structural use by birds identifiedhabitat associations that were not revealed in the correlationanalyses. This illustrates the importance of includingbehavioural observations in grassland bird habitat studies to80identify meaningful bird/habitat associations.6. ConclusionsThis study successfully met the research objectives ofcharacterizing grassland breeding bird communities of the studyarea, and of identifying associations between grassland habitatcharacteristics and breeding bird diversity and speciesabundances. The research hypothesis that bird diversity andspecies abundances were most often associated with vegetationstructure was not fully supported by this study. Birds were mostoften associated with topographic features. This may have beendue to correlations between topography and vegetation structure,or it may have been due to microsite differences associated withtopography. Nevertheless, vegetation structure was considered tobe the second factor of greatest influence on bird communities inthis study.As predicted, species diversity was greater in morestructurally complex habitats, given that complexity wasconsidered to be based primarily on vegetation height, verticalcover, and patchiness. Also as predicted, some species were morestrongly associated with structurally complex habitats than wereother species. For example, horned larks and long—billed curlewswere characteristic of short, open grass habitats, whereas vespersparrows and western meadowlarks were associated with tall, densegrass sites.The research hypothesis that breeding bird diversity and81species abundances were also associated with food resources wasnot supported in this study. This may have been due to samplingmethods that did not adequately sample types and numbers, and/ortime and spatial distributions of arthropods, or it may have beendue to a general lack of close coupling between grassland birdsand food resources.This study did not support the hypothesis that birddiversity and species abundances were affected by livestockgrazing. The prediction that grazing affected birds by alteringhabitat structure was supported somewhat, since vegetationheight, vertical cover, and patchiness declined with increasedgrazing pressure. The amount of variability in structuralcomponents accounted for by grazing however, was very low. Birddiversity declined with grazing, but species characteristic ofshort—grass habitats were not more abundant at higher grazinglevels, nor were the tall—grass species consistently moreabundant at lower grazing levels.The inability to define trends between grassland birds andlivestock grazing may mean that AUM5 did not adequately assessthe impacts of grazing intensity on birds, or that behaviouralinteractions between birds and livestock were occurring. Due tothe problems with defining associations between birds and foodresources, the prediction. that livestock grazing affectedgrassland birds by altering food resources, was not examined.827. Management RecommendationsThis study has provided preliminary information on grasslandbird communities of some of the Chilcotin grasslands, and throughthe identification of species habitat associations and thecreation of species management guidelines, has provided somedirection in managing grasslands to maintain avian diversity.Based on this study, grassland habitats in the study area shouldbe managed as a mosaic of habitat types ranging from short, openvegetation to tall, dense vegetation. Maintenance of rocks,shrubs, trees, and grass/shrub and grass/tree ecotones withinthese mosaics is also important. It may be possible to uselivestock grazing to create and/or maintain these habitatmosaics.It should be noted however, that the bird/habitatassociations and management guidelines in this study are specificto the study area, and are not necessarily applicable to otherareas or other time periods. Management and maintenance of aviandiversity throughout the Chilcotin-Cariboo grasslands shouldtherefore, be based on more detailed information than wasprovided in this study. A greater understanding of grasslandbird communities throughout the region could be achieved byundertaking the following procedures:1. Monitoring of grasslands should be done throughout a widerarea and longer time period than was covered in this study.From 1991—1993, five species of birds previously unknown inthe Chilcotin grasslands were recorded (McConnell831994, Roberts 1994, Van den Driessche et flj., in press).This suggests that the bird communities of these grasslandsmay be more diverse than previously thought. Future birdinventories should be made initially for five consecutiveyears, and then repeated one year in a five—year interval.These inventories should be made throughout the Cariboo andChilcotin grasslands. Also, despite the recent work byRoberts (1994), detailed information on birds of the lowelevation (i.e., sagebrush) grasslands is lacking in thisregion. These sites were too inaccessible to be properlycensused in this study;2. Future bird censuses should be done using spot—mappingrather than by point—count censuses. Point—counts were usedin this study because they are more efficient and providemore representative sampling over larger areas than doesspot-mapping (Verner 1985). Spot-mapping, however, providesbetter estimates of bird densities (Verner 1985). Becausespot-mapping is based on plotting of breeding territories,it would be easier to identify and measure the habitatvariables most strongly associated with those territories.This may provide more detailed and accurate habitatassociations than were found in this study;3. The correlation analyses in this study identifiedbird/habitat associations only. Cause and effect of birdhabitat use can not be identified by these analyses. Futureresearch should incorporate behavioural observations of84habitat use with bird censuses. As this study has shown,behavioural observations can identify impor.tant habitatrequirements that may be overlooked by simply measuringhabitat features around locations of bird sightings. Byrecording habitats used for such activities as mating,nesting, feeding, and roosting, and by noting theavailability of those habitats versus their use, details ofspecies habitat selection can be determined;4. The reproductive success of individual bird species indifferent grassland habitats should also be studied, as thiscould further identify and clarify bird/habitatassociations;5. Given the variability in grassland bird diets and foodresources, it is questionable whether further studies onfood resources would clarify bird/habitat associations inthe Chilcotin grasslands. If such studies are considerednecessary, information on diets of individual species, andon prey presence, abundance, and accessibility through timeand space should be collected; and6. Although the correlation analyses in this study identifiedsome associations between birds and livestock grazing, thecause for those relationships remains unclear. Thepotential impacts of livestock grazing on grasslands birdsmay be best determined through long-term replicated grazingtrials. Such trials would account for natural variabilityin bird populations, and would make statistical inferences85possible. Where replicated grazing trials are not feasible,livestock numbers, season and duration of grazing, and plantphenology and standing crop biomass on the study sitesshould be recorded and correlated with vegetation structure.Differences in climatic patterns and site conditionsbetween study areas should also be noted, as these caninfluence the effects of grazing on vegetation. In thisstudy, AIJM5 were used to quantify grazing intensity, as thisis the method used by the Ministry of Forests. A moreappropriate measure of grazing intensity may be the numberof livestock/ha. In addition to these recommendations,behavioural observations of disturbance effects of grazinglivestock on birds should be made. Experiments of livestocktrampling of artificial ground nests, and studies of thereproductive success of birds in grasslands with differentgrazing regimes would help clarify the potential impacts oflivestock grazing on birds.866 Literature CitedAllen, J.N. 1980. The ecology and behavior of the Long-billedCurlew in southwestern Washington. Wildl. Monogr. 73:3-67.American Ornithological Union. 1983. Check-list of North Americanbirds. Sixth ed. Allen Press, Inc., Lawrence, Kansas. 877pp.__________________________1984. Report of the meeting of thecommittee on classification and nomenclature. Auk 101:348.•1985. Thirty-fifth supplement tothe American Ornithologist’s Union Check-list of NorthAmerican birds. Auk 102:680-686.________1987. Thirty-sixth supplement tothe American Ornithologist’s Union Check-list of NorthAmerican birds. Auk 104:591—596.1989. Thirty-seventh supplement tothe American Ornithologist’s Union Check-list of NorthAmerican birds. Auk 106:532—538.Baldwin, P.H. 1971. Diet of the mountain plover at the PawneeNational Grassland, 1970—71. U.S. IBP Grassland Biome Tech.Rep. No. 134. 22pp. jjj R.A. Ryder 1980. Effects of grazingon bird habitats. Pages 51-64 Management of westernforests and grasslands for non—game birds. U.S. For. Serv.Gen. Tech. Rep. INT-86, Ogden, Utah.1973. The feeding regime of granivorous birds inshortgrass prairie in Colorado, U.S.A. Pages 237-248 jn S.C.Kendeigh and J. Pinowski, eds. Productivity, populationdynamics and systematics of granivorous birds. PolishScientific Publishers, Warsaw. 4lOpp. j R.A. Ryder 1980.Effects of grazing on bird habitats. Pages 51—64 jManagement of western forests and grasslands for non—gamebirds. U.S. For. Serv. Gen. Tech. Rep. INT—86, Ogden, Utah.Bicak, T,K., R.L. Redmond, and D.A. Jenni. 1982. Effects ofgrazing on Long-billed Curlew breeding behavior and ecologyin southwestern Idaho. Pages 74-85 Wildlife-livestockrelationships symposium: Proc. 10. J.M. Peek and P.D. Dalke,eds. Univ. Idaho For. Wildi. Range Exp. Stn., Moscow.Bock, C. E., J.H. Bock, W.R. Kennedy, and V.M. Hawthorne. 1984.Responses of birds, rodents, and vegetation to livestockexciosure in a semidesert grassland site. J. Range Manage.37:239—242.,and B. Webb. 1984. Birds as grazing indicator speciesin southeastern Arizona. J. Wildi. Manage. 48:1045-1049.87Breymeyer, A.I., and G.M. van Dyne. 1980. Grasslands, systemsanalysis, and man. Cambridge University Press, New York.950pp.British Columbia Ministry of Environment. 1991. Managing wildlifeto 2001: a discussion paper. B.C. Mm. Environ., Wildi.Branch, Victoria, B.C. 152pp.British Columbia Ministry of Environment, Lands,, and Parks. 1993.Red and Blue Lists. B.C. Mm. Environ. Lands, and Parks,Wildl. Branch, Victoria, B.C.Burt, W. H., and R.P. Grossenheider. 1976. A field guide to themammals. Third ed. Houghton Mifflin, Boston. 289pp.Cameron, E.S. 1907. The birds of Custer and Dawson counties,Montana. Auk 24:241-270.Campbell, R.W., N.K. Dawe, I. McTaggart-Cowan, J.M. Cooper, G.W.Kaiser, and M.C.E. McNall. 1990. The birds of BritishColumbia. Vol. 2. Royal British Columbia Museum, Can. Wildi.Serv. 514pp.Cannings, R.A., R.J. Cannings, and S.G. Cannings. 1987. Birds ofthe Okanagan Valley, British Columbia. Royal BritishColumbia Mus., Victoria, B.C. 420pp.Cannings, R.J. 1981. Notes on the nesting of Horned Larks on theChilcotin Plateau of British Columbia. Murrelet 62:21-23._____________,and w. Threfall. 1981. Horned Lark breedingbiology at Cape St. Mary’s, Newfoundland. Wilson Bull.93:519—530.Castrale, J.S. 1983. Selection of song perches by sagebrush—grassland birds. Wilson Bull. 95:67—655.Cody, M.L. 1966. The consistency of intra- and inter-continentalgrassland bird species counts. Am. Nat. 100:371—376._____•1968. On the method of resource division in grasslandbird communities. Amer. Nat. 102:107—147.1985. Habitat selection and open—country birds. Pages191-226 j Habitat selection in birds. M.L. Cody, ed.Academic Press.88Creighton, P.D. 1974. Habitat exploitation by an avian ground-foraging guild. Ph.D. Thesis, Colorado State Univ., Ft.Collins. 139pp. j R.A. Ryder. 1980. Effects of grazing onbird habitats. Pages 51-64 Management of western forestsand grasslands for non—game birds. U.S. For. Serv. Gen.Tech. Rep. INT-86, Ogden, Utah.Daubenmire, R. 1978. Plant geography. Academic Press. New York,New York. 338pp.Demarchi, D.A. 1988. A regional wildlife ecosystem classificationsystem for British Columbia. Pages 11—19 j H.A. Stelfox andG.R. Ironside, compilers. Land/Wildlife integration workshopNo. 3, Mount Ste.—Marie, Quebec, 16-19 September 1985.Ecological Land Classification Series No. 22. Can. Wildl.Serv., Ottawa, Ont.Douglas, G.W., G.B. Straley, and D. Meidinger. 1989. The vascularplants of British Columbia. Part 1: Gymnosperms anddicotyledons (Aceraceae through Cucurbitaceae). Special Rep.Series 1. B.C. Mm. Forests, Victoria, B.C. 208pp._________________________,and____________.1990. The vascularplants of British Columbia. Part 2: Dicotyledons(Diapensiaceae through Portulaceae). Special Rep. Series 2.B.C. Mm. Forests, Victoria, B.C. 158pp.and 1991. The vascularplants of British Columbia. Part 3: Dicotyledons(Primulaceae through Zygophyllaceae and Pteridophytes).Special Rep. Series 3. B.C. Mm. Forests, Victoria, B.C.177pp.Folse, L.J., Jr. 1981. Ecological relationships of grasslandbirds to habitat and food supply in East Africa. Pages 160-166 in The use of iuultivariate statistics in studies ofwildlife habitat. D.E. Capen, ed. USDA For. Serv. Gen. Tech.Rep. RN-87.Graul, W.D. 1980. Grassland management and bird communities.Pages 38—47 jfl Management of western forests and grasslandsfor non—game birds. U.S. For. Serv. Gen. Tech. Rep. INT—86,Ogden, Utah.Harrison, C. 1984. A field guide to nests, eggs, and nestlingsof North American birds. Second ed. Collins, Toronto. 416pp.Hewitt, G., E.W. Huddleston, R.J. Lavigne, D.N. Ueckert, and J.G.Watts. 1974. Rangeland entomology. Soc. Range Mgmt. RangeSci. Ser. No. 2. 127pp.89Hilden, 0. 1965. Habitat selection in birds. Ann.. Zool. Fenn.2:53—75. jj J.A. Wiens. 1973. Pattern and process ingrassland bird communities. Ecol. Monogr. 43:237-270.Hooper, T.D., and M.D. Pitt. 1994. Problem analysis forChilcotin-Cariboo grassland biodiversity. B.C. Mm.Environ., Lands and Parks, Wildl. Branch, Williams Lake,B.C., 2lOpp.Hooper, T.D., and J.-P. Savard. 1991. Bird diversity, density,and habitat selection in the Cariboo—Chilcotin grasslands:with emphasis on the Long-billed Curlew. Technical ReportSeries No. 142. Canadian Wildlife Service, Pacific and YukonRegion, British Columbia. lO2pp.Hope, G.D., W.R. Mitchell, D.A. Lloyd, W.R. Erickson, W.L.Harper, and B.M. Wikeem. 1991. Interior Douglas-fir Zone.Pages 153-166 j3 D. Meidinger and J. Pojar, compilers andeds. Ecosystems of British Columbia. B.C. Mm. For. Spec.Rep. Ser. 6, Victoria, B.C.Hopkins, R.B. 1980. Mixed prairie I. Am. Birds 34:67—68. H.A.Kantrud. 1981. Grazing intensity effects on. the breedingavifauna of North Dakota grasslands. Can. Field-Nat. 95:404-417.Jenni, D.A., R.L. Redmond, and T.K. Bicak. 1982. Behavioralecology and habitat relationships of Long-billed Curlews inwestern Idaho. Rep. Bur. Land Manage., Boise Idaho. 234pp.j l.A. Ohanjanian. 1985. The Long—billed Curlew, (Numeniusamericanus), on Skookumchuck Prairie — status report andenhancement plan. B.C. Mm. Environ., Wildl. Branch Rep.Jensen, H.P., D. Rollins, and R.L. Gillen. 1990. Effects ofcattle stock density on trampling loss of simulated groundnests. Wildi. Soc. Bull. 18:71—74.Johnsgard, P.A. 1979. Birds of the Great Plains. Univ. Nebr.Press, Lincoln, Nebr. 539pp.Johnson, D.H. 1972. Breeding bird populations of selectedgrasslands in east-central North Dakota. Am. Birds 26:971-975. j H.A. Kantrud. 1981. Grazing intensity effects on thebreeding avifauna of North Dakota grasslands. Can. Field—Nat. 95:404—417.____________1973. Breeding bird populations of selectedgrasslands in east-central North Dakota. Am. Birds 27:989-990. jn H.A. Kantrud. 1981. Grazing intensity effects on thebreeding avifauna of North Dakota grasslands. Can. FieldNat. 95:404—417.90____________1974. Breeding bird populations of selectedgrasslands in east-central North Dakota. Am. Birds 28:10301031..jfl H.A. Kantrud. 1981. Grazing intensity effects onthe breeding avifauna of North Dakota grasslands. Can.Field—Nat. 95:404—417.Kantrud, H.A. 1981. Grazing intensity effects on the breedingavifauna of North Dakota grasslands. Can. Faeld—Nat. 95:404—417.,and R.L. Kologiski. 1982. Effects of soils andgrazing on breeding birds on uncultivated upland grasslandsof the Northern Great Plains. USD1 Res. Pap. No. 15. J.fl K.D.DeSmet. 1989. Status report on the Long-billed Curlew(Numenius americanus) in Canada. COSEWIC Draft Rep.,Dept.Res., Winnipeg, Man._____and______________.1983. Avian associations ofthe northern Great Plains. J. Biogeo. 10:331—350.Karasiuk, D., H. Vriens, J.G. Stelfox, and J.R. McGillis. 1977.Study results from Suffield, 1976. Pages E33-E44 jfl Effectsof livestock grazing on mixed prairie range and wildlifewithin PFRA pastures, Suffield Military Reserve. J.G.Stelfox, compiler. Range-Wildlife Study Committee, Can.Wildi. Serv., Edmonton, Alta.Kelly, G.D., and W.W. Middlekauff. 1961. Biological studies ofDissosteira surcata Saussure with distributional notes onrelated California species (Orthoptera: AcrLdidae).Hilgardia 30:395—424.King, R. 1978. Habitat use and related behaviors of breedingLong—billed Curlews. M.S. Thesis, Colorado State Univ., Ft.Collins, Cob. 69pp.Kbopateck, J.M., R.J. Olson, C.J. Emerson, and J.L. Jones. 1979.Land-use conflicts with natural vegetation in the UnitedStates. Environ. Sci. Div. Publ. No. 1333, 22pp. Natl. Tech.Inform. Ser., U.S. Dept. Commerce, Springfield, Virginia22161. j W.D. Graul. 1980. Grassland management practicesand bird communities. Pages 38—47 j3 Management of westernforest and grasslands for non—game birds. U.S. For. Serv.Gen. Tech. Rep. INT-86, Ogden, Utah.Larson, D.L., and C.E. Bock. 1986. Determining avian habitatpreference by bird-centered vegetation sampling. Pages 37-43.in Wildlife 2000: Modelling habitat relationships forterrestrial vertebrates. J.L. Verner, M.L. Morrison, andC.J. Ralph, eds. Univ. Wisconsin Press, Madison, Wis.91Lein, M.R. 1968. The breeding biology of the Savannah SparrowPasserculus sandwichensis (Gmelin) at Saskatoon,Saskatchewan. M.A. Thesis, Univ. Saskatchewan, Saskatoon.l7lpp. .j11 H.A. Kantrud. 1981. Grazing intensity effects onthe breeding avifauna of North Dakota grasslands. Can.Field— Nat. 95:404—417.Lincoln, P.c. 1925. Notes on the bird life of North Dakota withparticular reference to the si.umuer waterfowl. Auk 42:50—64.jfl H.A. Kantrud. 1981. Grazing intensity effects on thebreeding avifauna of North Dakota grasslands. Can. Field-Nat. 95:404—417.Linsdale, J. 1938. Environmental responses of vertebrates in theGreat Basin. Am. Midl. Nat. 19:1—206..j D..E. Medin andW.P. Clary. 1990. Bird and small mammal populations in agrazed and ungrazed riparian habitat in Idaho. U.S.D.A. For.Serv. Intermountain Res. Stn. Res. Pap. INT-425.Little, E.L., Jr. 1977. Research in the pinyon-juniper woodland.Pages 8—19 j E.F. Aldon and T.J. Loring, tech. coord.Ecology, uses and management of pinyon—juniper woodlands:Proceedings of the workshop. USDA For. Serv. Gen. Tech. Pap.RM—39.Magurran, A.E. 1988. Ecological diversity and its measurement.Princeton Univ. Press, Princeton, N.J. l79pp.Maher, W.J. 1973. Birds: I. Population dynamics. Canadiancommittee for the International Biological Programme.Matador Project, Tech. Rep. 34. H.A. Kantrud. 1981.Grazing intensity effects on the breeding avifauna of NorthDakota grasslands. Can. Field—Nat. 95:404—417.__________1979. Nestling diets of prairie passerine birds atMatador, Saskatchewan, canada. Ibis 121:437—452.Martin, J.E.H. 1977. The insects and arachnids of Canada: Part 1:Collecting, preparing, and preserving insects, mites, andspiders. Canada Dept. of Agric. Pub. 1643. 182pp.McCallum, D.A., W.D. Graul, and R. Zaccagnini. 1977. The statusof the Long-billed Curlew in Montana. Auk 94:599-601. JnK.D. DeSmet. 1989. Status report on the Long-billed Curlew(Numenius americanus) in Canada. COSEWIC Draft Rep., Dept.Res., Winnipeg, Man.McConnell, S.D., R. Van den Driessche, T.D. Hooper, G.L. Roberts,and A. Roberts. 1994. First occurrence and breeding ofSprague’s Pipit Anthus spragueii, for British Columbia. Can.Field—Nat. 107:222—223.92McNicholl, M.K. 1988. Ecological and human influences on Canadianpopulations of grassland birds. Pages 1-25 Jj1 Ecology andconservation of grassland birds. P.D. Goriup, ed.International Council for Bird Preservation Tech. Pub. No.7.Meidinger, D. 1987. Recommended vernacular names for commonplants of British Columbia. B.C. Mm. For., Res. Branch,Victoria, B.C. 64pp.Moyles. D.L.J. 1981. Seasonal and daily use of plant communitiesby Sharp—tailed Grouse (Pediocetes Dhasianellus) in theparklands of Alberta. Can. Field—Nat. 95:287—291.Newton, I. 1980. The role of food in limiting bird numbers. Ardea68:11—30.Nicholson, A., E. Hamilton, W.L. Harper, and B.M. Wikeem. 1991.Bunchgrass zone. Pages 125—137 Jfl D. Meidinger and J. Pojar,compilers and eds. Ecosystems of British Columbia. B.C. Mm.For. Spec. Rep. Ser. 6, Victoria, B.C.Ohanjanian, l.A. 1985. The Long-billed Curlew, (Numeniusamericanus), on Skookumchuck Prairie — status report andenhancement plan. B.C. Mm. Environ., Wildi. Branch Rep.52pp.______________1986. The Long-billed Curlew in the EastKootenay — status report and enhancement schedule forSkookumchuck Prairie. B.C. Mm. Environ., Wildl. Branch Rep.l2pp.1987. Status report and managementrecommendations for the Long—billed Curlew (Numeniusamericanus) on the Junction. Mm. Environ, and Parks, Wildi.Branch, Williams Lake, B.C. 25pp.Owens, R.A., and M.T. Myres. 1973. Effects of agriculture uponpopulations of native passerine birds of an Alberta fescuegrassland. Can. J. Zool. 51:697—713.Page, J.L., N. Dodd, T.O. Osborne, and J.A. Carson. 1978. Theinfluence of livestock grazing on non—game wildlife. Cal—Neva. Wildl. 1978:159—173. Jj R.A. Ryder. 1980. Effects ofgrazing on bird habitats. Pages 51-64 Management ofwestern forests and grasslands for non—game birds. U.S. For.Serv. Gen. Tech. Rep. INT-86, Ogden, Utah.Pampush, G.J. 1980. Status report on the Long-billed Curlew inthe Columbia and northern Great Basins. Unpub. USFWS Rep.,Portland, Ore. j K.D. DeSmet. 1989. Status report on theLong-billed Curlew (Numenius americanus) in Canada. COSEWICDraft Rep., Dept. Res., Winnipeg, Man.93Pepper, G.W. 1972. The ecology of the sharp-tailed grouse duringspring and summer in the aspen parkiands of Saskatchewan.Sask. Dept. Nat. Res. Wildi. Rep. No. 1.Potter, P.E. 1972. Territorial behavior in Savannah Sparrows insoutheastern Michigan. Wilson Bull. 84:48-59. .in H.A.Kantrud. 1981. Grazing intensity effects on the breedingavifauna of North Dakota grasslands. Can. Field—Nat. 95:404-417.Rand, A.L. 1948. Birds of southern Alberta. Nat. Mus. Can. Bull.111. H.A. Kantrud. 1981. Grazing intensity effects on thebreeding avifauna of North Dakota grasslands. Can. Field-Nat. 95:404—417.Redmond, R.L., T.K. Bicak, and D.A. Jenni. 1981. An evaluation ofbreeding season census techniques for Long—billed Curlews(Numenius americanus). Pages 197-201 .jj Estimating numbersof terrestrial birds. C.J. Ralph, ed. Studies in avianbiology. No. 6. 197—201. jfl I. A. Ohanjanian. 1985. TheLong—billed Curlew, (Numenius americanus), on SkookumchuckPrairie — status report and enhancement plan. B.C. Mm.Environ., Wildi. Branch Rep.____________,and D.A. Jenni. 1985. Note on the diet of Long-billed Curlew chicks in western Idaho. Great Basin Nat.45: 85—86.and________.1986. Population ecology of theLong—billed Curlew (Numenius americanus) in western Idaho.Auk 103:755—767.Redpath, K. 1990. Identification of relatively undisturbed areasin the South Okanagan. Canadian Wildlife Service, Delta,B.C. jj1 W.L. Harper, E.C. Lea, and R.E. Maxwell. 1992.Biodiversity inventory in the South Okanagan. B.C. Mm.Environ., Wildlife Branch, Victoria, B.C. l6pp.Reed, J.M. 1986. Vegetation structure and Vesper Sparrowterritory location. Wilson Bull. 98:144—147.Renaud, W.E. 1980. The Long-billed Curlew in Saskatchewan: statusand distribution. Blue Jay 38:221-237. .j K.D. DeSmet. 1989.Status report on the Long-billed Curlew (Numeniusamericanus) in Canada. COSEWIC Draft Rep., Dept. Res.,Winnipeg, Man.Renken, R.B., and J.J. Dinsmore. 1987. Nongame bird communitieson managed grasslands in North Dakota. Can. Field—Nat.101:551—557.94Roberts, A. 1992. A report on the ecology of the JunctionWildlife Management Area. B.C. Mm. Environ., Wildl. Branch,Williams Lake, B.C. 63pp._________1994. Biodiversity. B.C. Mm. Environ., Lands, andParks, Wildi. Branch, Williams Lake, B.C.Rogers, L., R. Lavigne, and J.L. Miller. 1972. Bioenergetics ofthe western harvester ant in the shortgrass plainsecosystem. Environ. Entomol. 1:763—768.Rotenberry, J.T. 1980. Dietary relationships among shrubsteppepasserine birds: competition or opportunism in a variableenvironment? Ecol. Monogr. 50:93—110.________,and J.A. Wiens. 1978. Nongame bird communitiesin northwestern rangelands. Pages 32—46 j Workshop onnongame bird habitat management in the coniferous forests ofthe western United States. R.M. DeGraff, ed. USDA For.Serv. Gen. Tech. Rep. PNW-64.and_________.1980. Habitat structure,patchiness, and avian communities in North American steppevegetation: a multivariate analysis. Ecology 6:1228-1250.Roth, R.R. 1977. The composition of four bird communities insouth Texas brush-grasslands. Condor 79:417-425.Ryder, R.A. 1980. Effects of grazing on bird habitats. Pages 51-64.j Management of western forests and grasslands for non—game birds. U.S. For. Serv. Gen. Tech. Rep. INT—86, Ogden,Utah.Sadler, D.A.R. and W.J. Maher. 1976. Notes on the long—billedcurlew in Saskatchewan. Auk 93:382—384.Sas Institute, Inc. 1989. SAS. Cary, N.C.Shotwell, R.L. 1958. The grasshopper your sharecropper. Univ.Missouri Agr. Exp. Stn. Bull. 714. l6pp.Skinner, R.M. 1975. Grassland use pattern and prairie birdpopulations in Missouri. Pages 171—180.jfl M.K. Wali, ed.Prairie: A multiple review. Univ. North Dakota Press, GrandForks, N.D.Smith, C.C. 1940. The effect of overgrazing and erosion upon thebiota of the mixed-grass prairie of Oklahoma. Ecol. 21:381-397.95Springfield, H.W. 1976. Characteristics and management ofsouthwestern pinyon—juniper ranges: the status of ourknowledge. USDA For. Serv. Res. Pap. RM—160.Southwood, T.R.E. 1978. Ecological methods: with particularreference to the study of insect populations. Second ed.Chapman and Hall, London. 524pp.Sugden, J.W. 1933. Range restriction of the Long-billed Curlew.Condor 35:3-9.Systat, Inc. 1991. SYSTAT: The system for statistics. Evanston,IL.Taylor, R.L., and B. MacBryde. 1977. Vascular plants of BritishColumbia: a descriptive resource inventory. Univ. B.C. Tech.Bull. No. 4. Univ. B.C. Press, Vancouver, B.C. 754pp.Terres, J.K. 1980. The Audubon Society encyclopedia of NorthAmerican birds. Alfred A. Knopf, New York. llo9pp.Tester, J.R., and W.H. Marshall. 1961. A study of certain plantand animal interrelationships on a native prairie innorthwestern Minnesota. Occas. Pap. Minn. MUS. Nat. Hist.8:1—51. in R.A. Owens and M.T. Myres. 1973. Effects ofagriculture upon populations of native passerine birds of anAlberta fescue grassland. Can. J. Zool. 51:697—713.Townsend, J.E., and P.J. Smith. 1977. Proceedings of the seminaron improving fish and wildlife benefits in range management.USD1 Fish and Wildl. Serv. Prog., Washington, D.C. FWS/OBS77/1. Ji R.A. Ryder 1980. Effects of grazing on birdhabitats. Pages 51—64 ilL Management of western forests andgrasslands for non—game birds. USDA For. Serv. Gen. Tech.Rep. INT—86.Trost, C.H. 1972. Adaptations of Horned Larks (Eremohilaalpestris) to hot environments. Auk 80:506-527.Van den Driessche, R., S.D. McConnell, and T.D. Hooper. In Press.First breeding record of the Upland Sandpiper Bartramialoncicauda, for British Columbia. Can. Field—Nat.Verner, J. 1985. Assessment of counting techniques. Pages 247-302Current ornithology. R.P. Johnston, ed. Vol. 2. PlenumPub. Co., New York.Whittaker, R.H., and G.M. Woodwell. 1972. Evolution of naturalcommunities. Pages 137—159 j Ecosystem structure andfunction. J.A. Wiens, ed. Oregon State Univ. Press,Corvallis, Ore. Jj J.A. Wiens. 1973. Pattern and process ingrassland bird communities. Ecol. Monogr. 43:237—270.96Wiens, J.A. 1969. An approach to the study of ecologicalrelationships among grassland birds. Ornith. Monogr. 8:1-93..jfl J.A. Wiens. 1973. Pattern and process in grassland birdcommunities. Ecol. Monogr. 443:237—270.__________•1973a. Interterritorial habitat variation inGrasshopper and Savannah Sparrows. Ecology 54:877—884.1973b. Pattern and process in grassland birdcommunities. Ecol. Monogr. 443:237-270.l974a. Climatic instability and the “ecologicalsaturation” of bird communities in North Americangrasslands. Condor 76:385—400.1974b. Habitat heterogeneity and avian communitystructure in North American grasslands. Am. Midland Nat.91:195—213._1975. Avian communities, energetics, and functionsin coniferous forest habitats. Pages 146-182.jn Proceedingsof the symposium on management of forest and range habitatsfor nongame birds. Dixie R. Smith, tech. coord. USDA For.Serv., Wash., D.C.1976. Population responses to patchy environments.Ann. Rev. Ecol. Syst. 7:81-120..jfl R.A. Ryder 1980. Effectsof grazing on bird habitats. Pages 51-64.Management ofwestern forests and grasslands for non—game birds. U.S. For.Serv. Gen. Tech. Rep. INT-86, Ogden, Utah.__1977. Model estimation of energy flow in NorthAmerican grassland bird communities. Oecologia 31:135—151._ _and M.I. Dyer. 1975. Rangeland avifauna: theircomposition, energetics and role in the ecosystem. Pages146—181 jj Symposium on management of forest and rangehabitats for nongame birds, Tucson, Arizona. USDA For. Serv.Gen. Tech. Rep. WO-l._,and J.T. Rotenberry. 1979. Diet nicherelationships among North American grassland and shrubsteppebirds. Oecologia 42:253—292.and______________•1981. Habitat associationsand community structure of birds in shrubsteppeenvironments. Ecol. Monogr. 51: 21—41.and_______ .1985. Response of breedingpasserine birds to rangeland alteration in North Americanshrubsteppe locality. J. Applied Ecol. 22:655—668.97World Wildlife Fund Canada. 1989. Prairie conservation actionplan: 1989—1994. World Wildlife Fund Canada, Toronto,Ontario. 38pp.Wray, T., and R.C. Whitiuore. 1979. Effects of vegetation onnesting success of Vesper Sparrows. Auk 96:802—805.98Appendix 1. Bird species censused in the Chilcotin study area,British Columbia, 1991/92.Common Name Taxonomic NameCommon LoonGreat Blue HeronCanada GooseGreen-winged TealMallardNorthern PintailBlue-winged TealNorthern ShovelerGadwal 1Lesser ScaupWhite-winged ScoterBarrow’ s GoldeneyeBuffleheadCommon MerganserBald EagleNorthern HarrierSharp-shinned HawkNorthern GoshawkRed-tailed HawkAmerican KestrelMerlinGyrfalconBlue GrouseRuffed GrouseSharp-tailed GrouseVirginia RailSoraSandhill CraneKilldeerGreater YellowlegsLesser YellowlegsSpotted SandpiperUpland SandpiperLong-billed CurlewCommon SnipeBlack TernMourning DoveGreat Horned OwlShort-eared OwlCalliope HummingbirdRufous HummingbirdLewis’ WoodpeckerRed-naped SapsuckerDowny WoodpeckerHairy WoodpeckerNorthern FlickerOlive-sided FlycatcherGavia iimiierArdea herodiasBranta canadensisAnas creccaAnas platyrhynchosAnas acutaAnas discorsAnas clveataAnas streperaAythva affinisMelanitta fuscaBucehala islandicaBucehala albeolaMergus merganserHal iaeetus leucocephalusCircus cyaneusAccipiter striatusAccipiter qentilisButeo lamaicensisFalco sparveriusFalco columbariusFalco rusticolusDendracrapus obscurusBonasa u!nbellusTympanuchus phasianellusRallus limicolaPorzana carolinaGrus canadensisCharadrius vociferusTrincra melanoleucaTringa flavipesActitis inaculariaBartramia lonciicaudaNumenius americanusGallinago crallinagoChlidonias nigerZenaida macrouraBubo vircrinianusAsio flammeusStellula calliopeSe1ashorus rufusMelanerpes lewisSphyrapicus ruberPicoides pubescensPicoides villosusColates auratusContous borealis99Appendix 1. cont.Common Name Taxonomic NameWestern Wood—peweeAlder FlycatcherWillow FlycatcherLeast FlycatcherDusky FlycatcherPacific-slope FlycatcherSay’s PhoebeWestern KingbirdHorned LarkTree SwallowViolet-green SwallowNorthern Rough-winged SwallowCliff SwallowBarn SwallowBlack-billed MagpieAmerican CrowCommon RavenBlack-capped ChickadeeMountain ChickadeeRed-breasted NuthatchHouse WrenMarsh WrenRuby-crowned KingletMountain BluebirdVeeryHermit ThrushAmerican RobinSprague’s PipitEuropean StarlingWarbling VireoOrange—crowned WarblerYellow WarblerYellow-rumped WarblerNorthern WaterthrushWilson’s WarblerWestern TanagerLazuli BuntingChipping SparrowClay—coloured SparrowBrewer’s SparrowVesper SparrowSavannah SparrowLincoln’s SparrowGolden—crowned SparrowWhite—crowned SparrowRed-winged BlackbirdWestern MeadowlarkYellow—headed BlackbirdContopus sordidulusEmDidonax alnorumEmpidonax trailliiEmDidonax minimusEmpidonax oberholseriEmpidonax difficilisSavornis savaTyrannus verticalisEremophila alpestrisTachycineta bicolorTachycineta thalassinaStelcridopteryx serriennisHirundo pvrrhonotaHirundo rusticaPica picaCorvus brachvrhvnchosCorvus coraxParus atricapillusParus qambeliSitta canadensisTroglodytes aedonCistothorus palustrisRegulus calendulaSialia currucoidesCatharus fuscescensCatharus cruttatusTurdus micrratoriusAnthus spracrueiiSturnus vuicTarisVireo gilvusVermivora celataDendroica petechiaDendroica coronataSeiurus noveboracensisWilsonia pusillaPirancra ludovicianaPasserina amoenaSpizella passerinaSpizella pallidaSpizella breweriPooecetes crramineusPasserculus sandwichensisMelospiza lincolniiZonotrichia atricapillaZonotrichia leucophrvsAcrelaius phoeniceusSturnella necriectaXanthocephalus xanthocehalus100Appendix 1. cont.Common Name Taxonomic NameBrewer’s Blackbird EuDhagus cyanoceha1usBrown-headed Cowbird Molothrus aterRed Crossbill Loxia curvirostraPine Siskin Carduelis minus101Appendix 2. Bird species and habitat variable codes.Bird speciesHOLA - horned larkVESP — vesper sparrowLBCU - long-billed curlewWEME — western meadowlarkBRBL - Brewer’s blackbirdTRES — tree swallowMOBL - mountain bluebirdSAVS — savannah sparrowUPSA - upland sandpiperSPPI - Sprague’s pipitBRSP — Brewer’s sparrowSTGR - sharp-tailed grouseSEOW - short-eared owlHabitat variablesMAYVGCOV - vegetation cover in MayMAYVGHT - vegetation height in MayJUNVGCOV - vegetation cover in JuneJUNVGHT- vegetation height in JuneBAREGRND- % cover of bare groundGRASS- % cover of grassLITTER- % cover of litterSHRUBS- % cover of shrubsTREES- % cover of treesTREES_O — number of trees in the point—count areaBGRNDHI - heterogeneity index for % cover of bare groundMAYCOVHI - heterogeneity index for % vegetation cover in MayMAYHTHI - heterogeneity index for vegetation height in MayJUNHTHI - heterogeneity index for vegetation height in JuneELEV - elevationSPRNGAUM - number of spring grazing AUM5SMMERAUM - number of summer grazing AUMsFALLAUM - number of fall grazing AUMS102Appendix 3. Plant species recorded in the Chilcotin study area,British Columbia, 1992.Taxonomic Name Common NameAchillea millefoliumAqoseris alauca var. dasycephalaAaropvron cristatumAarovron sicatumAcrropvron trachycaulumAllium cernumAndrosace septentrional isAnemone multifidaAntennaria dimorphaAntennaria microphyllaAntennaria parvifloraAntennaria umbrinellaArabis holboelliiArtemisia campestrisArtemisia dracunculusArtemisia friciidaAstragalus acirestisAstracialus miserAstragalus tenellusBalsamorhiza sagittataBromus inermisBromus tectorumCalochortus macrocarpusCarex spp.Cerastium arvenseChrysothamnus nauseosusvar. albicaulisCirsium undulatumComandra umbellataCrepis atrabarbaDistichlis strictaErigeron compositus var. cilabratusErigeron flaciellarisErigeron linearisErigeron speciosus var. siecioususErioqonum heracleoidesFestuca saximontanaGaillardia aristataGalium borealeGeranium viscosissimumvar. viscosissimumGrindelia squarrosaHedysarum boreale ssp. mackenziiKoeleria macranthaLappula redowskii var. redowskiiLinum perenne ssp. lewisiiLithospermum ruderaleyarrowshort—beaked agoseriscrested wheatgrassbluebunch wheatgrassslender wheatgrassnodding oflionnorthern fairy—candelabracut—leaved anemonelow pussytoesrosy pussytoesNuttall ‘S pussytoesumber pussytoesHoelboell’s rockcressnorthern wormwoodtarragonprairie sagewortfield milk—vetchtimber milk-vetchpulse milk-vetcharrow—leaved balsamrootsmooth bromecheatgrasssagebrush mariposa lilysedge spp.field chickweedrabbit-brushwavy-leaved thistlepale comandraslender hawksbeardalkali saltgrasscut—leaved daisytrailing fleabaneline-leaved daisyshowy daisyparsnip—flowered buckwheatRocky Mountain fescuebrown—eyed Susannorthern bedstrawsticky purple geraniumcurly—cup gumweednorthern hedysarum5 unegrasswestern stickseedwestern blue flaxlemonweed103Appendix 3. cont.Taxonomic Name Common NameLomatium macrocarpumOpuntia fragilisOrthocarus luteusPenstemon procerus var. proceruscompressaPoa iuncifolia£2 pratensisPoa sandbergiiPotentilla gracilisPotentilla hippianaRosa acicularisRosa nutkanaSenecio canusSilene drummondii var. drummondiiSisyrinchium montanumSd idacio spathulataSpartina ciracilisSorobolus cryptandrusStipa comataStipa occidentalisStipa richardsoniiStipa sparteaSymphoricarpos occidental isTracioocion dubiusTracioocion pratensisZ ipadenus venenosuslarge-fruited desert-parsleybrittle prickly-pear cactusyellow owl—cloversmall — flowered penstemonCanada bluegrassalkali bluegrassKentucky bluegrassSandberg’ s bluegrassgraceful cinquefoilwoolly cinquefoilprickly roseNootka rosewoolly groundselDrummond’s ,ampionmountain blue—eyed—grassspike-like goldenrodalkali cordgrasssand dropseedneedle—and—thread grassstiff needlegrassspreading needlegrassporcupine grasswestern snowberryyellow salsifymeadow salsifymeadow death—camas104kppendix 4. Chi-square analysis of habitat structural use amonghorned larks, vesper sparrows, western meadowlarks,and mountain bluebirds in the Chilcotin study area,British Columbia, 1991/92.H0: There is no difference in habitat structural use amongthe four bird species.Ha: Habitat structural use is different among the four birdspecies.Species Habitat structure type TotalRocks Shrubs Trees Fences Logs RoadsHOLAf18 690 17 9 9 24 214 963F1b 358 142 245 68 19 132VESPf. 277 321 347 114 20 142 1221F. 454 179 311 86 24 167WEMEf. 72 61 303 15 9 32 492F. 183 72 125 34 10 68MOBLf. 32 24 74 64 4 7 205F1 76 30 52 14 4 28Total 1071 423 733 202 57 395 2881Observed frequency1) Expected frequencyX2= 1530.6P(X(15)=1530.6<0.001)Therefore, reject H0105Appendix 5. Chi-square analyses of habitat structural use byhorned larks, vesper sparrows, westernmeadowlarks, and mountain bluebirds in theChilcotin study area, British Columbia, 1991/92.H0: There is no difference between proportions of habitatstructures available and those used by each bird species.Ha: Proportions of habitat structures available and used by eachbird species are different.Species Habitat structure type Total(proportion available)Rocks Shrubs Trees Fences Logs Roads(0.503) (0.329) (0.118) (0.003) (0.028) (0.017)HOLAfa 690 17 9 9 24 214 963F1b 484.4 316.8 113.6 2.9 30.0 16.4x2 = 2862.1 P(X2(5)=2862.l<0.OO1) Therefore, reject H0 forHorned LarksVES Pf- 277 321 347 114 20 142 1221F1 614.2 401.7 144.1 3.7 34.2 20.8x2 = 4487.2 P(X(5)=4487.2<0.001) Therefore, reject H0 forVesper SparrowsWEMEf1 72 61 303 15 9 32 492F1 124.4 161.9 58.1 1.5 13.8 8.4x2 = 1409.1 P(X(5)=l409.1<0.001) Therefore, reject H0 forWestern MeadowlarksMOBLf. 32 24 74 64 4 7 205F. 103.1 67.4 24.2 0.62 5.7 3.5x2= 1530.6 P(X(5)=1530.6<0.001) Therefore, reject H0 forMountain Bluebirds8 Observed frequencyI) Expected frequency106Appendix 6. Multiple correlations between vegetation structureand topographic features for the Chilcotin studyarea, British Columbia, 1991/92.Topographic FeaturesVegetationStructure Elevation Slope AspectMAYVGCOV 0.11 0.15 0.48MAYVGHT 0.02 0.22 0.46JUNVGCOV —0.05 0.31 0.43JUNVGHT —0.14 0.31 0.32BAREGRND —0.27 0.15 —0.01GRASS 0.18 —0.19 —0.36LITTER 0.26 —0.22 0.20SHRUBS —0.80 0.80 0.17TREES 0.15 —0.11 0.42TREES_O 0.07 0.16 0.12MAYCOVHI —0.47 0.48 0.05MAYHTHI —0.36 0.45 005JUNHTHI —0.41 0.45 0.15BGRNDHI 0.14 —0.04 0.39variabilityexplained 10% 12% 9%107Appendix 7. Multiple correlations between vegetation structureand seasonal grazing (AUMs) for the Chilcotin studyarea, British Columbia, 1991/92.Grazing LevelsVegetationStructure SPRING-AU!’! SUNNER-AUM FALL-AUMMAYVGCOV —0.24 —0.15 —0.10MAYVGHT —0.27 —0.18 —0.14JUNVGCOV —0.22 —0.20 —0.12JUNVGHT —0.21 —0.24 —0.19BAREGRND 0.42 0.41 0.37GRASS —0.49 —0.34 —0.39LITTER —0.35 —0.26 —0.28SHRUBS 0.05 —0.22 —0.20TREES —0.20 —0.12 —0.12TREES_O —0.06 0.02 0.13MAYCOVHI 0.01 —0.13 —0.13MAYHTHI —0.13 —0.23 —0.22JUNHTHI —0.05 —0.22 —0.22BGRNDHI —0.30 —0.17 —0.18variabilityexplained 7% 5% 5%108


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items