POPULATION DYNAMICS, HABITAT USE AND MOVEMENTS OF GREATER SAGE-GROUSE IN MOFFAT COUNTY, COLORADO A Thesis Presented in Partial Fulfillment of the Requirements for the Degree ofMaster of Science With a . Major in Wildlife Resources In the College of Graduate Studies UNNERSITY OF IDAHO • By Doris Hausleitner July 8, 2003 Major Professor: Kerry P. Reese, Ph.D . • QL(pq~ &2 '6S" • • . 11 ·J13g; ~3 AUTHORIZATION TO SUBMIT THESIS This thesis of Doris Hausleitner, submitted for the degree of Master of Science with a major in Wildlife Resources and titled "Population dynamics, habitat use and movements of greater sage-grouse in Moffat County, Colorado" has been reviewed in final form, as . indicated by the signatures and dates given below. Permission is now granted to submit final copies to the College of Graduate Studies for approval. Major Professor Committee Members Department Administrator College Dean if2;£J~ Date 7-3 ~~w~~ Patricia A. Talcott Date lJ.f~.oi 7~~ Date I p~/05 ~Kerry Paul Reese ate Final Approval and Acceptance by the College of qraduate Studies ~ I THE UNIVERSITY OF IDAHO LIBRARY MANUSCRIPT THESIS The literary rights in an unpublished thesis submitted for the Master's degree and deposited in the University of Idaho Library are vested in the Regents of the University. This thesis is open for inspection, but it is to be used only with due regard for the literary rights involved . • • • 111 ABSTRACT • Long-term declines in greater sage-grouse (Centrocercus urophasianus) abundance may be associated with low productivity due to habitat degradation. Therefore, I investigated the effects of vegetation characteristics on nest site selection and success, and created predictive models using forward stepwise logistic regression. Greater sage-grouse nested in sites having greater sagebrush canopy cover, nest shrub height and visual obstruction than ... • random sites. Forb cover was greater at nest sites than random but grass cover and % exotic species at the nest bowl were greater at random sites than nest sites. Nest success was high with 58% (n = I08) of hens hatching 2: I egg. Nest success was positively associated with plant species richness, percent sagebrush, grass and forb cover and negatively associated with exotic herbaceous cover. Additionally, grass heights measured at the nest bowl and at I m from the nest bowl were parameters associated with nest success. Vegetation was reduced in 2002, a year with less spring precipitation, and may be associated with lower nest success. Habitat management to encourage greater sage-grouse nest success should promote native herbaceous understory growth in April and May and discourage sagebrush removal. • • • Habitat management guidelines are based, in part, on studies of greater sage-grouse nest site selection, which measure habitat characteristics post-hatch,> 30 days after initiation. Thus, these recommendations may be flawed due to the time lag in sampling. I investigated differences in habitat at initiation versus hatch using previous year's nest sites. A randomized complete block MANOVA indicated differences based on timing of sampling. The difference was largely due to grass height and percent grass cover. Grass heights at the nest bowl and at I m were I 0 em and 9 em at initiation and I6 em and I3 em at hatch, • IV respectively. Percent grass cover of the sample plot increased from 4 to 6% cover from initiation to hatch. Sampling techniques conducted at hatch sufficiently describes nest site habitat at selection apart from grass heights and grass cover. Despite extensive research into the habitat requirements of greater sage-grouse, summer night-roosting habitat has received no attention. Therefore, I investigated the vegetation characteristics of brood and night roost selection arid compared diurnal and nocturnal habitat use and selection. Predictive models were created using forward stepwise logistic regression. Greater sage-grouse selected brood-use areas with greater visual obstruction, greater forb cover and less bare ground than random. ' Night-roost locations were characterized by less bare ground and visual obstruction but greater percent forb cover than at random sites. Mean shrub cover and shrub height at night roost locations was lower (9% vs. 22%) and shorter (31 vs. 58 em) than at brood sites. Forb cover was an important predictor of both diurnal and nocturnal habitat use but did not meet the minimum requirements of 2: 1Oo/o suggested by management guidelines. Current habitat management guidelines attend only to diurnal habitat needs and should be revised in order to address the variance in temporal habitat selection. I described home range, daily movements and survival of greater sage-grouse .. monthly, seasonally and annually. Median daily movement during the brood-rearing period was greater for females with broods (184m/day) than those without broods (103m/day), and · was not associated with chicks/female at 6 weeks post-hatch. Lek to nest distance and brood home range sizes were greater for females that nested in the Axial Basin (4.6 km, 1,151 ha) than the Danforth Hills (2.6 km, 439 ha) indicating that habitat conditions may be causing greater movement in the Axial Basin. Distances moved from the last diurnal location to v • night roost sites (397m) suggested that females with broods were required to move 2 times their median daily movement in order to find suitable night-roost habitat and that seasonal movement estimates in the literature may be biased low. Annual home range sizes of yearlings and adults, and seasonal movements from the lek of capture indicated that the population of greater sage-grouse in the Axial Basin and Danforth Hills was non-migratory. Annual survival rates were greater for yearling (7 5%) than adult females (57%). Year, seasonal trend and age were variables that contributed to predictive annual survival models. Survival rates were lowest in the breeding and brood-rearing periods and increased through the fall and winter. Body condition indices, serum chemical constituents and disease prevalence were estimated for pre-laying female greater sage-grouse. Previous studies conjectured prematernal diet was associated with greater·sage-grouse productivity and survival but gave no specifics on condition with respect to body size or serum biochemical constituents. Thus I investigated pre-breeding body condition as a function of a body condition index and a serum chemical panel and how these parameters were associated with reproduction and survival. No differences were found in measures of fecundity, however the expense may be incurred through survival rates, which were related to the body condition index, serum protein and uric acid concentrations. Baseline biochemistry reference ranges for pre-breeding hens were established. These values varied by year and capture technique. Annual variation may be linked to precipitation and forage quality. Elevated creatinine kinase and chloride levels occurred in females captured by the CODA net-launcher in comparison to females captured by night-spotlighting. Differential biochemical values may have been caused by stress or diurnal variation in serum chemistry. Vl • Diseases tested included avian influenza, Salmonella pullorum/ S. typhoid, • Mycoplasma gallisepticum, M. synoviae, and M meleagridis. Individual grouse tested positive forM. synoviae, however there was no consistency of lab results among years, laboratories, or blood storage method. Additionally, no relationship between disease occurrence and reproduction or survival was detected and false positives have been reported with the plate agglutination-screening test. Given this information, the serology results need to be viewed cautiously as they were based on questionable disease incidence. ,~ t • Vll ACKNOWLEDGEMENTS This project would not have been possible without the financial and academic support of a number of agencies and individuals. First I would like to thank Colorado Division of Wildlife (CDOW), in particular Tony Apa and Rick Hoffman. Tony recognized the need for this work and provided all the funding and equipment required to run a project of this size. I am deeply grateful for his assistance in trapping the first season and in reviewing and editing my proposal and thesis. I must also extend the deepest gratitude to Rick Hoffman for his interest and genuine concern over field methods and sampling. He helped tremendously not only in guidance, but also in providing equipment, field support, vehicles, great passion and conversation. Additionally, Eric Rock was a tremendous help in locating leks, trapping, and providing energy and life-advice. I would like to thank Colowyo Coal Company for providing funding, housing and land access. I would particularly like to acknowledge Jarrell and Dottie who facilitated land Jf( . access and a comfortable field house. Katie Mills helped tremendously in data acquisition. Thank you also to National Fish and Wildlife Foundation, Trapper Mining Company, Moffat County, Yampa Valley Electric Company, North American Grouse Partnership, Bureau of Land Management, and Moffat Country Sage-Grouse Working Group for providing assistance with funding. Within BLM I would particularly like to acknowledge Mieke Bruch • for all her help with gathering data on land use. I sincerely thank all land owners for allowing this study to be conducted on their property, in particular, Kevin Schield, Nick and Ann Charchalis, Seeley Land and Livestock Company, Doug W elman, Walt Proctor, Walt Vlll and Nancy Bohrer, Carol Davidson Partnership Ltd., Dean Gent, Leonne Earl, Tom Kourlis, Ken Bekkedahl, Tom Deakon, Sterling and Mary Cook, Bill Moore and Ben Little. I would like to thank Kerry Reese for the boundless energy and time he has put into my academic and personal growth. He mustered enthusiasm in the field (even powered by soy cheese and humus sandwiches), assistance in creating and editing reports and continually challenging me. Most of all I would like to thank him for believing in me and helping me to believe in myself. I could not have asked for an advisor more suited to my needs and goals. I would also like to thank my committee members Dr. Patricia Talcott and Dr. Gerald Wright. Tricia attended my committee meeting days after having her first child, and provided invaluable guidance with the biochemistry and disease testing and with editing and reviewing reports. Thank you to Gerry for his insight and letting me snowshoe on his property! Data analysis would not have been possible without counseling from Dr. Kirk Steinhorst, Dr. E.O. Garton and advice from Paul Wik, Cam Collins, Robyn Vasterling, Jeff Manning and J enna Boisvert. I would like to thank Dr. Erik Stauber for teaching me how to draw blood on his pet doves. I also need to extend thanks to Dr. Mark Drew, Dr. Mike Dunbar, Dr. Singh Dhillon, Dr. Kendel Eyre and Michaela Erickson for guidance with disease testing, biochemistry panels and tissue handling in the field. Thank you also to Mountain Meat Packers for allowing me the use of their freezers and McCandless Animal Hospital for the silver nitrate sticks. Additionally, I must also acknowledge individuals for their physical and emotional support throughout this endeavor. I would like to thank all "grouse-house" residents, especially technicians Kyle Olree and Brooke Palmer for their year-round efforts in the field IX with continued energy and enthusiasm. Thank you for the seasonal help and incredible friendships from Heather Antonio, Zoe Miller, Sarah Quinlivan, Angie Lyman, and Matt Schenk. The study could not have been successful without their hard work and irregular hours. Thank you also to volunteers Mac Zimmerman, Andy Thompson, Amy Martin, Chuck Lowman, Bill Devergie, Jim Kiger, Chuck Woodward, Michele Harmon, Dean Gent, Tara Chenoweth, Mark and Tina Parchman, Mary Morris, Cathy DelGrosso, Annie Sagerson, Clee Seeling, Bob Nelson, and Julie Jessen who assisted in trapping and monitoring birds. Two volunteers in particular need an additional acknowledgement. I wish to thank Cam Collins for his tremendous trapping effort, field support and overall guidance throughout this project. Thanks also for the swing lesso~s and bike through Dinosaur! Additionally, I wish to express sincere thanks to Ellen Trimble, who braved the heat, shrubs, bugs and nights of northwest Colorado and provided us with renewed energy to finish off our vegetation sampling. I would like to thank my family for their continued words of encouragement, in particular, Anna Hausleitner for her efforts summing vegetation plots, keeping me well fed throughout this project, and many phone calls of support. It is impossible to extend enough thanks to Tom Abraham, who contributed to this project immensely. He gave up many nights and mornings and traveled great lengths so that we could become the dynamic trapping duo, running through the sagebrush-playing polka and Lucinda Williams: Additionally, he assisted in tic removal, vegetation sampling and monitoring on many summer days. He crossed the Axial Basin on foot on more than one occasion in support (defiance?) of the project. He also edited and provided valuable advice (and commas) on X reports and presentations. Most of all he provided emotional support and space when I most needed it, in order to finish this project and enable our relationship to evolve. Thank you to the Reese family group who helped make presentations smoother and provided literature, advice and friendship. Lastly I would like to thank all the graduate students who have made my stay at the University of Idaho a pleasure. I am not sure where to start in mentioning all the amazing friends and memorable occasions I have had here in Moscow. My life has been enriched with memories of sky-diving (oh ... er do I remember that?), hiking trips, running on Moscow Mountain, skiing, floor-hockey, ultimate (did I mention our championship title?), coop pizza, hair-dying episodes, theater, parties, dinners, potlucks, boggle, Kath's hot-tub and biking in White-pine. Thank you to all of you for the gift of your friendship. Xl TABLE OF CONTENTS ABSTRACT ........................ . ......................... ........................................... iii ACKNOWLEDGEMENTS ........................................................................... vii TABLE OF CONTENTS ............................................................................... xi LIST OF TABLES ...................................................................................... XV INTRODUCTION ...................................... .................................... . ... XV CHAPTER 1 .................................................................................... XV CHAPTER 2 .................... ............................................................... xvi CHAPTER 3 .................................................................................... xvi CHAPTER 4 ........................... . ....................................................... xvii CHAPTER 5 ....................................................................... . ........... xvii LIST OF FIGURES .................................................................................. xix CHAPTER 1 ................................................................................... xix CHAPTER 2 ............................................. ...................................... xix CHAPTER 3 .................................. : ................................................ xix CHAPTER 4 ...................................................................... .............. XX LIST OF APPENDICES ............................................................................. xxi INTRODUCTION ..................................................................................... 1 STUDY AREA ................................................... . ..................................... 3 GENERAL METHODS ............................................................................... 4 Lek surveys ................................. . ...................................................... 4 Capture...................................... . ...................................................... 4 Monitoring .......................................................................................... 5 GENERAL RESULTS ............................................................................ 5 Lek surveys .............. . ......................................................................... 5 Capture ............................................. . .. ............................................. 6 Bodymass ......................................................................................... 6 THESIS CHAPTERS ................................. . ............................................... 6 LITERATURE CITED ................................................................................ 8 CHAPTER 1. NEST SITE SELECTION AND SUCCESS OF GREATER SAGE-GROUSE HENS IN NOTHWEST COLORADO .......................................................................... 12 INTRODUCTION ................................................................................... 12 Reproduction ................................................................................... 12 Nesting habitat ................................................................................... 13 METHODS ............................................................................................. l4 Xll Monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . 14 Vegetation Srunpling ........................................................................... 14 Specific Methods ............................................................................... 16 Reproduction ........................................................................... 16 Vegetation Sampling .................................................................. 16 RESULTS .............................................................................................. 18 Reproductive Rates ............................................... ~ ............................ 18 Nest Predators ................................................................................... 19 Vegetation Srunpling ........................................................................... 20 Macrohabitat ........................................................................... 20 Microhabitat ............................................................................ 20 Individual Plant Frequency.......................................................... 23 DISCUSSION ......................................................................................... 23 Reproductive Rates ............................................................................ 23 Nest Predators .................................................................................. 26 Vegetation Sampling ........................................................................... 26 MANAGEMENT IMPLICATIONS .......................... : .................................... 29 LITERATURE CITED .............................................................................. 31 CHAPTER2. AN EVALUATION OF THE TIMING OF GREATER SAGE-GROUSE NEST VEGETATION SAMPLING .......................................................................... 52 INTRODUCTION .................................................................................... 52 METHODS ............................................................................................ 52 RESULTS .............................................................................................. 54 DISCUSSION ......................................................................................... 55 MANAGEMENT IMPLICATIONS ............................................................... 56 LITERATURE CITED ............................................................................... 58 CHAPTER3. NOCTURNAL AND DIURNAL HABITAT SELECTION OF GREATER SAGE-GROUSE DURING BROOD-REARING IN NORTHWEST COLORADO ......... 64 INTRODUCTION .................................................................................... 64 Brood-rearing Habitat ......................................................................... 64 Night-roost Habitat ............................................................................ 65 METHODS ............................................................................................. 65 Flock Size ........................................................................................ 67 Vegetation Sampling ........................................................................... 68 Early/late Brood-rearing ...................................................................... 69 RESULTS ............................ ·................................................................. 69 Brood size ......................................................................................... 69 Flock Size ........................................................................................ 70 Vegetation Sampling ........................................................................... 70 Macrohabitat ........................................................................... 70 Microhabitat ........................................................................... 72 XIV Biochemistry......................................................................... 136 Diseases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 7 Survival ......................................................................................... 137 Body Condition Index ............................................................... 137 Diseases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 9 RESULTS ............................................................................................. 139 Reproductive Parameters ............ . .. . ..................................................... 139 Body Condition Index ............................................................... 139 Biochemistry......................................................................... 139 Diseases ............................................................................... 140 Survival ..................................................................................... . ... 141 Body Condition ...................................................................... 141 Diseases ............................................................................... 142 DISCUSSION ........................................................................................ 142 Body Condition Index ............................................................... 142 Biochemistry......................................................................... 143 Diseases ............................................................................... 147 MANAGEMENT IMPLICATIONS ........................ . .................................... 149 LITERATURE CITED ............................................................................. 151 XV LIST OF TABLES INTRODUCTION Table 0.1. Peak female and male lek attendance (numbers and date) for greater sage-grouse in the Axiai'Basin and Danforth Hills, Colorado, 200112002 ................. 10 Table 0.2. Greater sage-grouse captures by year, sex and capture technique in northwestern Colorado, 2001 and 2002 ..................... ; .................................... 11 CHAPTER!. Table 1.1. Number of nests and female greater sage-grouse and their nest fate in northwest Colorado, 2001/2002 .................................................................... 37 Table 1.2. Cause-specific nest fate of radio-marked greater sage-grouse females in northwest Colorado, 2001/2002 ................................................................... 3 8 Table 1.3. Differences in macrohabitat characteristics (mean and standard error) between years at random sites in the Axial Basin and Danforth Hills, Colorado, 2001/2002 ..................... . ...................................................................... 39 Table 1.4. Means and standard errors of macrohabitat characteristics and results of chisquared tests for successful and non-successful greater sage-grouse nest sites and random sites in Axial Basin and Danforth Hills, Colorado, 2001/200 ....................... 40 Table 1.5. Microhabitat characteristics varying between locations at random sites (mean and standard error) in Moffat County, Colorado 2001/2002. ANOVA P-values are reported .............................................................................................. 41 Table 1.6. Microhabitat characteristics varying between years at random sites (mean and standard error) in northwest Colorado, 2001/2002. ANOVA P-values are reported ....... 42 Table 1. 7. Pattern of 5 Principal Components derived from analysis of 21 microhabitat variables from 75 random plots in Axial Basin and Danforth Hills, Colorado, 2001/2002 ........................................................................................... 43 Table 1.8. Logistic regression models of habitat variables predicting the binary variables for random and nest sites or successful and depredated nests of greater sage-grouse in northwest Colorado 2001/2002 .................................................................. 44 Table 1.9. Mean and standard error (SE) for habitat variables measured at successful, unsuccessful, and random nest sites of greater sage-grouse in Moffat County, Colorado 2001/2002 .................................................... . ......... .................. 45 XVl CHAPTER2. Table 2.1. Correlations between understory cover characteristics of greater-sage grouse nest habitat. Pearson correlation coefficients and associated P-values are reported ...... 61 Table 2.2. Mean and standard error (SE) for habitat variables included in MANOVA at initiation and hatch for 30 greater sage-grouse nests from 2001 measured in 2002 in the Axial Basin and Danforth Hills, Colorado ................................................... 62 Table 2.3. Coefficient of determination (R2) for habitat characteristics measured over a sampling gradient of initiation to hatch for 30 greater sage-grouse nest sites sampled in 2002 in northwest Colorado ..................................................................... 63 CHAPTER3. Table 3.1. Results of the multiple regression model using structural habitat variables to predict flock size of greater sage-grouse females and broods at night-roost locations during summer in Moffat County, Colorado, 2001/2002 ..................................... 94 Table 3.2. Results of chi-squared tests among physical macrohabitat characteristics (means and standard error) measured at greater sage-grouse brood, nocturnal and random sites in northwest Colorado, 2001/2002 ................................................ 95 Table 3.3. Microhabitat characteristics varying between years at random sites (mean and standard error) in Axial Basin and Danforth Hills, Colorado, 200112002 .................. 96 Table 3.4. Pattern of 4 principal components derived from analysis of 15 microhabitat variables from 94 random plots in Axial Basin and Danforth Hills, Colorado, 2001/2002 ............................................................................................ 97 Table 3.5. Logistic regression models of habitat variables predicting the binary variables for random and greater sage-grouse brood or night-roost sites and diurnal and nocturnal sites in northwest Colorado 2001/2002 .............................................. 98 Table 3.6. Mean and standard error (SE) for habitat variables measured at greater sagegrouse brood, night-roost and random sites in Axial Basin and Danforth Hills, Colorado, 2001/2002 ............................................. ................................... 99 Table 3.7. Microhabitat characteristics varying among early and late brood-rearing sites (mean and standard error) of greater sage-grouse in Axial Basin and Danforth Hills, Colorado, 2001. Brood-rearing periods were estimated as the time in which 90% of movements had been made to alternative cover types or long distances movements had ceased. In 2001 this date was 7 July ...................................................... 100 XVII • CHAPTER4. • .... Table 4.1. Median home-range estimates (95% fixed kernel, with least squares crossvalidation) and 95% confidence intervals around the median of greater sage-grouse brood-rearing females in northwest Colorado, 2001-2002 .................................. 124 Table 4.2. Median annual home-range estimates (95% fixed kernel, with least-squares cross validation) and 95% confidence intervals around the median of greater sage-grouse females in northwest Colorado, 2001/2002 and 2002/2003 .................. 125 Table 4.3. Annual and seasonal survival rates of radio-marked greater sage-grouse females in northwest Colorado 2001 and 2002. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . .. . . . .. 126 Table 4.4. Monthly survival estimates of adult greater sage-grouse females, 2001 and 2002, in northwest Colorado ..................................................................... 127 Table 4. 5. Monthly survival estimates of yearling greater sage-grouse females, 2001 and 2002, in northwest Colorado ..................................................................... 128 Table 4.6. Breeding, brood-rearing, fall and winter survival models varying by year, age and year + age for greater sage-grouse females in Moffat County, Colorado, 2001/2002. Models with 11 AICc::: 2 are considered competing models ............... 129 Table 4. 7. Annual survival models varying by year, age, season and time trend for greater sage-grouse females in Moffat County, Colorado, 200112002. Models with 11 QAICc S 2 are considered competing models .................................................. 130 CHAPTERS. Table 5 .1. Coefficient of determination (r2) for impact of spring body condition on reproductive parameters of greater sage-grouse females in Axial Basin and Danforth Hills, Colorado, 2001/2002 ...................................................................... 156 Table 5 .2. Blood biochemistry values for adult and yearling greater sage-grouse females, captured in March and April, Axial Basin and Danforth Hills, Colorado, 2001/2002 and for Oregon and Nevada captured 1999-2001 ............................................. 157 .. Table 5.3. Biochemistry values for 2 capture methods (mean and standard error) in prelaying greater sage-grouse females in northwest Colorado, 2001/2002 ................... 158 Table 5 .4. The influence of M synoviae on reproductive parameters (mean and standard error) of greater sage-grouse females in the Axial Basin and Danforth Hills, XVlll Colorado, 2001/2002 ........................................... ·................................... 159 Table 5.5. Logistic regression model of body condition and biochemical variables predicting the prevalence of M synoviae in pre-breeding greater sage-grouse females in northwest Colorado, 2001/2002 ............................................................... 160 Table 5.6. A priori survival models of greater sage-grouse females during the breeding and brood-rearing period tested for association with pre-breeding body condition in northwest Colorado, 200112002. Models with 11 QAICc::: 2 are considered competing models ................................................................................. 161 Table 5. 7. Breeding and brood-rearing survival models of greater sage-grouse females tested forM synoviae in northwest Colorado, 2001/2002. Models with 11 QAICc::: 2 are considered competing models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . ... . . . . . . . . . .. 162 • • XIX LIST OF FIGURES CHAPTER I. Figure 1.1. The 25 of the most common plant species at greater sage-grouse nest and random sites in Axial Basin and Danforth Hills, Colorado, 2001 and 2002. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A .......... 36 CHAPTER2. Figure 2.1. Regression of grass heights measured at 0 and 1m at greater sage-grouse nest sites at initiation (4/18-4/29) and nest hatch (5/17-6/4) in Moffat County, Colorado 2002 ................................................................................................... 59 CHAPTER3. Figure 3.1. The 25 most common plant species at greater sage-grouse brood use locations and random sites in Axial Basin and Danforth Hills, Colorado, 2001. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A .......... 90 Figure 3.2. The 25 most common plant species at greater sage-grouse night-roost locations and random sites in Axial Basin and Danforth Hills, Colorado, 2001. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A .......... 91 Figure 3.3. The 25 most common plant species at greater sage-grouse brood use locations and random sites in Axial Basin and Danforth Hills, Colorado, 2002. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A .......... 92 • Figure 3.4. The 25 most common plant species at greater sage-grouse night-roost locations and random sites in Axial Basin and Danforth Hills, Colorado, 2002. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A ......... 93 XX CHAPTER4. Figure 4.1. Mean monthly elevation and standard error of radio-marked greater sage-grouse females in the Axial Basin and Danforth Hills, Colorado in 2001 and 2002 ...................... . .... . ............................ ....... . ... ..... . ......................... 122 Figure 4.2. Annual survival model {S (T)}, time trend in seasonal survival of greater sage-grouse females in the Axial Basin and Danforth Hills, Colorado in 2001 and 2002. Survival rates and standard errors reported ............ . ............................... 123 .. . XXl LIST OF APPENDICES Appendix 1-A. Vegetation species and their native status identified at microhabitat plot sites of greater sage-grouse nests and random sites in the Axial Basin and Danforth Hills Colorado, 2001-2002. Code is the abbreviated genus and species name (first 2 letters of each). Nomenclature is taken from the PLANTS Database (USDA 2003) ......................................................................... ········· ............... 46 Appendix 1-B. Climatic variables from Craig 4 SW weather station in Colorado, 20012002 (NOAA 2003). MNTM= mean monthly temperature, TPCP= total monthly precipitation, and difference between 2002 and 2001 ................... . ..................... 51 1 INTRODUCTION Greater sage-grouse are the largest grouse in North America. The Latin name, Centrocercus urophasianus, is derived from the greek word "kentron" meaning spiny, "kerkos" meaning tail, and "oura phasianos" meaning tail of a pheasant (Gill 1966). Centrocercus is one of 10 genera in·the sub-family Tetraonidae, of the order Galliformes (fowl-like birds). They are members of the family Phasianidae to which the turkey (Meleagris spp. ), quail ( Callipepla spp. ), partridge (Alectoris spp.), pheasant (Phasianus spp.) and grouse belong (Madge and McGowan 2002). There are 2 species of sage-grouse in the genera Centrocercus (Young et al. 2000), the greater sage-grouse and Gunnison sage-grouse, Centrocercus minimus. Two weakly described subspecies include Eastern (C. u. urophasianus) and Western (C. u. phaois) greater sage-grouse (Hupp and Braun 1991, Schroeder et al. 1999). Sage-grouse depend on sagebrush (Artemisia spp.) for food or cover at all times of the year. Their distribution is associated with habitat dominated by sagebrush in western North America (Schroeder et al. 1999). Greater sage-grouse have been reduced to populations in southeastern Oregon, northeast and east border of California, southern Idaho, northern two-thirds ofNevada, portions of northeast, north and south Utah, portions of western Colorado, Wyoming (except northwest and southeast comers), east and southwest Montana, northwest and southwest S. Dakota, North Dakota, southeastern Alberta, • southwestern Saskatchewan, and small populations in central Washington (Schroeder et al. 1999). Gunnison sage-grouse are found in 8 isolated populations in Southwestern Colorado and adjacent San Juan County, Utah (Young et al. 2000). 2 Greater sage-grouse have experienced continued declines from the early 1900's to present with relief in the 1940's and 1950's (Homady 1916, Patterson 1952, Autenrieth 1981, Connelly et al. 2000). It is estimated that breeding populations have declined by 17-47% in the last 30 years (Connelly and Braun 1997). All populations of greater sage-grouse are under review for listing under the Endangered Species Act (Connelly et al. 2000). In Canada, greater sage-grouse have been listed provincially as endangered or threatened (Aldridge 2000). Within Colorado, greater sage-grouse historically occurred in at least 23 and possibly 27 counties (Braun 1995). Northern Colorado held and holds the highest density of greater sage-grouse and sagebrush habitat (Braun 1995). Currently, there are 15 counties in which greater sage-grouse occur in Colorado and populations are considered as persistent (> 500 breeding birds) in only 5 of these (Braun 1995). Braun (1998) noted lek count declines of 45 and 82% in Colorado since 1980. Lek surveys indicating population trends have been conducted in Moffat County from 1958-1997 and on Colowyo mine property in the Axial Basin in 1981, 1983 and 1994 to present (Hoffman 1979, Braun 1998, Monarch 2000). Mean number of males per lek has declined in Moffat County since the late 1960's. Braun (1998) reported an 82% decrease in the number of males on strutting grounds from 1978-1998 and a 52o/o decline in the number of active leks. Other than declining population trends, little is known of the Axial Basin and Danforth Hills population of greater sage-grouse. The objectives of my research were to (1) estimate reproductive parameters and nesting and brood-rearing habitat use; (2) characterize night-roost habitat selection; (3) assess current nest habitat sampling techniques; (4) ascertain 3 annual and seasonal movements, home-range sizes, cause-specific mortality and survival; and (5) assess pre-breeding hen body condition and disease prevalence and their possible influences on productivity and survival. STUDY AREA The central part of my study area was the Axial Basin and Danforth Hills, located in Moffat County, 30 km southwest of Craig, Colorado. The topography consisted of rolling hills and the study area was approximately 200,000 ha in area. The Axial Basin and Danforth Hills comprised approximately 36,000 ha and 52,000 ha of the study area, respectively. Elevation in the Axial Basin and Danforth Hills ranged from 1,818-2,000 m and 2,001 - 2,388 m, respectively. The Yampa River flows from east to west through the ·Axial Basin. The climate of Moffat County is .semi-arid, receiving 20.3 to 50.8 em of annual precipitation. The mean arinual temperature was 6.3 C (Braun and Hoffman 1979). The landscape consisted of private, state, and federally owned/administered sagebrush rangeland (Artemisia tridentata ), pinyon (Pinus edulis )-juniper (Juniperus monosperma, J. osteosperma, J. scopulorum ), mountain shrub communities with Gambel' s oak (Quercus gambelii), serviceberry (Amelanchier sp.), snowberry (Symphoricarpos oreophilus) chokecherry (Prunus virginiana ), and aspen (Populus tremuloides), private agricultural land in the Conservation Reserve Program (CRP), active strip mines, reclaimed strip mine lands, and agricultural lands. The sagebrush community was dominated by an overstory of big sagebrush (A. tridentata) and snowberry. The predominant grasses within this community included western wheatgrass (Pascopyron smithii), Kentucky bluegrass (Poa pratensis), Sandberg bluegrass (Poa secunda), cheat grass brome (Bromus tectorum ), and needle and 4 thread grass (Hespero-stipa comata). Dominant forbs included lupine (Lupinus sericeus), wild onion (Allium spp.), arrowleafbalsamroot (Balsamorhiza sagittata), and yarrow (Achillea millefolium). GENERAL METHODS Lek Surveys Active leks were surveyed a minimum of 3 times from 17 March to 8 May. Colorado Division of Wildlife protocol (2001) was used to survey 16leks in the Axial Basin and Danforth Hills. Capture Grouse were captured on or near 11lek sites during April and May 2001/2002. Three methods of trapping were employed, night-tinie spotlighting by vehicle and backpack, walkintraps, and a CODA net-launcher. Individuals were sexed and aged as yearlings or adults using wing characteristics (Dalke et al. 1963). They were banded and radio-tagged with necklace-mounted transmitters (Leonard et al. 2000) equipped with 4-hour mortality sensors and a range of 1.5 km. Guaranteed battery life was 18 months and a 30-cm antenna lay between the wings and down the back of the bird. A hand-held Y agi antenna, attached to a receiver/scanner was used to locate the radio-tagged grouse. Morphological characteristics of grouse measured at time of capture included weight and keel length. Blood was taken from the cutaneous ulnar vein for disease and biochemical analysis. 5 )- Monitoring Females were monitored until nesting was evident. Nest fate was determined when monitoring indicated that nesting efforts had ceased. Nests were categorized as successful(~ 1 egg hatched) or unsuccessful (Wallestad and Pyrah 1974). During the spring and summer, females with broods were located 1-2 times a week. Locations were collected for each brood from the time of hatch to 10 weeks because at this age chicks are considered independent from the hen (N. Burkepile, personal communication). Hens without broods were located weekly. During the fall and winter, locations were obtained on all birds 2:: 1 time per month. Grouse status (alive/dead) was monitored weekly. All use locations were recorded in Universal Transverse Mercator (UTM) coordinates using a hand-held Global Positioning System (GPS). Female body mass was analyzed using a 2-way factorial ANOVA (Ott and Longnecker 2001) to determine whether body mass varied with age or year. GENERAL RESULTS Lek Surveys We attempted to record peak attendance days in 2091 and 2002 (Table 0.1). Peak female attendance occurred in the first week of April and was similar among years. Two potential leks, Tic Tac and East Box Elder, found in 2001 were inactive in 2002. Additionally, historical Temple Gulch lek was found inactive when surveyed in 2002. Historicallek sites SG- 1- 3, 5 and 6, were devoid of males both.years. Dramatic declines in male attendance from 2001 to 2002 were observed at West Box Elder Lek. Domestic sheep occupied this lek from 28 March to 10 May in 2002. 6 Capture In March and April2001 and 2002 I captured 144 grouse. One hundred sixteen grouse (81 %) were captured using the spot-light capture method, 26 (18%) were captured using the CODA net launcher and 2 were captured using walk-in-traps (Table 0.2). One hundred females and 15 males were radio-marked and 6 were re-captures. The remaining grouse were captured in order to validate our blood testing procedures. Of the radio-marked females, 54 were adults and 46 were yearlings. Body mass Mean body mass of yearling females was 1,424 g (SE = 19, n = 27) in 2001, and 1,426 g (SE = 15, n = 19) in 2002. Body mass of adult greater sage-grouse females was 1,625 g (SE = 20, n = 35) and 1,564 g (SE = 24, n = 22) for 2001 and 2002, respectively. Weights did not vary between years (F = 3.12, P = 0.0805). However, yearling female weight (1,425 g, SE = 13, n = 46) was less (F = 72.01, P < 0.001) than that of adult females (1,601 g, SE = 16, n =57). THESIS CHAPTERS Chapters in my thesis are written separately, with the intent of individual publication of each. Chapter 1 describes reproductive parameters and models nest site selection and success of greater sage-grouse in northwest Colorado. Additionally, this chapter includes a description of cause-specific nest depredation. Chapter 2 is a techniques chapter that evaluates the timing of nest vegetation sampling. Chapter 3 describes brood productivity and diurnal and nocturnal habitat selection of brood females. Chapter 4 reports annual and seasonal movements, cause-specific mortality and survival of greater sage-grouse females. 7 Chapter 5 describes the fitness of greater sage-grouse females associated with disease and the pre-breeding body condition of females . ... 8 LITERATURE CITED Aldridge, C.L. 2000. Reproduction and habitat use by sage grouse (Centrocercus urophasianus) in a northern fringe population. M.S. Thesis University of Regina, Regina, Saskatchewan. 109pp. Autenrieth, R.E. 1981. Sage grouse management in Idaho. Wildlife Bulletin 9. Idaho Department ofFish and Game, Boise, Idaho. Braun, C.E. 1995. Distribution and status of sage grouse in Colorado. Prairie Naturalist 27: 1-9. Braun, C.E. 1998. Sage grouse declines in western North America: what are the problems? Proceedings of the Western Association of State Fish and Wildlife Agencies 78:139-156. Braun, C.E., and D.M. Hoffman. 1979. Vulnerability and population characteristics of sage grouse in Moffat County. Job Progress report. W-37-R-32, Work Plan 3, Job. 11. Colorado Division ofWildlife, Fort Collins 163-199. • Connelly, J.W., and C.E. Braun. 1997. Long-term changes in sage grouse Centrocercus urophasianus populations in western North America. Wildlife Biology 3:229-234. Connelly, J.W., M.A. Schroeder, A.R. Sands, and C.E. Braun. 2000. Guidelines to manage sage grouse populations and their habitats. Wildlife Society Bulletin 28:967-985. Dalke, P.D., D.B. Pyrah, D.C.. Stanton, J.E. Crawford, and E.F. Schlatterer. 1963. Ecology, productivity, and management of sage grouse in Idaho. Journal of Wildlife Management 27:811-841. Gill, R.B. 1966. A literature review on the sage grouse. Department of Fish and Parks Game Research Division and Cooperative Wildlife Research Unit, Final Report 6: 1-38. Hoffman, D .M. 1979. Investigations of the distribution and status of sagebrush and sage grouse in the Moffat County area. Colorado Game and Fish. ProjectW-37-R-32. Work Plan 3. Job 10. Job Completion report. 95-161. Hornady, W.T. 1916. Save the sage grouse from extinction, a demand from the civilization to the western states. New York Zoological Park Bulletin 5:179-219. Hupp, J.W., and C.E. Braun. 1991. Geographic variation among sage grouse in Colorado. Wilson Bulletin. 103: 255-261. Leonard, K.M., K.P. Reese, and J.W. Connelly. 2000. Distribution, movements and habitats of sage grouse Centrocercus urophasianus on the Upper Snake River Plain of Idaho: changes from the 1950s to the 1990s. Wildlife Biology 6:281-286. •~ t l 9 Madge, S., and P. McGowan. 2002. Pheasants, partridges and grouse: A guide to the pheasants, partridges, quails, grouse, guineafowl, buttonquails, and sandgrouse of the world. Princeton University Press, Princeton, New Jersey. 488 pp. Monarch, J. 2000. 1999 Grouse studies on Colowyo Coal Company L.P. Properties in the Danforth hills. Unpublished report, Cedaredge, CO. 31 pp. Ott, R.L., and M. Longnecker. 2001. An introduction to statistical methods and data analysis, fifth edition. Duxbury, Pacific Grove, CA. 1152pp. Patterson, R.L. 1952. The sage grouse in Wyoming. Sage Books, Inc., Denver, CO. 341pp. Schroeder, M.A., J.R. Young, and C.E. Braun. 1999. Sage grouse (Centrocercus urophasianus) In A. Poole and F. Gills (eds). The Birds ofNorth America, No. 425. The Birds of North America, Inc., Philadelphia, P A. Wallestad, R.O., and D. Pyrah. 1974. Movement and nesting of sage grouse hens in central Montana. Journal of Wildlife Management 35:129-136. I ~ L I Young, J.R., C.E. Braun, S.J. Olyer-McCance, J.W. Hupp, and T. W. Quinn. 2000. A new species of sage-grouse (Phasianidae: Centrocercus) from southwestern Colorado. Wilson Bulletin 112: 445-453. l I I I 10 ,. )- r. ~ ... • I ~ .. .. I L- Table 0.1. -Peak female and male lek attendance (numbers and date) for greater sage-grouse in the Axial Basin and Danforth Hills, Colorado, 2001/2002. 2001 2002 Males Females East Box Elder a 8 (4/1) 0 0 0 Gossard/ SG- 12 8 (4/5) 6 (4/5) 8 (3/26) 15 (3/26) 0 1 (4/1) 0 . 4 (3/21) Juniper 2/ SG-13 16 (3/20) 9 (4/4) 18 (4/11) 9 (4/7) Morgan Gulch 2/ SG-1 0 30 (4112) 23 (4/5) 23 (4/5) 58 (4/5) New Juniper Gulch a SG 5c 5 (3/28) 3 (4/3) 0 0 0 0 0 0 SG6c 0 0 0 0 SG-112 c 0 0 0 0 SG-3 c 0 3 (4/6) 0 1 (4/22) SG-4 0 0 9 (4/11) 9 (4/11) SG-7 SG-8 b 27 (4/8) 21 (4/3) 37 (5/4) 16 (4/9) 4 (4/15) 3 (4/15) 1 (4/2) 0 Temple Gulch (SG-14) 7 (3/30) 0 0 0 Tic Tac a 4 (4114) 0 0 0 West Box Elder/ SG-11 33 (3/25) 52 (4/1) 15 (3/25) 38 (4/3) LekNames Juniper 1 c Males Females I ~ I ~ ~ a Potentiallek 2001 UTM coordinates different from Monarch (2000) c Historicallek site b 11 Table 0.2. Greater sage-grouse captures by year, sex and capture technique in northwestern Colorado, 2001 and 2002. 2002 2001 Spotlighting CODA net launcher Walk-intrap a Spotlighting CODA net launcher Females 52 12 2 31 11 Males 12 3 0 21 0 TOTALS 64 15 2 52 11 a Capture technique attempted only in 2001 12 CHAPTER 1. NEST SITE SELECTION AND SUCCESS OF GREATER SAGEGROUSE IN NORTHWEST COLORADO INTRODUCTION There are several factors responsible for the decline in greater sage-grouse populations. The most important are habitat loss, fragmentation and degradation (Schneegas • ... 1967, Klebenow 1972, Braun 1998, Leonard et al. 2000). Degradation of sagebrush has been ongoing since early settlement across western rangelands. Recent landscape fragmentation has favored generalist predators, thus predation pressure may have increased as a function of habitat changes (Storch and Willebrand 1991). For many grouse species, nest depredation is an important source of reduced productivity (Reynolds et al. 1988). Long-term declines in greater sage-grouse abundance may be associated with low productivity due to habitat degradation. Thus, I investigated the effects of vegetation characteristics on nest success and created models of habitat characteristics to predict nest selection and success. Reproduction Incubation of greater sage-grouse lasts 25 to 29 days (Schroeder 1997). Mean clutch size varies from 6.0 to 9.5 throughout the range of the species (Connelly et al. 2000) and adults tend to lay larger clutches and are more successful than yearlings (77% vs. 44%) (Wallestad and Pyrah 1974). In Oregon, peak hatching occurs from the last week of May to the second week in June (Call and Maser 1985). Braun (1981) reported peak hatching dates in southern Moffat County, Colorado to occur from 15-21 June. 13 Nest success reported in the literature ranged from 15% (Gregg et al. 1994) to 77% (Wallestad and Pyrah 1974) and varied with local vegetation and weather conditions (Gill 1966). Braun (1981) and Swenson (1986) stated that older females have greater nest success and less variance than yearling females. Bergerud (1988) reported an average nest success of 38o/o and indicated that juvenile females renested less frequently than adult females. Contrastingly, Connelly et al. (1993) found nest success (53%) and renesting rates between age classes to be similar. Nesting habitat Spatial variation and diversity of habitat available to a prey population are likely to enhance the stability of that population (Reynolds et al. 1988). Specifically, increased spatial understory heterogeneity and cover at nest sites reduces the risk of nest detection by predators, and protects the eggs from environmental conditions (Bowman and Harris 1980, Cambell et al. 2002). The basic requirement of a greater sage-grouse nest site is concealment (Patterson 1952). Nests are almost exclusively under sagebrush (Patterson 1952, Klebenow 1969, Braun et al. 1977, Wakkinen 1990, Connelly et al. 1991, Musil et al. 1994) and are • created by scratching out a depression (Schroeder et al. 1999). Nests are infrequently constructed under other shrubs, and meet with lower success (Connelly et al. 1991). Nests are generally under shrubs with more ground, lateral, and shrub cover than at random sites (Klebenow 1969). Successful nests tend to be in areas of higher canopy cover (2: 15%) than unsuccessful nests (Klebenow 1969, Wallestad and Pyrah 1974, Gregg et al. 1994). In North Park, Colorado, nest shrub cover averaged 44% (n = 19) (Schoenberg 1982). Bowman and Harris (1980) indicated canopy cover strongly influenced prey vulnerability. Medium height shrubs (30-70 em) have been deemed important for nest success (Wallestad 14 and Pyrah 1974, Schoenberg 1982, Gregg et al. 1994, Delong et al. 1995, Connelly et al. • 2000). Klebenow (1969), Wakkinen (1990), Gregg et al. (1994) and Delong et al. (1995) indicated tall grass cover(~ 18 em) as a parameter important to nest success. METHODS Monitoring Radio-marked females were monitored until nesting was evident. Nest fate was determined when monitoring indicated that nesting efforts had ceased. Nests were • • categorized as successful (2: 1 egg hatched) or unsuccessful (Wallestad and Pyrah 1974) . Nest predators were classified as either avian or mammalian based on patterns of egg and nest site destruction (Sargeant et al. 1998). If eggs were not present or evidence was confounding, the nest predator type was classified as unknown (Lariviere 1999). Nest sites were recorded in Universal Transverse Mercator (UTM) coordinates using a hand-held Global Positioning System (GPS). Vegetation Sampling • ... Microhabitat refers to fine-scale habitat resolution within a patch of habitat at a nest site. These measurements were made :S 7 days after nesting efforts had ceased for each successful and unsuccessful nest and at an equal number of random sites. Random locations were selected from northing and easting UTM coordinates falling within the home range of the population and were measured within 1 week of vegetation sampling at the nest site use area using identical techniques. Random sites were selected from within a sagebrush community within the study site so that I could estimate the macro and micro-habitat variables selected for nesting. Ten meter transects were placed in the cardinal directions 15 intersecting at the center of the nest bowl. The nest shrub species and height were recorded. Canopy cover of the shrub overstory was estimated using line-intercept (Canfield 1941). Height of the nearest shrub within 1 m of the transect line was measured at 2.5 m, 5 m, 7.5 m and 10m. Grass height (GHEIGHT) was measured for the nearest, tallest grass part at the points where the edge of the nest bowl and the transect intercept, and at the 1 m point on each transect. Percent forbs and grass cover by lowest possible taxa, percent exotics, bare ground • • • and litter canopy cover were estimated using 40 by 50 em microplots (adapted from Daubenmire 1959, Mosley et al. 1989). Eleven cover classes were used as suggested by Boisvert (2002). Cover classes were delineated as: (Trace 1: 0-2%, Trace 2: 3-9%, 1: 1019%,2: 20-29%,3:30-39%,4: 40-.49o/o, 5: 50-59%, 6: .60-69%,7:70-79%, 8: 80-89%,9: 90100%). Two microplots were placed at the interception point of the transects, with 1 comer on the edge of the nest bowl going north or south along the transect. Subsequent plots were placed with the outermost comer along the transects at 2.5, 5 and 10m. Overhead concealment of the nest was measured using a 12 x 12 em cover-board .. separated into 25, 3 x 3 em squares (modified from Jones 1968). The cover-board was placed over the nest bowl and from a height of 1.5 m, and the number of squares covering> 50% were recorded. Visual obstruction (VISOB) at nest sites was measured using a 1.5 m cover pole demarcated by decimeter bands (Robel et al. 1970). The pole was placed at the nest bowl and the number of increments covered were recorded along each transect from 2.5, 5, and 10m and at a height of 0.5 and 1.5 m to simulate the view of potential nest predators. Measurements read at 5.0 m were correlated to those measured at 2.5 m and thus eliminated from further analysis (r2 = 0.67 - 0.84, P < 0.001 ). 16 I recorded physical macrohabitat characteristics at all sites. These were slope, aspect, elevation, and cover type. Cover type, as delineated by Colorado Gap was shrub steppe (sage brush), mountain shrub, CRP, riparian, grassland or agricultural field (Colorado Gap Analysis Project 2001). Agricultural fields consisted of either wheat or hay fields. Species richness was estimated by counting the number of forb and grass species present in the microplots. Additionally, distance to nearest anthill, visible roadway, telephone pole, power line, and fence post was estimated using a range finder. There were 156 plant species identified at nest and random sites over the 2 years of study (Appendix 1-A). Thus, analyzing cover by individual plant species was not feasible compared to the number of plots sampled. Instead, I summarized individual plant species by their frequency of occurrence within· each plot and ranked them by their abundance within • nest and random locations. Specific Methods Reproduction. Nest initiation was compared between years and between ages of hens using a chisquared goodness of fit test. Nest success was compared between years, ages ofhens and location using a chi-squared goodness of fit test (Ott and Longnecker 2001). Clutch initiation date and clutch size between ages and years were compared using ANOVA (Ott and Longnecker 2001). Vegetation sampling. I partitioned the variables of slope, aspect, elevation, distance to nearest structure, distance to nearest 2-track and distance to nearest anthill into categories. The categories for slope were partitioned into 3 categories (0-5°, 6-10°, > 10°). Aspect was partitioned into 5 17 categories: 0° (no aspect), 315-45° (northerly), 46-135° (easterly), 136-225° (southerly) and 226-314° (westerly). Elevation was characterized as low(< 2,020 m), medium (2,021-2,140 m) and high(> 2,141 m). Distances to nearest visible structure and nearest 2-track were categorized as 0-100 m, 101-200 m, and> 200m. Distance to nearest anthill was partitioned into 3 classifications (0-25 m, 26-50 m, > 50 m). Prior to examination, variables were assessed for correlation strength (Ott and Longnecker 2001). I used chi-squared tests to determine whether there were differences in macrohabitat between years and for successful and unsuccessful nests. Vegetation variables were divided into 2 groups, those measured at the nest bowl and those measured in the surrounding area (2.5, 5, and 10m). Grass height measurements were summarized at 0 and 1 m. Means of cover estimates were obtained by using the mid-point value of each cover class. Variables were assessed for normality and transformed when necessary. I used a multivariate analysis of variance (MANOVA) to assess microhabitat variables at random locations between years, locations (Axial Basin or Danforth Bills) and the year*location interaction. Canonical variables were used to assess for multicolinearity. Microhabitat variables that varied between years were assessed for correlation strength and then entered in a forward step-wise logistic regression with the principle components and the class variable year. A principal components analysis (PCA) was conducted on the random locations to reduce the amount of variables by combining highly correlated variables (Johnson and Wichern 2002). Nest sites were scored with the principal components generated from the pooled random plots. Forward step-wise logistic regression was used to model the importance of habitat variables in predicting each of2 binary variables; nest or random and successful or non-successful nests (Beier 1989). A significance level of P ~ 0.10 18 was used to determine which variables entered and remained in the model. This analysis was performed in SAS 8.2 (SAS Institute 1999). RESULTS Reproductive Rates Nest initiation rates for 2001 (86%, n = 50) were similar (i = 42, P = 0.518) to those reported in 2002 (90%, n = 69), therefore years were pooled to compare between ages. The nest initiation rate for adults and yearlings was 92% (n = 85) and 79% (n = 34), respectively • ()( = 3.57, P = 0.059). Nest success for 2001 and 2002 was 64%, n = 42 and 49%, n = 65, respectively (i = 2.34, P = 0.130) (Table 1.1 ). Years were pooled for comparison between ages and -locations. Nest success for adults (58%, n = 81) did not differ (i = 1.12, P = 0.292) from that of yearlings (46%, n = 26). Nest success for individuals in the Axial Basin (54o/o, n =53) was not different (i = 0.53, P = 0.473) than that of individuals in the Danforth Hills (61 %, n = 41). In 2001, only 1 female of 12 who lost their clutch attempted to renest and was successful. In 2002, 3 hens of20 attempted to renest, 2 were successful. Hen success over both years was 57% (n = 103). Mean clutch initiation date for 2001 was 26 April (SE = 1.76, n = 25). Mean clutch initiation date for 2002 was 21 April (SE = 0.91, n =51), 5 days earlier (F = 4.74, P = 0.033) than that of2001. Mean clutch initiation date did not differ (F= 2.08, P = 0.154) for adult (22 April, n =58) and yearling females (25 April, n = 19). Average clutch size for 2001 (n = 39) and 2002 (n = 42) was equal at 6.67, SE= 0.25 (F = 0.49, P = 0.484). Mean clutch size for adult females (7.04, SE = 0.20, n =59) was larger (F= 13.81, P < 0.001) than that of yearlings (5.68, SE = 0.29, n = 22). 19 Nest Predators 1>- Of the depredated nests (n = 40) over both years, 70% (n = 28) were attributed to mammalian predators, 5% (n = 2) avian, and 25% (n = 10) unknown (Table 1.2). One female abandoned her nest in 2001, while in 2002, 4 females abandoned nests. The nests depredated by avian suspects exhibited no nest material displacement and shells remained in the nest or were dispersed vertically. Eggs exhibited small openings in the side and had conspicuous yolk. Of the suspected mammalian depredations, none showed aerial dispersal and 91% showed ground dispersal of nest material. Of those, 36% exhibited 26-50% of nest material displacement and 23% exhibited 51-7 5% displacement. Thirteen percent of suspected mammalian predators cached 1-3 eggs outside of the nest under > 3 em of soil. Seventyseven percent of suspected mammalian depredations showed some digging at the nest site. ·The mean number of dig sites was 1.8, and 53% were 11-20 em, 24% were> 20 em, 18% were 6-10 em, and 6% were 1-5 em in width. There was> 1 shell found at 86% of mammalian-suspected nest depredations. Of these, 18o/o had small holes so that > % of the shell was intact and 60% had large holes so that 12 to :Y4 of the shell was still intact. The remaining eggs had < Y2 eggshell intact. The location of these openings in eggshells were sometimes mixed within a nest. Of 73 eggs, 73% of openings were on the side, 19o/o were on the side-end and 8% were on the end. Seventy-five percent of eggs were clean and 25% had conspicuous yolk remaining in the shell (n = 89). Hair samples were found at 3 nest depredation sites and 10 females were killed while incubating. Of the nests abandoned, 2 were due to researcher monitoring and 1 female abandoned immediately following the passage of a flock of sheep accompanied by sheep dogs. The reason for abandonment of the remaining nests is unknown. 20 Vegetation Sampling Macrohabitat. Of the 94 known nest sites, 90% (n = 85) occurred in a sage-steppe community, 5% (n = 5) in mountain shrub, 3% (n = 3) in CRP and 1% (n = 1) in a wheat field. No differences were detected among relative physical macrohabitat characteristics between years at random sites (P > 0.128) and so data were pooled (Table 1.3). There were no differences between proportion of use of physical macrohabitat characteristics measured at successful and non-successful nests (Table 1.4). There were no differences between relative use at random and nest sites except that nest sites tended to fall within the closest category (< 25 m) to the nearest anthill at a greater proportion (X= 8.44, 3 df, P = 0.038, n = 159) than at random sites (32 vs. 14%) (Table 1.4). Microhabitat. A difference was found between microhabitat characteristics at random sites for location* year (Wilk's 'A= 0.49; F = 1.89; 25,45 df; P = 0.031), location (Wilk's 'A= 0.46: F = 2.11, 25,45 df; P = 0.014) and years (Wilk's 'A= 0.30: F= 4.12, 25,45 df; P < 0.001). Assessment of the canonical coefficients revealed multicolinearity, thus the structural coefficients were used to assess which variables contributed to overall differences. The interaction, location*year was ordinal, and therefore I was able to interpret the main effects (Johnson and Wichern 2002). The habitat variables contributing to this difference included percent shrub cover (F= 6.9, P = 0.011), forb cover at the intersection (F= 12.27, P = 0.001), grass cover at the intersection (F = 7.10, P = 0.100), average grass cover (F = 4.76, P = 0.033) and visual obstruction between 0.5-1.0 m read from 0.5 m height and a distance of 21 10m (F = 4.56, P = 0.042). The herbaceous components contributing to this difference were accentuated in the Axial Basin due to the drought condition in 2002. The habitat variables that differed from the Axial Basin and Danforth Hills at the intersection were shrub height, grass height, percent litter and percent bare ground. Additionally, percent sagebrush cover, percent forb cover, percent bare ground and percent litter averaged over the plot, species richness and 6 visual obstruction differed between locations (Table 1.5). The habitat characteristics that varied between years at random locations were species richness, cover board reading, grass height at the intersection, grass height at 1 m, and VISOB > 50cm from 10.0 m and a reading height of 0.5 m (Table 1.6). None of these variables were highly correlated (l :S 0.34). A PCA of75 random sites minus the above 5 variables produced a description of the structural habitat gradients within the study area. With 21 structural variables, the PCA defined 5 principal components, accounting for 78% of the overall variance. Each of the principal components considered had an Eigenvalue> 1 (Table 1.7). The 5 principal components retained for further analysis each described an ecologically interpretable gradient. The first component was defined by cover at the intersection and was accompanied by shrub height at 0 m and visual obstruction readings. The second component varied along a gradient of ground cover with litter and exotics on the negative end of the scale and bare ground on the positive end. The third principal component was defined by a gradient of grass cover with grass and bare ground cover on the positive end and litter on the negative end. The fourth component described a complex gradient of structure with the scale ranging from herbaceous cover to tall shrubs, a high proportion of shrub cover, sagebrush, and a high percent of exotics. This addressed the overall 22 physiognomy of the sage-steppe community. Principal component 5 was characterized by the understory component and involved the forb and grass components of the community (Table 1. 7). Logistic regression with forward step-wise entry found PC 1, PC 5, grass height at 0 m and VISOBa23 to be significant predictors of nest occurrence (Table 1.8). Variables of importance within PC 1 included shrub cover, shrub height at 0 m, and visual obstruction readings (Table 1.7). Grass and forb cover were the main components of PC 5. The logistic regression equation was Logit (Y) = 1.0 + 0.9 (PC 1) - 0.6 (PC 5)- 0.08 (GHEIGHTO)- 0.7 (VISOBa23) + e. Nest shrub height and all visual obstruction readings were higher at nest sites than at random (Table 1.9). Mean percent shrub cover was 34% at nest sites and 24o/o at random sites. Forb cover was greater at nest sites than random and the alternative was found for grass cover. Grass height at 0 m was 13.8 em at nest sites and 15.5 em at random sites (Table 1.9). Logistic regression was also used to predict nest success. The habitat variables that entered the model as significant predictors of nest success were PC 5 and grass height measured at 0 and 1 m (Table 1.8). The logistic regression equation was Logit (Y) = -1.5 + 0.5 (PC 5) + 0.1 (GHEIGHTO) + 0.1 (GHEIGHT1) +e. Variables of importance within PC 5 were grass and forb cover (Table 1.7). Forb cover was greater at the nest bowl and within the plot site for successful nests (Table 1.9). Average grass cover was greater over the plot for successful nests. Mean grass heights measured at 0 and 1 m were 11.7 em and 13.5 em for unsuccessful nests and 15.4 em and 18.2 em for successful nests. 23 Individual Plant Frequency. Because the MANOVA indicated that there were no differences among percent forbs or grasses between years, I pooled plant frequency data. Of 94 nest sites, only 2 occurred under a shrub species other than sagebrush. One female nested under a rubber rabbitbrush (Ericameria nauseosa ), and 1 nested under a Russian thistle (Sa/sola iberica). Both were successful. Shrub species that occurred at > 60 % of nest and random sites were sagebrush and dead sagebrush (Figure 1.1). Forbs found in> 60% of plots included clover (Trifolium spp.), pale madwort (Alyssum alyssoides), longleaf phlox (Phlox longifolia), maiden blue eyed Mary ( Collinsia parviflora), false dandelion (Agoseris glauca) and lupine (Lupin us sericeus). Grasses that occurred in> 60% of nest sites included western wheatgrass · (Pascopyron smithii) and cheat grass (Bromus tectorum ). Greater species occurrence at nest sites may indicate selection for these species. Females may be selecting false dandelion, lupine, and salt and pepper parsley (Lomatium orientale) and avoiding copper mallow (Sphaeralcea coccinea) (Figure 1.1 ). DISCUSSION Reproductive Rates Nest initiation rates for adult and yearling females were 92% and 79%, respectively. Although this was not statistically significant it verifies previous reports which indicate adults have a higher nest initiation rate than yearlings (Bergerud 1988, Connelly et al. 1993). While Connelly et al. (1993) reported initiation rates of78% for adults and 55% for yearlings in southeast Idaho, my results are similar to those reported in Washington by Sveum (1995) of80% (n = 95) and Schroeder (1997) who observed 100% (n = 129). I believe my nest 24 initiation rates may be biased low because insufficient monitoring could potentially have missed some nesting efforts. Clutch initiation dates ranged from 8 April to 20 May. Renesting efforts ranged from 15 May to 25 May. These are earlier than nesting dates reported for North Park, Colorado, which were between 21 April and 21 June (Dargon and Keller 1940). Braun (1981) reported peak hatching dates in southern Moffat County from 15-21 June. Peak hatch dates in the Axial Basin and Danforth Hills occurred from 17-26 May. Peak hatch dates were earlier than reported by Braun (1981) indicating that I did not miss many initial nesting efforts. Clutch initiation dates were earlier in 2002 likely due to drier conditions and a milder winter. Regional climate maps for northwest Colorado showed that there was a 21.3 em decrease in precipitation from 2001 to 2002 for April to June and a+ 3°C departure of average temperature for June 2002 (Appendix 1-B). Additionally, December - February 2002 received 25-50% of normal precipitation (NOAA 2003). Mean clutch size in the Axial Basin (6.67) was smaller than previous studies in Colorado, which reported clutch sizes of7.5 (n = 69) (Keller et al. 1941) and 7.0 (n = 29) (Peterson 1980). Adults laid larger clutches but had similar nest success to yearlings. Commonly, adult clutch sizes are 0.2-2.1 eggs greater than yearlings (Schroeder et al. 1999). Connelly et al. (1994) also found nest success to be similar between ages. Combined nest success for yearling and adult females over 2 years in Axial Basin was 55% (n = 107). My values are normal to high in comparison with other Colorado studies, which range between 25% (Carr 1967) and 58% (Peterson 1980), similar to results in Idaho (52%, n = 166), and high compared to nest success rates in Oregon (14.5-30%) and Washington (36.7-40.9%) (Schroeder et al. 1999). Renesting effort was restricted to 4 females over 2 years. 25 The Axial Basin was characterized by habitat with less species richness, sagebrush, forb and litter cover and more bare ground than in the Danforth Hills (Table 1.5). Land in the privately-owned Danforth Hills is grazed by livestock, while the Axial Basin consists of a patchwork of private, BLM and state lands, and is used for grazing, crops and CRP. The dry winter and spring in 2002 exacerbated herbaceous habitat conditions in the Axial Basin, an effect not observed in the Danforth Hills. Habitat restrictions in the Axial Basin may have caused females to move further to nest sites and increase their brood home ranges (Chapter 4). Altitudinal differences, soil types, wildlife and livestock grazing may be factors influencing the vegetation characteristics in the 2 locations. Elevation ranges from 1,818 2,000 min the Axial Basin and from 2,001 - 2,388 min the Danforth Hills. Large numbers · of elk winter in the Axial Basin and their influence on vegetation was unmeasured. The grazing regime as it exists now in the Danforth Hills consists of cattle and sheep rotated through pastures from mid May until October. Total area of these allotments is 51,696 ha. The grazing regime in the Axial Basin, as per BLM data, includes sheep, cattle and horses, and grazing occurs on a total of 35, 938 ha of private, federal and state lands. Utilization occurs year-round on most allotments. Laycock (1979) indicated that a high level of spring grazing by cattle and sheep in a sage-steppe community reduces the vigor and production of herbaceous species. Grazing can move the sage-steppe community into a lower successional state with a diminished herbaceous understory (Laycock 1991). Wild and domestic ungulate grazing would add to this impact. The combination of abiotic factors coupled with large ungulate grazing and livestock grazing may contribute to the differences between habitat characteristics in the Axial Basin and Danforth Hills. 26 Nest Predators Potential nest predators in Moffat County include badgers (Taxidea taxus), Richardson's ground squirrels (Citellus richardsoni), common ravens (Corvus corax), American crows (Corvus brachyrhynchos), black-billed magpies (Pica hudsonia), coyotes (Canis latrans), weasels (Mustela sp.), red foxes (Vulpes vulpes), raccoons (Procyon lotor), porcupine (Erethizon dorsatum) and gulls (Larus sp.). Interpreting cause-specific nest depredation from egg and nest remains can be confounded due to inter-specific overlap and intra-specific variation in nest destruction, stage of development, researcher experience, partial nest depredation, scavenging and parental activity at depredated nests (Lariviere 1999). Thus, despite inferring only mammalian or avian predation, my results should be viewed cautiously. Seventy percent of depredated nests in the Axial Basin and Danforth Hills were attributed to mammalian predators, although species specific nest depredations could not be estimated from available data. This was consistent with data from North Park, Jackson County, Colorado where badgers and ground squirrels caused the majority of nest depredations (Gill1964). Common ravens and coyotes preyed upon 76% of nests (n = 21) in Wyoming (Allred 1942). Autenrieth (1981) determined common ravens to be the most important nest predator in Idaho. Similarly, in Oregon, common ravens accounted for 66% of 131 depredated nests (Batterson and Morse 1948). Avian predators in the Axial Basin and Danforth Hills accounted for only 5% of destroyed nests. Vegetation Sampling The only macrohabitat characteristics to vary between random and nest sites was distance to nearest anthill. This may suggest that greater sage-grouse are selecting nest sites with close proximity to anthills in order to provide early sustenance for chicks. Insects r - - - - . , . - - - - - - - - - - - - - - - -- -- -- - - -- - -- - - - -- - - - - - - 27 dominate the diet of chicks in their first 2 weeks, and ants (Hymenoptera) may provide a consistent food resource (Knowlton and Thomley 1942, Klebenow and Gray 1968, Drut et al. 1994). No macrohabitat differences existed between successful and unsuccessful nests. These results may be due to little habitat diversity within the physiographic characteristics at my study sites (Porter and Church 1987, Aberg et al. 2000). Despite the apparent homogeneity of the shrub-steppe community in my study area, selection was apparent at the microhabitat scale. Greater sage-grouse females in the Axial Basin and Danforth Hills selected nest sites with greater shrub cover, greater shrub height and greater grass and forb cover than random sites. Klebenow (1969) noted that nest sites were generally under shrubs with more ground, lateral, and shrub cover than at random sites. Bowman and Harris (1980) indicated canopy· cover strongly influenced prey vulnerability. I ~ Successful nests tended to be found in areas of higher canopy cover (2: 15%) than unsuccessful nests (Klebenow 1969, Wallestad and Pyrah 1974, Gregg et al. 1994). In North Park, Colorado, nest cover averaged 44% (n = 19) (Schoenberg 1982). Additionally, medium height shrubs (30-70 em) have been deemed important for nest success (Wallestad and Pyrah 1974, Schoenberg 1982, Gregg et al. 1994, Delong et al. 1995, Connelly et al. 2000). ·Females in the study area selected nest sites that were characterized by a mean shrub height of 56.5 em and nest shrub height of79.4 em. Percent grass and forb cover was also greater at nest sites than at random. Grass height was also selected as a predictor of nest occurrence and was greater at random locations than at nest sites. The difference, however, was < 2 em and may not be biologically significant. _Guidelines for protection of greater sage-grouse breeding habitats suggest 15-35% canopy cover of sagebrush and perennial herbaceous cover averaging 2: 18 em in height and a 28 diversity of forbs (Connelly et al. 2000). Mean canopy cover of sagebrush at random sites in the Axial Basin and Danforth Hills was 17%, suggesting that the study site has appropriate canopy cover within the sage-steppe community. Grass height at the nest bowl and at 1m were variables that predicted nest success. This is consistent with other observational studies in Oregon and Idaho. Klebenow (1969), Wakkinen (1990), Gregg et al. (1994) and Delong et al. (1995) indicated tall grass cover (2: 18 em) as a parameter significant for nest success. While guidelines suggest a grass height of2: 18 em for nesting (Connelly et al. 2000), high nest success of 64% in 2001 and 49% in 2002 suggest that these guidelines may not be applicable to Colorado. Rather, minimum grass height, based on successful nests within the study area, should range between 15-18 em (Table 1.9). Apparent nest success for 2001 was greater than 2002 by 15%. Although this was not statistically significant, it may be biologically relevant. The nest habitat characteristics that varied among years were those that would have been influenced by the hotter, drier conditions in 2002 (Appendix 1-B). Species richness, cover board readings, grass height at the intersection of the transects and at 1 m, and visual obstruction readings were all lower in 2002 than 2001 (Table 1.6). There was a reduction of 5-7 em in grass height between 2001 and 2002 and visual obstruction at the intersection was lower in 2002. These are likely reasons that nest success was lower in 2002 and suggests that grass height, although not limiting the population in years with high precipitation may be an important element limiting nest success in drought years. Preference for individual plant species was apparent for forbs such as false dandelion, lupine, and salt and pepper parsley. These species may be selected for their structural characteristics; particularly lupine which ranges from 50-100 em tall (Kershaw et al. 1998). 29 Similarly, Apa (1998) found the family Leguminosae, which includes lupine, to be critical in the classification of greater sage-grouse nest sites from random sites in Idaho. False dandelion was found in the crops of greater sage-grouse carcasses in the study site as a forage item. Salt and pepper parsley may provide cover at nest sites. Globe mallow was avoided at nest sites, perhaps because this plant occupies dry, open habitat types (Kershaw et al. 1998). MANAGEMENT IMPLICATIONS There is concern over decreasing greater sage-grouse populations throughout their range. Causes are frequently attributed to habitat alteration due to habitat fragmentation, land conversion, overgrazing, introduction of exotic weeds, pesticides and altered fire regimes .(Miller and Eddleman 2000). My results indicated the importance of shrub cover, native herbaceous cover, .and grass height on nest success. Successful nests in the study site featured grass heights of 15 em and 18 em at the nest bowl and at 1m from the nest bowl, respectively. Current habitat guidelines suggest grass heights be maintained at 2:. 18 em. However, given the normal to high nest success in the population, grass heights of 15-18 em may be adequate. Nest success in the study site varied by year, as did visu~l obstruction and grass height, which illustrates the interaction of these habitat components. Management actions should maintain native forb and grass communities as well as the shrub component in the sage-steppe community. The Axial Basin compared to the Danforth Hills has less shrub cover, visual obstruction, forb cover over the nest plot and grass height at 0 m, all variables deemed important for nest site selection and success. Difference between the 2 areas may be due to elevation, soil type, wildlife grazing and the timing and intensity of livestock grazing. A 30 large-scale, controlled grazing study that also evaluates the impact of elk foraging in winter could determine what may be influencing differences in habitat characteristics between the Axial Basin and Danforth Hills. While Beck and Mitchell (2000) stated that livestock grazing tends to be the most detrimental to greater sage-grouse during their nesting activities, nest success did not differ between locations with different grazing regimes in my study. Because of the importance of leks to breeding greater sage-grouse, I recommend that leks not be used as feeding sites for sheep during the breeding season (mid-March to mid-April). 31 LITERATURE CITED Aberg, J. G. Jansson, J.E. Swenson, and G. Mikusinski. 2000. Difficulties in detecting habitat selection by animals in generally suitable areas. Wildlife Biology 6:89-99. Allred, W.J. 1942. Predation data and the sage grouse. Wyoming Wildlife 7:3-4. Apa, A.D. 1998. Habitat use and movements of sympatric sage and Columbian sharp-tailed grouse in southeastern Idaho. Ph.D. Dissertation, University of Idaho, Moscow, Idaho. Autenrieth, R.E. 1981. Sage grouse management in Idaho. Wildlife Bulletin 9. Idaho Department ofFish and Game, Boise, Idaho. Batterson, W.M., and W.B. Morse. 1948. Oregon sage grouse. Oregon Game and Fish Commission Fauna Series 1. Portland. 29 pp. Bergerud, A.T. 1988. Population ecology ofNorth American grouse. Pages 578-648 in A.T. Bergurud and M.W. Gratson, eds. Adaptive strategies and population ecology of northern grouse. University of Minnesota Press, Minneapolis. Beck, J.L., and D.L. Mitchell. 2000. Influences of livestock grazing on sage grouse habitat. Wildlife Society Bulletin 28:993-1002 . . • Beier, P. 1989. Use of habitat by mountain beaver in the Sierra Nevada. Journal of Wildlife Management 53:649-654. Boisvert, J.H. 2002. Ecology of Columbian sharp-tailed grouse associated with conservation reserve program and reclaimed surface mine lands in northwestern Colorado. M.S. Thesis, University of Idaho, Moscow 184pp. Bowman, G.B., and L.D. Harris. 1980. Effect of spatial heterogeneity on ground nest depredation. Journal of Wildlife Management 44:806-813. Braun, C.E. 1981. Vulnerability and population characteristics of sage grouse in Moffat County. Final Report, Federal Aid Wildlife Restoration Project W-37-R, Work Plan 3, Job 11. Colorado Division of Wildlife, Fort Collins. 29-73. Braun, C.E. 1998. Sage grouse declines in western North America: what are the problems? Proceedings of the Western Association of State Fish and Wildlife Agencies 78:139-156. Braun, C.E., T. Britt, and R.O. Wallestad. 1977. Guidelines for maintenance of sage grouse habitats. Wildlife Society Bulletin 5:99-106. 32 Call, M.W., and C. Maser. 1985. Wildlife habitats in managed rangelands- the great basin of southeast Oregon. Sage grouse. U.S Department of Agriculture, Forest Service. General Technical Report PNW-187. Cambell, S., A. Smith, S. Redpath, and S. Thirgood. 2002. Nest site characteristics and nest success in red grouse Lagopus lagopus scoticus. Wildlife Biology 8:169-174. Canfield, R.H. 1941. Application of the line interception method in sampling range vegetation. Journal of Forestry 39:388-394. Carr, H.D. 1967. Effects of sagebrush control on abundance, distribution and movements of Sage grouse. Colorado Game and Fish Project W-37-R-20. Work Plan 3, Job 8a. Job Completion Report. 106pp. Colorado Gap Analysis Project. 2001. "Land cover classification system". [http://ndis.nrel.colostat.edu/co/gap. Connelly, J.W., M.A. Schroeder, A.R. Sands, and C.E. Braun. 2000. Guidelines to manage sage grouse populations and their habitats. Wildlife Society Bulletin 28:967-985. Connelly, J.W., W.L. Wakkinen, A.D. Apa, and K.P. Reese. 1991. Sage grouse use of nest . sites in southeastern Idaho. Journal of Wildlife Management 55:521-524. Connelly, J.W., K.P. Reese, W.L Wakkinen, M.D. Robertson, and R.A. Fischer. 1994: Sage grouse ecology report. Idaho Department ofFish and Game Job Completion Report. W160-R-19. Subproject 9. Boise. Connelly, J.W., R.A. Fischer, A.D. Apa, K.P. Reese, and W. Wakkinen. 1993. Renesting by sage grouse in southeastern Idaho. Condor 95:1041-1043. Dargon, L.M., and R.J. Keller. 1940. Sage grouse surveys. Colorado Game and Fish Commission. Progress Report. 4-R. 28pp. Daubenmire, R. 1959. A canopy coverage method of vegetation analysis. Northwest Science 33:43-64. Delong, A.K., J.A. Crawford, and D.C. Delong, Jr. 1995. Relationships between vegetational structure and predation of artificial sage grouse nests. Journal of Wildlife Management 59:88-91. Drut, M.S., J.A. Crawford, and M.A. Gregg. 1994. Brood habitat use by sage grouse in Oregon. Great Basin Naturalist 54:170-176. 33 ,. Gill, R.B. 1964. Effects of sagebrush control on distribution and abundance of sage grouse. Colorado Game, Fish and Parks Department. Project W-87-R-17, Work Plan 8, Job 8. Job Completion Rep. 185pp. Gill, R.B. 1966. A literature review on the sage grouse. Department ofFish and Parks Game Research Division and Cooperative Wildlife Research Unit, Final Report 6:1-38. Gregg, M.A., J.A. Crawford, M.S. Drut, and A.K. Delong. 1994. Vegetational cover and predation of sage grouse nests in Oregon. Journal ofWildlife Management 58:162-166. Hitchcock, A.S. 1971. Manual of the grasses of the United States. Dover Pub., Inc. New York. 1051pp. Johnson, R.A. and D.W. Wichern. 2002. Applied multivariate statistical analysis, fifth edition. Prentice-Hall Inc., Upper Saddle River, NJ. 767pp. Jones, R.E. 1968. A board to measure cover used by prairie grouse. Journal of Wildlife Management 32:28-31. Keller, R.J., H.R. Shepard, and R.N. Randall. 1941. Survey of 1941: North Park Jackson County, Moffat County, including comparative data of previous seasons. Colorado Game and Fish Commission. Sage grouse survey 3. 31 p. Kershaw, L., A. MacKinnon, and J. Pojar. 1998.·Plants of the Rocky Mountains. Lone Pine Publishing, Edmonton. 384pp. Klebenow, D.A. 1969. Sage grouse nesting and brood habitat in Idaho. Journal of Wildlife Management 33:649-662. Klebenow, D .A. 1972. The habitat requirements of sage grouse and the role of fire in management. Proceedings from Tall Timbers Fire Ecology Conference 12:305-315. Klebenow, D.A., and G.M. Gray. 1968. Food habits of juvenile sage grouse. Journal of Range Management 21:80-83. Knowlton, G.F., and H. F. Thomley. 1942. Insect food of sage grouse. Journal ofEconomic Entomology 35:107-108. Lariviere, S. 1999. Reasons why predators cannot be inferred from nest remains. Condor 101:718-721. Laycock, W.A. 1979. Management of sagebrush. Rangelands 1:207-210. Laycock, W.A. 1991. Stable states and thresholds of range condition on North American rangelands: a viewpoint. Journal of Range Mana.gement 44:427-433. 34 Leonard, K.M., K.P. Reese, and J.W. Connelly. 2000. Distribution, movements and habitats of sage grouse Centrocercus urophasianus on the Upper Snake River Plain of Idaho: changes from the 1950s to the 1990s. Wildlife Biology 6:281-286. Miller, R.F. and L.L. Eddleman. 2000. Spatial and temporal changes of sage grouse habitat in the sagebrush biome. Oregon State University Agricultural Experimental Station Technical Bulletin 151 :34pp. Mosley, J.C., S.C. Bunting, and M. Hironaka. 1989. Quadrat and sample sizes for frequency sampling mountain meadow vegetation. Great Basin Naturalist 49:241-248. Musil, D.D., K.P. Reese, and J.W. Connelly. 1994. Nesting and summer habitat use by translocated sage grouse ( Centrocercus urophasianus) in central Idaho. Great Basin Naturalist 54:228-233. National Oceanic and Atmospheric Administration. 2003. "Climate prediction center". [http://www.cpc.ncep.noaa.gov/products/monitoring_and_data/us_prec.html]. Ott, R.L., and M. Longnecker. 2001. An introduction to statistical methods and data analysis, fifth edition. Duxbury, Pacific Grove, CA. 1152pp. Patterson, R.L. 1952. The sage grouse in Wyoming. Sage Books, Inc., Denver, CO. 341pp. Peterson, B.E. 1980. Evaluation of the effects of changes in hunting regulations on sage grouse populations: evaluation of censuses of females . Colorado Division of Wildlife, Federal Aid Wildlife Restoration Project W-37-R-33, 115-116. Porter, W.F., and K.E. Church. 1987. Effects of environmental pattern on habitat preference · analysis. Journal of Wildlife Management 51 :681-685. Reynolds, C.J., P. Angelstam, and S. Redpath. 1988. Predators, their ecology and impact on gamebird populations. Pages 72-97 in P.J. Hudson and M.R.W. Rands (eds.), Ecology and management of game birds. BSP Professional Books, Oxford, England. 263 pp. Robel, R.J., J.N. Briggs, A.D. Dalton, and L.C. Hulbert. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management 23:295-297. Sargeant, A.B., M.A. Sovada, and R.J. Greenwood. 1998. Interpreting evidence of depredation of duck nests in the prairie pothole region. U.S. Geological Survey, Northern Prairie Wildlife Research Center, Jamestown, ND and Ducks Unlimited, Inc., Memphis, TN. 72pp. SAS Institute Inc. 1999. Version 8.2. SAS Institute, Carey, North Carolina, USA. l 35 Schneegas, E.R. 1967. Sage grouse and sagebrush control. Trans. North American Wildlife Conference 32:270-274. Schroeder, M.A. 1997. Unusually high reproductive effort by sage grouse in a fragmented habitat in north-central Washington. Condor 99: 933-941. Schroeder, M.A., J.R. Young, and C.E. Braun. 1999. Sage grouse (Centrocercus urophasianus) In A. Poole and F. Gills (eds). The Birds ofNorth America, No. 425. The Birds ofNorth America, Inc., Philadelphia, PA. Schoenberg, T.L. 1982. Potential impacts of strip mining on sage grouse movement. Colorado Game and Fish. Project W-37-R-35. Work Plan 3. Job 12. Job Completion Report. 59-142. Storch, I., and T. Willebrand. 1991. Workshop summary: management implications of nest and brood predation in grouse. Ornis Scadinaviva 22:62-69. Sveum, C.W. 1995. Habitat selection by sage grouse during the breeding season in southcentral Washington. M.S. thesis, Oregon State Univ., Corvallis. Swenson, J .E. 1986. Differential survival by sex in juvenile sage grouse and gray partridge. Ornis Scandinavica 17: 14-1 7. U.S Department of Agriculture, Natural Resource Conservation Service. 2003. The PLANTS Database [http://plants.usda.gov]. National Plant Center, Baton Rouge, Lousiana, lJSA.. Wakkinen, W.L. 1990. Nest site characteristics and spring-summer movements of migratory sage grouse in southeastern Idaho. M.S. Thesis. University of Idaho, Moscow, Idaho. Wallestad, R.O., and D. Pyrah. 1974. Movement and nesting of sage grouse hens in central Montana. Journal of Wildlife Management 35:129-136. Whitson, T., L.C Burrill, S.A. Dewey, D.W. Cudney, B.E. Nelson, R.D. Lee, and R. Parker. 2000. The weeds of the west, 9th Edition. Western Society of Weed Science, Newark, CA. 626pp. '-4.: ' ~ "f j.-, '!" =- ~ ~ .,.........-- ~ ~ ~~~.,..,.,. ~ 120 1• Q) r- (.) c: Q) '':::J 80 . - - f- 0 - - r- (.) (.) ~ 60 - 1' 1 - - - 1- C1) (.) 40 Q) - - .' a. 20 - 1- - r- - - - ~ - 1- - f- - 1- r- r- : ~ ~ - . ' c::: r - - 0 '- I Nest sites . D Random sites . 100 1- . ·.i : - ~ - 1- ~ I- . 1- 1- - f- - 1- 1- . 1- - :· - ; ,. t·· 1;, ["• 0 0:: t- 0:: <( E (/) co Q. E :::J .E ·c t- -- <( t- <( "'0 (9 ._ 0 co t- I" I· 1- Q. 0 Q. 0 () - - 1- 1- I' CD CD 0 (.) CD ..c 1- (.) c 0 1- ~ CD w Plant Species Figure 1.1. The 25 of the most commonly plant species at greater sage-grouse nest and random sites in Axial Basin and Danforth Hills, Colorado, 2001 and 2002. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A. w 0\ 37 Table 1.1. Number of nests and female greater sage-grouse and their nest fate in northwest Colorado, 200112002. Nests Females Successful Unsuccessful Successful Unsuccessful 2001 27 15 27 14 2002 32 33 32 32 Total 59 48 59 45 Adult 47 34 47 31 Yearling 12 14 12 14 Total 59 48 59 45 Year ~ Age r ~ ~ 38 Table 1.2. Cause-specific nest fate of radio-marked greater sage-grouse females in northwest Colorado, 200112002. Successful Year Other Depredated Mammalian Avian Unknown Abandoned 2001 30 10 1 2 1 2002 30 18 1 8 4 ... ""- . ~ , . ...., • ,.;: .....,. • ~- J , ...._ ~ ,. ~ ~~--.:: ,...~ :>lilt: 'f "¥ T "''I" Table 1.3. Differences in macrohabitat characteristics (mean and standard error) between years at random sites in the Axial Basin and Danforth Hills, Colorado, 2001/2002. Habitat Variable 2001 2002 (n =53) (n = 49) i df P-value Slope C) 4.9 (0.4) 7.3 (1.0) 12.8 3 0.802 Aspect C) 172.7 (16.1) 188.6 (17 .0) 78.8 4 0.361 2032 (16) 2061 (18) 84.0 3 0.449 Distance to nearest structure (m) 216.5 (43.7) 193.6 (33.5) 37.0 3 0.288 Distance to nearest 2-track (m) 166.1 (23.7) 162.1 (25.4) 52.5 3 0.128 2.6 (0.1) 2.6 (0.2) 0.7 3 0.853 Elevation (m) Distance to nearest anthill * * Categorical data only (1 = 0-25 m, 2 = 26-50 m, 3 > 51 m) w \0 J ' -4..- .- ' . ~ ~ ... .... ,.. .. ~ ' t ..1. • -"" . -~ j ~ ~ ~ ~ ~ ~ ~ T T Table 1.4. Means and standard errors of macro habitat characteristics and results of chi-squared tests for successful and non-successful greater sage-grouse nest sites and random sites in Axial Basin and Danforth Hills, Colorado, 200112002. Random (n=93) Nests (n=100) P-value Successful (n=50) Unsuccessful (n=42) P-value Slope 6.0 (0.5) 5.2 (0.5) 0.441 5.4 (0.6) 4.7 (0.7) 0.685 Aspect 180(11) 208 (12) 0.075 202 (16) 201 (19.0) 0.988 Elevation 2046 (12) 2045 (10) 0.775 2045 (13) 2040 (16) 0.442 Distance to nearest visible structure 206.7 (29.7) 205.1 (36.7) 0.290 237.2 (66.9) 184.4 (39.1) 0.496 Distance to nearest 2-track 164.3 (17.2) 146.0 (14.1) 0.189 155.6 (20.6) 129.7 (20.2) 0.538 Distance to nearest ant-hill* 2.3 (0.1) 2.6 (0.1) 0.038 2.3 (0.1) 2.3 (0.1) 0.870 Habitat Variable * Categorical data only +::. 0 I 41 I ~ Table 1.5. Microhabitat characteristics varying between locations at random sites (mean and standard error) in Moffat County, Colorado 2001/2002. ANOVA P-values are reported. Danforth Hills Species richness Axial Basin (n=41) 15.1 (0.6) Cover board Habitat variable ~ • f~ ~ F-value P-value 18.1 (0.7) 8.98 0.004 12.2 (1.4) 14.3 (1.5) 1.49 0.226 Percent shrub cover a 19.7 (2.2) 29.4 (2.8) 2.86 0.095 Percent sagebrush cover a 23.2 (1.7) 31.7 (2.1) 9.82 0.002 Intersection shrub height 40.7 (5.0) 62.2 (6.9) 5.44 0.023 Average shrub height 45.7 (4.1) 58.1 (5.1) 2.39 0.127 Grass height Om 13.2 (0.9) 18.6 (2.0) 18.17 <0.001 Grass height 1m 14.6 (0.8) 16.1 (1.6) 1.27 0.263 Percent forb cover at 0 m a 4.7 (0.9) 8.7 (1.3) 1.95 0.167 Average forb cover a 4.4 (0.5) 8.1 (1.0) 7.63 0.007 Percent grass cover at 0 m a 4.8 (0.8) 8.5 (2.8) 0.51 0.478 Average grass cover a 4.8 (0.6) 7.9 (2.0) 1.21 0.275 Percent exotic cover at 0 rna 31.0 (4.1) 34.4 (4.6) 0.01 0.918 Average exotic cover a 32.6 (3.3) 27.3 (3.1) 1.17 0.283 Percent bare ground 0 m a 28.5 (4.8) 17.8 (4.7) 3.81 0.055 Average bare ground a 28.0 (3.8) 17.5 (3.1) 5.35 0.024 Percent litter cover 0 m a 70.5 (4.6) 79.5 (4.6) 3.93 0.051 Average litter cover a 72.5 (3.6) 80.2 (2.9) 3.76 0.056 VISOBa11b 11.7 (2.3) 20.9 (3.0) 4.46 0.038 VISOBa13b 26.4 (2.6) 37.7 (2.6) 4.39 0.040 VISOBa21b 3.2 (1.4) 7.3 (1.9) 2.38 0.128 VISOBa23b 8.9 (1,.9) 21.9 (3.0) 9.07 0.004 VISOBb11b 8.5 (1.8) 16.9 (2.9) 5.58 0.021 VISOBb13b 19.1 (2.2) 30.3 (2.7) 6.91 0.011 VISOBb21b 2.1 (1.3) 4.9 (1.8) 1.84 0.179 VISOBb23b 4.6 (1.4) 11.6 (2.2) 4.77 0.032 (n=32) . a Arcsin transformed for MANOV A, untransformed mean and SE are reported. b Visual obstruction readings where the letter responds to reading height (a= 0.5m, b = 1.5 m), the first number responds to the 50cm demarcations on the pole (1 = 0 m- 0.5 m, 2 = 0.5 m- 1.0 m) and the second number responds to reading distance from the pole (1 = 2.5 m, 2 = 5.0 m, 3 = 10 m). . ., • 43 ~> ~ Table 1. 7. Pattern of 5 Principal Components derived from analysis of 21 microhabitat variables from 75 random plots in Axial Basin and Danforth Hills, Colorado, 2001/2002 . • Structural Feature ~ "1 .. Percent shrub cover PC1 PC2 PC3 0.237 PC5 0.325 Percent sagebrush cover Shrub height at 0 m PC4 0.305 0.249 • Average shrub height 0.348 ... Forb cover at 0 m -0.228 0.546 Average forb cover -0.220 0.525 .. • ), Grass cover at 0 m 0.309 0.375 Average grass cover 0.312 0.280 Exotic herbaceous cover at 0 m 0.505 Average exotic herbaceous cover ~ -1 I' -0.302 Bare ground 0 m -0.151 0.388 Average bare ground -0.166 0.328 Litter 0 m -0.385 Average litter -0.337 VISOB 0.5 m from 2.5 m 0-50 em 0.310 VISOB 0.5 m from 10m 0-50 em 0.313 VISOB 0.5 m from 2.5 m 0.5-1 m 0.287 VISOB 1.5 m from 2.5 m 0-0.5 m 0.306 0.392 0.323 -0.319 1-r -0.202 \ ... VISOB 1.5 m from J10 m 0-0.5 m 0.330 VISOB 1.5 m from 2.5 m 0.5-1 m 0.257 VISOB 1.5 m from 10m 0.5-1 m 0.295 -0.218 Eigenvalue 7.1 3.2 2.9 1.8 1.5 Percent of variation explained 33.7 15.3 13.7 8.5 7.2 Cumulative percent 33.7 49.0 62.7 71.1 78.4 44 • ~ ~ ~ Table 1.8. Logistic regression models of habitat variables predicting the binary variables for random and nest sites or successful and unsuccessful nests of greater sage-grouse in northwest Colorado 2001/2002 . Dependant Variable I ~ ,., ... .. Random/ Nest sites · ~- ~ .. ~ Successful/ Unsuccessful nests f ( Parameters Regression Coefficient (SE) Waldi df p PC 1 0.913 (0.160) 32.77 1 <0.001 PC5 -0.563 (0.167) 11.36 1 0.001 Grass height 0 m -0.080 (0.033) 5.89 1 0.015 Visual obstruction 0.5-1.0 m read from a height of0.5 m and a distance 10 m -0.656 (0.213) 9.50 1 0.002 PC5 0.547 (0.237) 5.34 1 0.021 Grass height 0 m 0.069 (0.042) 2.71 1 0.100 Grass height 1 m 0.087 (0.038) 4.40 1 0.022 45 Table 1.9. Mean and standard error (SE) for habitat variables measured at successful, unsuccessful, and random nest sites of greater sage-grouse in Moffat County, Colorado 2001/2002. Random Total nests Successful Unsuccessful nests Variable (n=93) (n=53) (n=40) (n=75) Species richness 18.9 (0.6) 19.9 (0.8) 17.6 (0.7) 16.6 (0.5) Cover board 17.7 (0.7) 18.6 (1.0) 16.5 (1.1) 12.9 (1.0) . Percent shrub cover a 34.1 (1.4) 33.9 (1.9) 34.2 (2.2) 24.0 (1.8) Percent sagebrush cover a 26.9 (1.4) 25.6 (1.7) 28.6 (2.2) 16.6 (1.5) ~ Nest shrub height 79.4 (2.5) 78.7 (3.2) 80.4 (4.0) 50.1 (4.2) "" Average shrub height 56.5 (2.2) 55.7 (3.1) 57.5 (2.9) 51.1 (3.2) Grass height 0 m 13.8 (0.7) 15.4 (0.9) 11.7 (0.9) 15.5 (1.0) Grass height 1 m 16.2 (0.8) 18.2 (1.1) 13.5 (1.0) 15.0 (0.8) Percent forb cover at 0 m a 6.9 (0.8) 7.7(1.4) 5.8 (0.7) 6.3 (0.8) Average forb cover a 8.4 (0.7) 9.3 (1.0) 7.2 (0.9) 6.2 (0.6) Percent grass cover at 0 m a 3.7 (0.4) 3.7 (0.6) 3.8 (0.7) 6.4 (1.3) Average grass cover a Percent herbaceous exotic cover at 0 m a Average herbaceous exotic cover a Percent bare ground 0 m a 4.4 (0.3) 4.8 (0.6) 3.9 (0.4) 6.1 (0.9) 30.5 (2.7) 30.5 (3.2) 30.5 (4.6) 31.9 (3.0) 29.3 (2.2) 28.6 (2.9) 30.2 (3.4) 29.5 (2.3) 4.9(1.1) 6.1 · (1.8) 3.5 (0.8) 23.3 (3.3) Average bare ground a 17.4 (1.6) 17.5 (2.1) 17.3 (2.5) 23.7 (2.6) Percent litter cover 0 m a 87.9 (1.8) 87.7 (2.6) 88.3 (2.3) 74.9 (3.2) Average litter cover a VISOBa11 b VISOBa13 b 77.5 (2.0) 78.5 (2.5) 76.2 (3.3) 75.5 (2.4) 38.4 (1.0) 46.8 (0.6) 39.0 (1.2) 47.4 (0.7) 37.7 (1.8) 46.1 (1.1) 15.7 (1.9) 31.5 (1.9) VISOBa21 b 13.5 (1.3) 13.4 (1.7) 13.6 (1.9) 4.8 (1.1) VISOBa23 b 24.6 (1.6) 23.8 (2.0) 25.8 (2.5) 14.6 (1.8) VISOBb11 b 33.2 (1.2) 32.6 (1.6) 34.0 (1.7) 12.2 (1.6) VISOBb13 b 40.5 (1.0) 40.8 (1.2) 40.1 (1.6) 24.1 (1.7) VISOBb21 b 9.4(1.1) 9.0 (1.4) 10.0 (1.7) 3.2 (1.0) , ~ ~ VISOBb23 b 15.9 (1.3)" 14.9 (1.7) 17.3 (2.1) 7 ~5 (1.3) a Arcsin transformed for MANOV A, untransformed mean and SE are reported b Visual obstruction readings where the letter responds to reading height (a= 0.5m, b = 1.5 m), the first number responds to the 50cm demarcations on the pole (1 = 0 m- 0.5 m, 2 = 0.5 m- 1.0 m) and the second number responds to reading distance from the pole (1 = 2.5 m, 2 = 5. 0 m, 3 = 10 m). 46 Appendix 1-A. Vegetation species and their native status identified at microhabitat plot sites of greater sage-grouse nests and random sites in the Axial Basin and Danforth Hills Colorado, 2001-2002. Code is the abbreviated genus and species name (first 2 letters of each). Nomenclature is taken from the PLANTS Database (USDA 2003). Scientific name • FORBS Achillea millefolium Agoseris glauca Allium spp. Alysum alyssoides Antennaria spp. Antennaria microphylla Arabis spp. Arabis drummondii Arcticum minimus Astragalus/Oxytopis Astragalus spp. Astragalus purshii Balsamorhiza sagittata Calochortus nuttallii Capsella bursa-pastoris Cardaria draba Castilleja miniata Came/ina microcarpa Chenopodium Chenopodium berlandieri Chenopodum multifidum Chenopodum vulvaria Chorispora tenella Cicuta douglasii Cirsium arvense Cirsium undulatum Cirsium vulgare Collinsia parviflora Collomia linearis Comandra umbellate Conium maculatum Crepis intermedium Cryptantha glomerata Cynoglossum officinale Delphinium bicolor Common Name Yarrow False dandelion Wild onion (genus only) Pale madwort Pussytoes (genus only) Littleleaf pussytoes Rockcress (genus only) Drummond's rockcress Common burdock Astragalus or Oxytropis Locoweed (genus only) Woollypod milk-vetch Arrow-leaved balsamroot Sego lily Shepards purse White top Indian paintbush Smallseed falseflax Pigweed, goosefoot N etseed lambsquarters Cutleaf goosefoot Stinking goosefoot Blue mustard Douglas' water hemlock Canada thistle W avyleaf thistle Bull thistle Maiden blue-eyed mary Narrow leafed collomia Bastard toadflax Poison hemlock Gray hawksbeard Cryphaea moss Hound's tongue Low larkspur Code Native Status Native a Native a Native a Introduced 1 b ALAL Native a Native a ANPA Native a Native a ARDR No record a ARMI Native a ASOX Native a Native a ASPU Native a BASA Native a CANU Introduced a CABU Introduced a CADR -Nativea CAMI CAMIC Introduced 1 b Native a Native a CHBE Native a CHMU Introduced a CHVU Introduced 1 2 3 a b CHTE Native a CIDO Introduced 1 2 3 ab CIAR Native b CIUM Introduced a CIVU Native a COPA Native a COLI Native a COUM Introduced 1 b COMA Native a CRIN Native a CRGL Introduced 1 b CYOF Native a DEBI ACMI AGGL 47 )- ,. Appendix 1-A. Continued. . .. ... ~ ~ ~- Scientific name Common Name Code Native Status Descurainia sophia Echinocereus spp. Erysimum asperum Erigeron engelmannii Erigeron filifolius Erigonum ovalifolium Tansy mustard Flixweed Hedgehog cactus (genus only) Western wallflower Engelmann's fleabane Thread leafed daisy Cushion buckwheat DESO Introduced I 3 a b Native a Native a Native a Native a Native a Erigonum umbellatum Sulphur buckwheat Euphorbia esula Fritillaria atropurpurea Galium boreale Gentian linearis Geranium caespitosum Glycyrrhiza lepidota Grindelia squarrosa Helianthus H eterotheca villosa Huechera parvifolia Hydrophy llum Ipomea spp. Ipomopsis aggregata Kochia scoparia Lactuca occidentalis Lactuca serriola Lathyrus ochroleucus Lepidium perfoliatium Linum lewisii Linaria vulgaris Lithospermum ruderale Lomatium orientale Lomatium triternatum Lpomopsis aggregata Lupinus minimus Lupinus sericeus Lygodesmia juncea Leafy Spurge Leopard lily Northern Bedstraw Narrow leaf gentian Common wild geranium Wild Licorice Curly-cup gum weed Sunflower Golden aster Saxifrage Water leaf Morning glory (genus only) Scarlet gilia Kochia Bluebird sitckseed Prickly lettuce Creamy peavine Clasping pepperweed Western Blue flax Yellow toad flax Puccoon Salt and pepper parsley Parsley Scarlet gilia Cushion lupine Silky lupine Skeletonweed Machaeranthera canescens Hoary aster Medicago sativa Melilotus officinalis Mentha spicata Mertensia longiflora Alfalfa Yellow sweet clover Spearment Bluebells ,... ~ ... ~- .. II.- -~ ..... .... ~ ~ ~ ERAS EREN ERFI EROV ERU M EUES FRAT GABO GELI GECA GLLE GRSQ HEVI HUPA Native a Introduced a 1 2 Native a Native a Native a Native a Native a Native a Native a Native a No record a Native a Introduced a Native a Introduced 1 b Native a Introduced 1 b Native a Introduced 1 b Native a Introduced 123 ab Native a Native a Native a Native a Native a Native a Native a IPAG KOSC LAOC LASE LAOC LEPE LILE LIVU LIRU LOOR LOTR LPAG LUMI LUSE LYJU MAC Native a A MESA Introduced a MEOF Introduced 1 2 b MESP Introduced a MELO Native a 48 ._ . Appendix 1-A. Continued. l. .,. ....,- ~ r 1>- Scientific name Common Name Mictrostersis Microsteris Nemophilia breviflora Nemophilia Sanfoin Onobrychis viciaefolia Opuntia polyacantha Starvation cactus Oxytropis sericeus Locoweed Penstemon (genus only) Penstemon spp. Penstemon gairdneri Gardner's beardtongue Phlox hoodii Phlox · Phlox Phlox longifolia Plantain Plantago Plantago hookeri California plantain Plantago patagonica Woolly plantain Knotweed (genus only) Polygonum spp. Cinquefoil Potentilla spp. Pseudostellari jamesiana Tuber starwort Pricklefruit buttercup Ranunculus muricatus Watercress Rorippa nasturtium Rumex spp. Sorrel Sa/sola iberica Russian thistle Senecia canus Woolly groundsel Senecio integerrimus Butterweed, Golden ragwort Solidago spp. Goldenrod (genus only) Nightshade (genus only) Solanum spp. Copper mallow Sphaeralcea coccinea Taraxacum officinale Dandelion Field pennycress Thlaspi arvense Townsendii hookeri Easter daisy Salsify Tragapogan dubius Trifolium arvense Rabbit foot clover Trifolium fragiferum Strawberry clover Verbascum thapsus Common mullein Viola nuttallii Violet Viscia americana American vetch Viscia vilosia Winter vetch Zigadenus venenosus Death camas GRASSES Achnatherum hymenoides Aegilops cylindrical Agropyron cristatum Agropyron dasystachyum Indian Rice Grass Jointed goatgrass Crested wheatgrass Thickspike wheatgrass Code Native Status SPCO TAOF THAR TOHO TRDU TRAR TRFR VETH VINU VIAM VIVI ZIVE No Record a Native a Introduced a Native a Native a Native a Native a Native a Native a Native a Native a Native a Native a Native a Native a Introduced 1 b Native a Both a Introduced 2 b Native a Native a Native b No Record a Native a Introduced 1 b Introduced 1 b Native a Introduced 1 Zb Introduced a Introduced a Introduced 1 b Native a Native a Introduced a Native a ACHY AECY AGCR AGDA Native ac Introduced 1 a Introduced 2 a Native c NEBR ONVI OPPO OXSE PEGA PHHO PHLO PLHO PLPA PSJA RAMU RONA SAID SECA SEIN 49 • .._ Appendix 1-A. Continued . ~ Scientific name Common Name Code Native Status .. Agropyron griffithii Agropyron intermedium Agropyron repens Agropyron riparum Agropyron trachycaulum Agrostis stolonifera Bromus inermis Bromus japonicus Bromus marginatus Bromus tectorum Carex spp. Dactylis glomerata Elymus cinereus Elymus elmoides Festuca spp. F estuca occidenta/is Hespero-stipa comata Hilaria spp. Hordeum jubatum Koelaria cristata Melica bulbosa Nassella virdula Pascopyron smithii Phleum pratense Poa annua Poa bulbosa Poa compressa Poa pratensis . Poa fendleriana Poa secunda Scirpus spp. Triticum aestivum Vulpia octojlora Montana wheatgrass Intermediate wheatgrass Quackgrass Stream bank wheatgrass Slender wheat grass Red Top Bent Smooth brome Japanese brome Mountain brome Cheat grass Sedge (genus only) Orchard grass Basin Wild Rye Squirrel-tail Fescue (genus only) Western fescue Needle and thread grass Curly mesquite (genus only) Foxtail barley June Grass Onion grass Green needle grass Western wheat grass Timothy Annual bluegrass Bulbus bluegrass Canada bluegrass Kentucky bluegrass Mutton grass Sandberg blue grass Sedge (genus only) Wheat Six -weeks grass AGGR AGIN AGRE AGRI AGTRA AGST BRIN BRJA BRMA BRTE Native ac Introduced 1 a Introduced 1 2 a Native ac Native ac Native ac Introduced 1 a Introduced 1 2-a Native ac Introduced c Native a Introduced 1 2 a Native a c Native a Introduced a Native a c Native a c Native a Native a c Native ac Native a c Native a c Native a c Introduced 1 2 a Introduced a Introduced 1a Introduced 1a Introduced 1 a c Native a c Native a c Native a Introduced 1 2 a c Native a SHRUBS Acer glabrum Amelanchier alnifolia Artemisia cana Artemisia ludoviciana Artemisia tridentata Atriplex canescens Ceratoides spp. Douglas maple Serviceberry, Saskatoon Silver sage Mugwort, Medicine sage Big sagebrush Four-wing salt brush Winterfat ~ ... 4t r ~ " ... .._ .... ~ DAGL ELCI ELEL FEOC HECO HOW KOCR MEBU NAVI PASM PHPR POAN POBU POCO POPR POFE POSE TRAE vuoc ACGL AMAL ARCA ARLU ARTR ATCA Native a Native a Native a b Native a b Native a b Native a b Introduced a t • ~ r ~ 50 / Appendix 1-A. Continued . Scientific name Common Name Code Native Status Ericameria nauseosus Chrysothamnus viscidiflorus Berberis fendleria Gutierrezia sarothrae Hippochaeta spp. Juniperus spp. Mahonia Repens Purshia tridentata Prunus virginiana Quercus gambelii Ribies spp. Rosa woodsii Sarcobatus vermiculatus Salix spp. Symphoricarpos oreophilus Tetradymia spp. Gray Rabbitbrush Douglas Rabbitbrush Fendler Barberry Broom snakeweed Horsetail (genus only) Juniper (genus only) Creeping barberry Antelope brush Choke cherry Gambel's Oak Gooseberry (genus only) Wood's rose Greasewood Willow (genus only) Mountain snowberry Horsebrush (genus only) ERNA CHVI BEFE GUSA Native a Native a Native a Native a b No record a Native a Native ab Native a b Native a Native a b MARE PUTR PRVI QUGA ROWO SAVE SALIX SYOR Sources: a The PLANTS Database (USDA 2003) b The Weeds of the West (Whitson et al. 2000) c Manual of the grasses of the United States (Hitchcock 1971) 1 Introduced from Europe Introduced from Asia 3 Noxious weed status 2 Nativeab Nativeab Native a Native a Native a 51 Appendix 1-B. Climatic variables from Craig 4 SW weather station in Colorado, 2001-2002 (NOAA 2003). MNTM= mean monthly temperature, TPCP= total monthly precipitation, and difference between 2002 and 2001. ,.. r ..._ ~ January February March April May June July August September October November December 2001 -10 -5 1 8 11 16 N/A 20 N/A 9 3 -7 MNTMCC) 2002 Difference -8 +2 -8 -3 -3 -4 0 8 11 0 +3 19 22 N/A -1 19 15 N/A -2 7 N/A N/A N/A N/A 2001 85 113 110 117 192 63 91 134 91 . 134 129 55 TPCP (mm) 2002 Difference -18 67 24 -89 119 +19 120 +3 13 -179 26 -37 71 -20 +62 196 157 +66 134 0 N/A N/A N/A N/A 52 CHAPTER 2. AN EVALUATION OF THE TIMING OF GREATER SAGE-GROUSE NEST VEGETATION SAMPLING INTRODUCTION Microhabitat sampling for nest use and selection studies of ground nesting sagesteppe birds (Reynolds 1981, With and Webb 1993) and upland game birds (Storaas and Wegge 1987, Kilbride et al. 1992, Rumble and Hodorff 1993, McKee et al. 1998) are made after a female ceases nesting so as not to disrupt incubation. This is also prevalent in habitat studies of greater sage-grouse (Klebenow 1969, Wakkinen 1990, Connelly et al. 1991, Gregg et al. 1994, Musil et al. 1994, Apa 1998, Aldridge 2000). These measurements have limitations; they do not accurately represent vegetation composition and structure at the time of nest site selection as plant growth and/or senescence is ignored (Reese et al. 1987). Habitat management guidelines of greater sage-grouse currently use post-hatch vegetation characteristics although these may be sampled> 30 days after nest site selection. The impact of this time lag on habitat characteristics is unknown. My objectives were (1) to estimate habitat characteristics at the time of nest initiation at previous years nest sites, (2) to compare those to habitat characteristics measured at the time of hatch, and (3) to predict habitat characteristics incorporating the time lag. METHODS To describe the vegetation greater sage-grouse select at the time of nest initiation, vegetation was sampled in 2002 at nest initiation (18 April - 29 April) and at hatch (18 May4 June) at 30 nest sites randomly chosen from 40 nests active in 2001. Sixteen of these had 53 • • been successful, and 14 had been unsuccessful. All vegetation measurements were the same as in Chapter 1 with the following exception: grass and forb cover were not quantified to nearest taxon. Vegetation variables were divided into 2 groups, those measured at the nest bowl and those measured in the surrounding area (2.5, 5, and 10m). Grass height measurements were summarized at 0 and 1 m. Means of cover estimates were obtained by using the mid-point value of each cover class. Variables were assessed for normality and arcsin transformed when necessary. A correlation analysis was performed on the habitat variables to reduce the dimensionality. Visual obstruction measurements read at 5.0 m were correlated to those measured at 2.5 m (r2 = 0.57- 0.90, P < 0.001) and measurements read from 0.5 m were correlated to those read at 1.5 m (/ = 0. 78 - 1.00, P < 0.001 ), thus the latter in both cases was ·removed from further analysis. The remaining visual obstruction readings were correlated to nest shrub height and also removed from further analysis (r2 = 0.36- 0.74, P < 0.004). Percent shrub cover was correlated to big sagebrush cover and therefore percent shrub cover was removed from further consideration (r2 = 0.36, P < 0.006). Understory cover measurements at the nest bowl and averaged over the nest plot were correlated, thus the r average values were retained as the values consisted of a greater number of samples (Table 2.1). I used a randomized complete block design and a multivariate analysis of variance (MANOVA) to assess differences in the remaining 10 microhabitat variables due to timing of vegetation sampling. Paired nest plots were blocked. Evaluation of univariate ANOVA and canonical analysis was used to determine which variables contributed to overall differences. Habitat variables that accounted for differences between initiation and hatch were assessed l 54 for linear correlation based on the date of vegetation sampling (1 - 48 days), corresponding to nest initiation and hatch (SAS Institute 1999). RESULTS ... A difference was detected in vegetation by time of measurement (Wilk's 'A= 0.27; F = 5.32; 10,20 df; P = 0.001) for the 10 habitat variables. The habitat variables contributing to this difference included grass heights at 0 and 1m, and average percent grass cover (Table 2.2). Other habitat characteristics, especially those associated with sagebrush heights and coverage, did not differ from initiation to hatch (Table 2.2). Grass height at 0 and 1 m showed a significant linear correlation between initiation and hatch (Table 2.3). The linear regression equations that resulted were grass height in em at 0 m = 8.93 em + 0.17 (date of vegetation sampling) and grass height at 1m = 7.98 em+ 0.12 (date of vegetation sampling) (Figure 2.1 ). Four outliers were removed from the percent grass cover data set and then • analyzed for linear correlation. There was no linear relationship for percent grass cover (Table 2.3). • Using the slope of the regression equations as a constant, 45 days post-initiation as the sampling time, and habitat management guidelines of Connelly et al. (2000), which suggest grass height at hatch to be ~ 18 em, grass height at both 0 and 1 m at initiation needs to be~ 10.4 em. However, these management guidelines may have limited applicability in Colorado given that no habitat studies from Colorado are included in deriving this minimum standard. Nest success for the population of greater sage-grouse in the Axial Basin and Danforth Hills was normal to high for both years with 64% and 49% success in 2001 and 2002, respectively. Mean grass heights were 18.9 and 18.3 em at successful nest in 2001 and 55 13.4 and 13.8 em in 2002 (Chapter 1). With these values as guidelines at hatch, grass height at initiation should range from 5.8 - 11.3 em at the nest plot and 8.4- 12.9 em at 1 m. DISCUSSION Characteristics of vegetation at the time of initiation may best represent structure females are cueing into for nest site selection. At nest initiation females selected shrubs with a mean height of 82 em, within a stand of shrubs with average height of 52 em. Sagebrush canopy cover was 30%. These components corresponded to greater sage-grouse habitat t characteristics associated with nest sites at hatch and to those recommended in the literature (Connelly et al. 2000). Grass cover was 4% and heights at initiation were 9-10 em, which were lower than those reported at time of hatch and as recommended for nesting habitat (Connelly et al. 2000). These were the variables that contributed most to the difference between sampling times . • Grass height at the nest bowl and at 1 m formed predictive linear relations between sampling times from initiation and hatch. Grass height and cover were deemed important for nest site selection and success in the Axial Basin and Danforth Hills (Chapter 1). Linear correlation between variables at initiation and hatch indicated that hens selected characteristics of the herbaceous understory that would serve them best throughout incubation. Additionally, correlations enable managers to predict the change in habitat characteristics over the time frame from initiation to hatch and allow some freedom in the timing of sampling over this period. The assumptions of this sampling procedure were that there was no variation in grazing by domestic and wild ungulates between years and that the same nest sites in 2001 56 represented those selected in 2002. The degree of nest fidelity gives the second assumption credibility. The average distance adult females moved between consecutive nests was 0.45 km (Chapter 4). Measurements made at actual nest sites in 2002 (n =54) were compatible to those at nest sites from 2001 at hatch. Grass heights measured in 2001at the nest bowl and 1 m were similar to those reported at hatch with heights of 12.5 em, SE = 0.8 and 13.6 em, SE = 0.8. MANAGEMENT IMPLICATIONS Data collected in this study in 2002 of nests from 2001 provided a more accurate measurement of the nest site characteristics that greater sage-grouse females were selecting without having to disturb incubating females. The only habitat characteristics that differed from the time of initiation to hatch were grass heights and grass cover. Because these variables showed a linear correlation between initiation and hatch, their values can be predicted over the time period between initiation and hatch. Given that nest success for the population is normal to high for the species, I used grass heights measured at hatch in the regression equations to predict a minimum guideline of 6 - 13 em grass height for nesting measured at clutch initiation. Apart from grass measurements, no other habitat variables differed between sampling times, indicating that current habitat sampling techniques provide an adequate description of nest site characteristics at selection. Decisions regarding management of nest habitat conditions may be based on range measurements made in the spring. Grazing of cattle begins by 1 May in the Danforth Hills • and by early March in the Axial Basin on areas used for nesting by greater sage-grouse. Additionally, controlled burning practices in the Danforth Hills commence in late April and 57 early May. Therefore, when assessing range conditions at these time periods, minimum structural variables of habitat at the time of nest initiation need to be considered for any ground-nesting species that relies on grass for cover during incubation . • 58 LITERATURE CITED Aldridge, C.L. 2000. Reproduction and habitat use by sage grouse (Centrocercus urophasianus) in a northern fringe population. M.S. Thesis University of Regina, Regina, Saskatchewan. 109pp. Apa, A.D. 1998. Habitat use and movements ofsympatric sage and Columbian sharp-tailed grouse in southeastern Idaho. Ph.D. Dissertation, University of Idaho, Moscow, Idaho. Connelly, J.W., M.A. Schroeder, A.R. Sands, and C.E. Braun. 2000. Guidelines to manage sage grouse populations and their habitats. Wildlife Society Bulletin 28:967-985. Connelly, J.W., W.L. Wakkinen, A.D. Apa, and K.P. Reese. 1991. Sage grouse use of nest sites in southeastern Idaho. Journal of Wildlife Management 55:521-524. Gregg, M.A., J.A. Crawford, M.S. Drut, and A.K. Delong. 1994. Vegetational cover and predation of sage grouse nests in Oregon. J oumal of Wildlife Management 58: 162-166. l I Kilbride, K.M., J.A. Crawford, K.L. Blakely, and B.A. Williams. 1992. Habitat use by breeding female California quail in western Oregon. Journal of Wildlife Management 56:85-90. Klebenow, D.A. 1969. Sage grouse nesting and brood habitat in Idaho. Journal of Wildlife Management 33:649-662. McKee, G., M.R. Ryan, and L.M. Mechlin. 1998. Predicting prairie chicken nest success from vegetation and landscape characteristics. Journal of Wildlife Management 62:314321. Musil, D.D., K.P. Reese, and J.W. Connelly. 1994. Nesting and summer habitat use by translocated sage grouse (Centrocercus urophasianus) in central Idaho. Great Basin ··Naturalist 54:228-233. Reese, K.P., J.A. Kadlec, and L.M. Smith. 1987. Characteristics of islands selected by nesting Canada geese, Branta canadensis. The Canadian Field Naturalist 101:539-542. Reynolds, T.D. 1981. Nesting of the sage thrasher, sage sparrow and Brewer's sparrow in southeastern Idaho. Condor 83:61-64. Rumble, M.A., and R.A. Hodorff. 1993. Nesting ecology of Merriam's turkeys in the Black Hills, South Dakota. Journal of Wildlife Management 57:789-801. SAS Institute Inc. 1999. Version 8.2. SAS Institute, Carey, North Carolina, USA. 59 Storaas, T., and P. Wegge. 1987. Nesting habitats and nest predation in sympatric populations of capercaillie and black grouse. Journal of Wildlife Management 51:167172. Wakkinen, W.L. 1990. Nest site characteristics and spring-summer movements of migratory sage grouse in southeastern Idaho. M.S. Thesis. University of Idaho, Moscow, Idaho. With, K.A., and D.R. Webb. 1993. Microclimate of ground nests: the relative importance of radiative cover and breaks for three grassland species. Condor 93:401..:413. I I i ~ I~ ~ 1 .. • -~ • c~ ....- 30 I 25 ., rI • Grass height at 0 m 0 Grass height at 1 m • - - Regression line 20 ~ s (.) -+-> ,.s:::: bl) 'Q) ~ oe • "-"' 15 • • - U'l U'l ro 1-4 C!J 10 0 0 I 4/15 0 • • 0 4/22 ••• i v 0 ~ 0 • 0 ~ 5 ., • • 0 s• 0 • a 516 5/13 5/20 ov oO 0 0 • • ~ • 4/29 0 5/27 6/3 Date of Vegetation Sampling Figure 2.1. Regression of grass heights measured at 0 and 1m at greater sage-grouse nest sites at initiation (4/18-4/29) and nest hatch (5/17-6/4) in Moffat County, Colorado 2002. 0\ 0 ' • ' .. ~ Table 2.1. Correlations between understory cover characteristics of greater-sage grouse nest habitat in northwest Colorado. Pearson correlation coefficients and associated P-values are reported. Forbs 0 m Forbs 0 m P-value Average forbs P-value Average forbs Grass 0 m Average grass Bare ground 0 m Average bare ground Litter 0 m Average litter 1 0.604 < 0.001 1 Grass 0 m P-value 0.078 0.554 -0.108 0.412 1 Average grass P-value 0.119 0.366 0.081 0.537 0.586 < 0.001 1 Bare ground 0 m P-value -0.227 0.081 -0.306 0.018 0.111 0.397 0.204 0.118 1 Average bare ground P-value -0.267 0.039 -0.450 < 0.001 0.036 0.784 0.024 0.857 0.396 0.002 1 Litter 0 m P-value 0.120 0.360 0.270 0.037 -0.475 < 0.001 -0.434 0.001 -0.821 < 0.001 -0.455 < 0.001 1 Average litter P-value 0.218 0.094 0.376 0.003 -0.204 0.117 -0.186 0.156 -0.350 0.006 -0.928 < 0.001 0.545 < 0.001 1 0\ ~ j 62 Table 2.2. Mean and standard error (SE) for habitat variables included in MANOVA at initiation (4/18-4/29) and hatch (5/17-6/4) for 30 greater sage-grouse nests from 2001 measured in 2002 in the Axial Basin and Danforth Hills, Colorado. Variable • .• Initiation SE Hatch SE Cover board 15.4 0.05 13.6 1.4 Percent sagebrush cover a 29.9 2.6 29.9 2.8 Nest shrub height (em) 81.9 3.5 82.8 3.7 Average shrub height (em) 51.9 3.1 52.7 3.3 Grass height 0 m (em) b 10.0 0.6 15.6 1.1 Grass height 1 m (em) c 8.5 0.4 13.3 0.7 Average forb cover a 3.8 0.7 5.1 0.9 Average grass cover ad 3.5 0.4 5.5 0.8 Average bare ground a 15.9 2.5 14.0 2.3 Average litter cover a 80.5 2.4 82.1 2.5 a Arcsin transformed for MANOV A, untransforrned mean and SE are reported b F = 43.36, P < 0.001 c F = 32.99, P < 0.001 d F = 5.86, P = 0.022 63 Table 2.3. Coefficient of determination (r) for habitat characteristics measured over a sampling gradient of initiation to hatch for 30 greater sage-grouse nest sites sampled in 2002 in northwest Colorado. . ~ ... .. .. Variable F-value P-value r2 Grass height at 0 m 22.70 < 0.001 0.281 Grass height at 1 m 24.32 < 0.001 0.295 Average grass cover 2.93 0.093 0.052 ( 64 CHAPTER 3. NOCTURNAL AND DIURNAL HABITAT SELECTION OF GREATER SAGE-GROUSE DURING BROOD-REARING IN NORTHWEST COLORADO INTRODUCTION Brood-rearing habitat An important feature of brood-rearing habitat of greater sage-grouse is cover adjacent to food in the form of insects and forbs (Klebenow 1972, Sveum et al. 1998). Insects dominate the diet of chicks in their first 2 weeks (Peterson 1970), with the main insects consumed being grasshoppers (Orthoptera), beetles (Coleoptera), and ants (Hymenoptera) (Knowlton and Thomley 1942, Klebenow and Gray 1968, Drut et al. 1994). As chicks develop, their diet shifts to forbs and succulent shrubs (Patterson 1952, Klebenow and Gray 1968, Klebenow 1969, Peterson 1970, Klebenow 1972). This may also coincide with habitat desiccation (Fischer et al. 1996a). Distinct summer habitat preferences are exhibited by some populations (Schoenberg 1982, Connelly et al. 2000). Early brood-rearing sites tend to be in big sagebrush habitat close to nest areas. Wakkinen (1990) noted a shift in late brood-rearing habitat to more. mesic sites as forbs began to desiccate. Fischer et al. (1996a) related this movement to vegetal moisture content of::; 60% water. Summer migration has been documented as either upwards to upland meadows (Savage 1969), or downward in elevation to riparian areas following a gradient of forbs (Klebenow 1969). 65 Night-Roost Habitat Despite extensive research into the habitat requirements of greater sage-grouse, summer night-roosting habitat has received no attention. Previous research has included winter night-roost sites (Beck 1977) and male night-roost preferences during the breeding season (Patterson 1952, Wallestad 1971, Klebenow 1972, Emmons and Braun 1979). Authors have speculated that during the spring and summer greater sage-grouse select open cover types in which to breed and feed (Klebenow 1972, Blus et al. 1989, Nelle et al. 2000) and dense habitat in order to roost. By late summer, grassland grouse populations abandon the tactic of inconspicuousness and join gang broods (Bergerud 1988). Patterson (1952) indicated that a single greater sage-grouse roost area in the fall and winter might encompass dozens of acres and involve hundreds of birds. There is no indication in the literature as to what role, if any, social structure plays on nocturnal habitat use. A more complete understanding of greater sage-grouse habitat use is imperative for reversing population declines. By considering temporal variation in habitat use patterns habitat management guidelines can promote and protect critical habitat needs. The objectives of this study were to (1) model brood productivity and brood and night-roost habitat selection and (2) compare nocturnal and diurnal habitat use during the brood-rearing period. METHODS Females with broods were located, at randomly selected time periods from dawn to dusk, via radio-telemetry 1-2 times per week. Additionally, individual females with and 66 without broods were located at 2: 1 hour before sunrise and 2: 1 hour after sunset from 4-1 0 weeks post-hatch. A count of individuals at each night roost location was made using binoculars and a spotlight (Wakkinen et al. 1992). At each diurnal location cover type, date, time, UTM coordinates, slope, aspect, distance to nearest visual obstruction and distance to nearest 2-track were recorded. An estimate of brood size was made 6 weeks post- hatch by flushing radio-marked females that had been successful at hatching a nest. Productivity was calculated by dividing the total number of chicks flushed at 6 weeks by the number of females alive and radio-marked at the start of the breeding season. A Van der Waerden test was used to assess variability in productivity among years, as the data did not fit a normal distribution (Conover 1999) Individuals located nocturnally were restricted to those within a 15-minute walk from the nearest road, thus distance to road was not included in our analysis as our sample design biased this variable. Each week, for 10 weeks post-hatch, 25o/o of radio-marked hens with broods were randomly chosen for diurnal vegetation sampling. I estimated the center of brood activity by locating the hen visually. Sites were marked and sampled within 3 days after the brood had left the vicinity. All vegetation measurements at brood use locations were the same as at nest sites (Chapter 1) with the following exceptions: 1) grass and forb height were quantified by measuring the nearest individual within 1 m of plot center, and within 1 m of each transect at 2.5, 5.0, 7.5 and 10m, 2) a single 40 x 50 em microplot was placed at the point where the transects met and at 2.5, 5.0, 7.5, and 10m along each transect, and 3) no cover board was used at the transect intersection. Random locations were selected from northing and easting UTM coordinates falling within the home range of the population and were measured within 1 week of vegetation , 67 . sampling at the brood use area using identical techniques. Random sites occurring in shrubsteppe, mountain shrub, CRP, riparian, wet meadow, and agricultural cover types were accepted as potential brood-rearing locations. There were 173 plant species identified at brood, night roost and random sites over the 2 years of study (Appendix 1-A). Thus, analyzing cover by individual plant species was not feasible compared to the number of plots sampled. Instead, I summarized individual plant species by their frequency of occurrence within each plot and ranked them by their abundance within brood, night-roost and random locations. Flock size A correlation matrix of structural habitat characteristics was examined to control for multicollinearity. Visual obstruction readings read at a height of 1.5 m were correlated (/ > 0.80, P < 0.001) to those read at 0.5 m and therefore removed from flock size analysis. Additionally 0-0.5 m and 0.5-1.0 m demarcations of visual obstruction read from a height of 0.5 m and a distance of 5 m were correlated thus the visual obstruction reading from 0.5-1.0 m was also removed from further analysis. Forward stepwise multiple regression analysis was used to determine the model that would best predict flock size with the fewest structural habitat characteristics (Ott and Longnecker 2001). A significance level of P ~ 0.10 was used to determine which variables entered and remained in the model. Variables considered were shrub, grass and forb height, visual obstruction readings from 0-0.5 m and read from 2.5, 5 and 10m., and visual obstruction readings from 0.5-1.0 m and read from 5 and 10m. Vegetation Sampling I partitioned the variables of slope, aspect and distance to nearest structure into the same categories as described in Chapter I. Elevation received an additional categorization; 68 low(< 2,020 m), medium (2,021-2,140 m), medium- high (2,141-2,160 m) and high(> 2,160 m). Prior to examination, variables were assessed for correlation strength (Ott and Longnecker 2001). I used chi-squared tests to determine whether there were macrohabitat differences between years at brood sites. Chi-squared tests were also conducted between random and brood-use sites, random and night roost locations, and diurnal and nocturnal sites. The mean of each vegetation variable over the plot was used in all calculations. Means of cover estimates were obtained by using the mid-point value of each cover class. Variables were assessed for normality and arcsin transformed when necessary. I used a multivariate analysis of variance (MANOV A) to assess whether microhabitat variables differed at random locations between years (P ~ 0.05) (Johnson and Wichern 2002). Those that differed were assessed for correlation strength. A principal component analysis was conducted on the characteristics of the random locations that did not differ between years (Johnson and Wichern 2002). Brood and night-roost sites were scored with the principal components generated from the random plots. Forward step-wise logistic regressions included the principle components, habitat variables that differed between years and the class variable year. These were used to test the importance of the habitat variables in predicting 1 of2 binary variables; brood site or random, night-roost or random and diurnal or nocturnal use ·site. A significance level of P ~ 0.10 was used to determine which variables entered the model. This analysis was performed in SAS 8.2 (SAS Institute 1999). Early/Late Brood-rearing Brood habitat has typically been separated into that used for early and late broodrearing (Drut et al. 1994). I further defined this split as the date when 90% of radio-marked 69 hens changed cover type to either mountain shrub or riparian/wet-meadow communities and/or had moved up or down in elevation. During 2001, early brood-rearing appeared to end 7 July and late brood-rearing terminated at the end of data collection in August. In 2002, early brood-rearing ended by 14 July and late brood-rearing continued through August when field work was completed. I used a correlation analysis to reduce the dimensionality of the data set and then a MANOVA to assess whether habitat differences occurred between early and late brood-rearing time-periods. Analysis of canonical variables and univariate ANOVA was used to determine which variables contributed to the differences. RESULTS Brood Size Brood sizes at 6 weeks ranged from 0 to 8 chicks. In 2001 the mean number of chicks per female that successfully hatched a nest was 3.5, SE = 0.4, n = 26, which did not differ (t = 1.9, P = 0.064) from the mean number of chicks per female, 2.4, SE = 0.4, n = 29 in 2002. Pooling the data over both years resulted in an estimate of 3.0 chicks/successful female at 6 weeks of age. Mean productivity in 2001 (1.5 chicks/female alive at the beginning of nesting, SE = 0.3, n = 60) did not differ (Z = 8.05, P = 0.214) from 2002 (1.0 chick/female, SE = 0.2, n = 73). Flock Size Flock sizes at nocturnal use sites ranged from 1 to 30 individuals, with mean flock size at 6.0 individuals (SE = 0.82, n = 58). Forward stepwise multiple regression analysis produced a model with visual obstruction from 0-0.5 m, read from a height of 0.5 m and a distance of 5 m and visual obstruction reading from 0.5-1.0 m, read from a height of 0.5 m 70 and a distance of 10m, as the only significant predictors of flock size at night-roosts (Table 3.1). Flock size declined with increasing visual obstruction when read close to the flock (5 m) at ground level (0-0.5 m) and increased with increasing visual obstruction from the furthest reading distance (10m) and between the heights of 0.5-1.0 m. Vegetation Sampling M aero habitat. Proportional use of aspect and elevation differed between years at brood-use sites (X = 15.43, 4 df, P = 0.004, X= 12.90, 3 df, P = 0.005), respectively. The difference in aspect was due to a higher proportion of locations falling in the 0° slope category in 2001 (25%) than 2002 (7%). For elevation the difference resulted from higher proportional use of the low elevation category (50 vs. 35%) in 2001 than in 2002. The means, however, between years for these variables fell within the same category (aspect: 136-187°, elevation: 20212140 m). These characteristics were not pooled for further analysis. In 2001, greater sagegrouse broods were found at a greater proportion within the high elevation category (17 vs. 2%) and less in the medium-high category (9 vs. 26%) (X= 14.74, 3 df, P = 0.002) compared to random (Table 3.2). Additionally, relative brood use tended to be less (10 vs. 41 %) within aspects ranging from 136-314° (X= 20.13, 4 df, P = 0.001) than random. In 2002, mean elevation at greater sage-grouse brood and random sites was 2,117, SE = 23, n = 92 and 1,978, SE = 17, n = 93, respectively and proportional use did not differ (X= 6.58, 3 df, P = 0.373). Relative use of aspect at brood sites did not differ from random (X= 4.25, 4 df, P = 0.097). Combining years, no relationship was detected between proportional brood and random site use with respect to distance to nearest visible structure (X= 0.30, 2 df, P = l 71 0.862), distance to nearest two-track (X= 4.38, 2 df, P = 0.112) or mean slope (X= 5.13, 2 df, P = 0.077) (Table 3.2). No differences were detected among relative use of physical macrohabitat characteristics at night-roost sites between years and so the data were pooled (P > 0.341). Additionally, no relationship was found with regards to relative use of slope (X= 3.39, 2 df, P = 0.183) and distance to structure (X= 0.14, 2 df, P = 0.930) for night-roost and random locations (Table 3.2). A relationship was found with regards to relative use of aspect (X= 14.27, 4 df, P = 0.007), and elevation (X= 29.21, 3 df, P < 0.005) for greater sage-grouse I l l night-roost locations. Greater sage-grouse night-roost locations were found at a greater proportion within the northerly aspect category (43 vs. 19%) and within the medium-high (38 vs. 16o/o) elevation category than random sites. Comparison of physical macrohabitat characteristics among diurnal and nocturnal brood sites revealed no differences of relative u~e of slope (X= 0.12, 2 df, P = 0.941) or distance to nearest visible structure (X= 0.01, 2 df, P = 0.994) (Table 3.2). In 2001, no difference was detected with proportional use of aspect (X= 6.20, 4 df, P = 0.185), however, night-roost locations tended to fall within the medium-high elevation category to a greater extent (39 vs. 9%) than diurnal locations (X= 11.76, 3 df, P = 0.008). No relationships between diurnal and nocturnal locations were detected for relative use of aspect or elevation in 2002 (X= 5.60, 4 df, P = 0.231; X= 0.07, 3 df, P = 0.996) (Table 3.2). Microhabitat. Microhabitat characteristics differed at random sites (F = 5.65, P < 0.005) between years. Vegetation characteristics that varied between years at random locations were species richness, percent cover offorbs, exotics and bare ground, visual obstruction from 0-0.5 m 72 measured from a reading location of 5 m and a reading height of 0.5 m, visual obstruction from 0-0.5 m measured from a reading location of 10m and a reading height of 0.5 m, and visual obstruction from 0.5- 1.0 m measured from a reading location of 2.5 m and a reading height of0.5 m (Table 3.3). Visual obstruction from 0-0.5 m read from reading height of0.5 mat 10m and visual obstruction 0.5-1.0 m read from a reading height of0.5 m and a distance of2.5 m were correlated (r2 = 0.83, P < 0.005, n = 94), thus the latter reading was removed from further analysis. Additionally, species richness and percent forb cover were correlated (r2= 0.68, P < 0.001, n = 94). Species richness was removed from further analysis. I conducted a PCA on the 15 habitat characteristics of 94 random sites, minus the 5 variables; percent cover of forbs, exotics and bare ground, and visual obstruction from 0-0.5 m, read from a height of 0.5 m and from the distances 5 and 10m which contributed to the significant difference between years. This produced a description of the structural habitat gradients within the study area. The PCA defined 4 principal components, accounting for 78% of the overall variance. Each of the principal components considered had an Eigenvalue > 1 (Table 3.4). The 4 principal components each described an ecologically interpretable gradient. The first component, described a complex gradient of structure with the scale ranging from herbaceous cover to tall shrubs, high proportion of shrub cover including sagebrush and high visual obstruction from 0-1 m. This addressed the overall physiognomy of the sage-steppe community. The second and third components related to the shrub-steppe under-story community. The second component varied along a gradient of herbaceous cover from a high percent grass cover at the positive end of the gradient, with grass and forb height and percent litter at the negative end of the gradient. The third principal component was defined by grass I~ l k 1,.- 73 height on the positive end of the gradient and visual obstruction at ground level(< 50 em) on the negative end. The fourth component defmed community complexity with high shrub cover, sagebrush cover and grass height on the positive end of the gradient and visual obscurity from 0-1 m read from 2.5 mat a height of 1.5 m falling on the negative end of the gradient (Table 3 .4). Ten variables entered all3 forward step-wise logistic regressions. These included year, principle components 1-4, and the variables that differed between years; percent forb, exotic and bare ground cover, and 2 visual obstruction readings, VISOBa12 and VISOBa13. Logistic regression found percent forb cover, percent bare ground, and visual obstruction readings from 0-0.5 m, a reading height of 0.5 m and from 5 and 10m, to be significant predictors ofbrood occurrence (Table 3.5). The logistic regression equation that resulted was: Logit (Y)= -0.9 + 31.7 (PCTFORB)- 3.7 (BRGROUND) -0.8 (VISOBa12)- 0.7 (VISOBa13) + e (Table 3.5). The logit transformation allows one to describe the relationship between the binary response variable and multiple explanatory variables and the e represents error in the model (Ott and Longnecker 2001). Percent forb cover and visual obstruction readings from 0-0.5 m, a reading height of0.5 m and from 5 and 10m were all greater at brood sites than at random (Table 3.6). Mean bare ground cover was the only variable less at brood sites than at random sites. Logistic regression was also used to predict night-roost selection. The habitat variables that entered the model as significant predictors of night-roost locations were percent forb and bare ground cover, VISOBa12 and VISOBa13 (Table 3.5) .. The logistic regression equation that resulted was: Logit (Y) = 0.7 + 14.5 (PCTFORB)- 2.9 (BRGROUND)- 0.9 (VISOBa12)- 0.4 (VISOBa13) + e (Table 3.5). Percent forb cover was l 74 greater at night-roost locations than at random locations, while bare ground, VISOBa12 and VISOBa13 values were all less at night-roost locations than at random locations (Table 3.6). Finally, logistic regression was used to predict nocturnal versus diurnal occurrence. The habitat variables that entered the model as significant predictors of night-roost locations were PC 1, percent cover of exotics, and VISOBa12 (Table 3.5). The logistic regression equation that resulted was: Logit (Y) = 1.5- 0.7 (PC 1)- 2.2 (EXOTICS)- 1.5 (VISOBa12) + e (Table 3.5). PC 1 involved percent shrub cover, sagebrush cover, shrub height, and all the visual obstruction readings (Table 3.4). All the values within PC 1, percent cover of exotics and visual obstruction 0-0.5m read from a height of 0.5 m and from a distance of 5 m were greater at diurnal sites than at nocturnal sites (Table 3.6). I l Individual Plant Frequency Because the MANOV A indicated that there were vegetation differences between I years, particularly among forb cover (Table 3.3), I analyzed plant frequency separately for 2001 and 2002. In 2001, the most abundant shrub species at brood sites was sagebrush (Figure 3.1). Forbs found in> 60% of plots included pale madwort, maiden blue-eyed Mary, longleaf phlox, onion (Allium spp.), clover and common salsify (Tragapogon dubius) (Figure 3.1). Grasses that occurred in> 60% ofbrood sites included western wheatgrass, cheat grass and Sandberg bluegrass (Poa secunda). The same species were abundant at random sites with some exceptions. Dead sagebrush and copper mallow occurred at > 60% of random locations and common salsify only occurred at 30% of random locations (Figure 3.1 ). The most abundant shrub species at night-roost sites was dead sagebrush (Figure 3.2). Nocturnal sites had less sagebrush occurring than random. Forbs found in> 60% of nightroost locations include pale madwort, buttercup (Ranunculus muricatus), longleaf phlox, 75 maiden blue-eyed Mary and common salsify. Grasses at> 60o/o of night-roost locations included Sandberg bluegrass, western wheatgrass and cheat grass. Common salsify occurred at 64% of night-roost plots and only 31% of random locations (Figure 3.2). In 2002, the most abundant shrub species at brood sites were sagebrush and dead 1.,- sagebrush (Figure 3.3). Forbs found in> 60% of plots included only longleaf phlox and maiden blue eyed Mary. Abundant grasses included western wheatgrass and Sandberg bluegrass. In 2002, the majority of the most abundant species of forbs (maiden blue-eyed Mary, common salsify, lupine) occurred more frequently at brood sites than random indicating that hens were associating with these forbs for brood-rearing (Figure 3.3). The most abundant shrub species at night-roost locations in 2002 was dead sagebrush (Figure 3.4). Forbs and grasses at> 60% of night-roost locations included pale madwort, longleaf phlox, western wheat grass, cheat grass, and Poa spp. Pale madwort, sagebrush and globe mallow all occurred more frequently at random locations than night-roost locations in 2002. Early/Late Brood-rearing Early brood-rearing locations were primarily in shrub-steppe (58%). Broods were also found in mountain shrub (19%), burnt shrub-steppe communities (13%), CRP (8%) and riparian/ wet meadows (2%). Late brood-rearing sites were found with increased frequency in mountain shrub (37%) and riparian (11 %) and reduced frequency in shrub-steppe (37o/o), burnt shrub-steppe (10%) and CRP (4%). The MANOVA on random plots indicated a difference among habitat variables between years (Table 3.3), thus timing was assessed separately for 2001 and 2002. Analysis of early and late brood-rearing indicated a difference in 2001 (Wilk's 'A= 0.47; F= 1.9; 22,36 df; P = 0.047) but not in 2002 (Wilk's 'A= 0.29; F= l 76 1.12; 22,10 df; P = 0.445). The habitat variables contributing to the differences in early and late brood-rearing in 2001 were visual obstruction reading 0.5-1.0 m and a reading height of 1.5 m and from a distance of 10m, visual obstruction readings 0-0.5 m and read from a reading height of 1.5 m and a distance of 2.5 and 10 m and percent forb cover and percent sagebrush cover (Table 3.7). Forb cover was greater at late-brood rearing sites while sagebrush cover was reduced. Visual obstruction decreased at the ground level but increased from 0.5- 1.0 m (Table 3.7). ,. DISCUSSION Brood Size The use of flush counts as an estimate of brood size is inherently biased. Estimates / can be either too low if chicks fail to flush or too high if females adopt chicks or broods congregate (N. Burkepile, personal communication). I used flush counts at 6 weeks of age to reduce these biases. An assumption of flush counts is that these biases are consistent between years and locations; therefore brood counts can be used as an index of trend. Mean brood size at 6 weeks post-hatch in 2001 (3.5) was greater than that in 2002 (2.4) by> 1 chick. Although not significant, this difference may have resulted from less forage and cover due to the hot, dry conditions in 2002. Estimates of brood productivity from harvest data collection on Cold Spring Mountain, Blue Mountain and eastern Moffat County, Colorado from 1976-1998 ranged from 0.2-5.0 chicks/female with a mean of2.0 chicks/female (CDOW, unpublished data). Mean productivity from 1976-1998 from North Park, Jackson County was 1.5 chicks/female (CDOW, unpublished data). Productivity of2.25 chicks/female has been considered necessary to maintain a stable or increasing population 77 (Connelly et al. 2000). However, data in North Park shows chicks/female< 2 from 19852000, a period when the population size increased. Estimates of brood productivity in the Axial Basin and Danforth Hills (1.5 and 1.0 chicks/female in 2001 and 2002) were below the management recommendations and the means reported in other populations in Colorado. Population trends in Moffat County indicated a decline in population abundance although nest success and adult survival do not appear to be limiting factors (Chapter I, Chapter IV). Although brood counts are likely biased low, these productivity estimates warrant further research into chick survival which may be limiting the population. Macrohabitat Differences in relative use of macrohabitat characteristics between brood and random sites were found in 2001 as a greater proportion of brood use-sites were in the high elevation category(> 2,160 m) and less in the medium-high elevation category (2, 141- 2,160 m) compared to random. Additionally, relative brood use was lower in the southerly and westerly aspect categories (136-314°) than random. These results may have been from a greater number of females raising broods in the Danforth Hills than at random from within the study site. Typically females with broods move upwards in elevation during the late brood-rearing period, following a gradient of forbs (Fischer et al. 1996a). The Danforth Hills, which provided much of the late brood-rearing habitat, are higher in elevation, steeper in gradient, and are primarily northerly in aspect. I ~ The differences occurring among relative use of physi~al macrohabitat characteristics between night-roost locations and random sites were with respect to aspect and elevation; The r.easons for these differences are likely due to our sampling protocol. Locations were only obtained for grouse that were within a 15 minute walk from the nearest two-track. 78 Many night-roost locations were restricted to the Danforth Hills (37 vs. 21), which had a higher concentration of roads than that of the Axial Basin. Similarly, data for 2001 revealed that nocturnal locations were at higher elevations in a greater proportion than at diurnal locations. Again, I sampled more night-roost (71 %) than brood locations (44%) in the Danforth Hills than in the Axial Basin. Macro habitat characteristics were basically used in proportion to random. These results may be due to the fact there was little habitat diversity within the physiographic characteristics at my study sites (Porter and Church 1987, Aberg et al. 2000). Despite the apparent homogeneity of the shrub-steppe community, selection was apparent at the microhabitat scale. ·Brood-site selection Habitat variables that were significant predictors ofbrood site selection were percent cover of forbs, bare ground, and visual obstruction readings from 0-0.5 m. Greater sagegrouse females selected brood-rearing sites with higher forb cover and less bare ground than at random locations. Indeed, there was twice as much forb cover at brood sites (8%), than at random locations (4%). These cover values are based on the means of midpoints and would all fit into the vegetation cover category of 3-9%. Previous studies reported broods selecting areas related to forb cover (Klebenow 1969, Peterson 1970, Autenreith 1981, Drut et al. 1994, Apa 1998, Sveum et al. 1998). My forb cover values were similar to those reported by Dunn and Braun (1986) in Moffat County (5%), Schoenberg (1982) in North Park Colorado (7%) and Apa (1998) in southeastern Idaho (8o/o). However, other rese,archers have reported forb cover values at brood sites as greater than these; 17-27% in central Montana (Wallestad 1971) and 10-25% in Oregon and Washington (Drut et al. 1994, Sveum et al. 1998). Forbs 79 have greater protein potential than other vegetation types (Peterson 1970) and have been associated with invertebrate biomass in sagebrush cover-types (Jamieson et al. 2002). Johnson and Boyce (1990) showed experimentally that insufficient insects in the diet-of chicks resulted in deaths from malnutrition. Similarly, arthropod abundance within the first 10 days of hatching was directly correlated to survival in partridge chicks (Alectoris rufa and Perdix perdix) (Green 1984). Greater sage-grouse females are likely selecting sites with greater forb cover in order to provide forage in the form of invertebrates and forbs for their broods. Less bare ground at brood sites than at random sites supports this hypothesis. Preference for individual plant species was more apparent in 2002, possibly due to drought conditions. Forbs and grasses such as maiden blue-eyed Mary, onion, Sandberg bluegrass, common salsify and clover were common in 2001 but scarce in 2002. The selection for areas of greater forb cover was apparent in 2002 as forb species occurred less frequently at random locations than brood-use sites. Common salsify has been reported as a major forage item for greater-sage grouse and occurred with twice the frequency at use locations than at random. Globe mallow and cheat grass were not as prevalent at brood sites than at random sites, as these plants indicate dry, open habitat types (Kershaw et al. 1998). The major forbs known to be important in greater sage-grouse diets include common dandelion (Taraxacum officina/e), milkvetch (Astragalus spp.), sego lily (Calochortus spp.), yarrow (Achillea millefolium), prickly lettuce (Lactuca spp.), lupine, and hawksbeard (Crepis spp.) (Fischer et al. 1996b). Of these, common dandelion, lupine, and yarrow occurred more frequently at brood sites than random, suggesting selection for these forage items. 80 Early/Late Brood-rearing In 2001 there were differences in early and late brood-rearing with respect to forb cover, sagebrush cover and visual obstruction readings. Females with broods moved to areas with greater forb cover, less visual obstruction at ground level and greater visual obstruction from 0.5-1..0 m. Increased use of the mountain shrub communities in the late brood-rearing period supports this trend of greater forb cover, less ground level visual obstruction but greater vertical visual obstruction. Mean sagebrush canopy cover at brood sites in the study area was 11%, however it decreased from 13% to 8% as hens moved into mountain shrub communities from early to late brood-rearing. In 2002, no differences were observed between early and late-brood rearing. Sample size for the late brood-rearing period was small, thus results should be viewed cautiously. Due to the drought in 2002, there was not an abrupt change in habitat, but rather a continuous selection for areas with greater moisture and thus more forbs and insects. Night-roost Habitat Selection Nocturnal locations (n =58) were in burnt shrub-steppe (52%), shrub-steppe (38%), grassy meadows (7%), CRP (2%) and riparian (2%) cover types. My results indicated that greater sage-grouse females choose night-roost sites with less bare ground and visual obstruction, but greater forb cover than that at random sites. Night-roost habitat information for greater sage-grouse is limited primarily to males during the breeding season. Patterson (1952), Wallestad (1971) and Klebenow (1972) noted that roosts near leks chosen by males and non-nesting females contained the heaviest and densest sagebrush. In contrast, Emmons and Braun (1979) found 90% of male sage-grouse chose nocturnal roosts directly on leks in order to maintain their strutting positions. Canopy cover of these roost sites was only 8. 7o/o 81 and shrub height averaged 18.7 em. Shrub canopy cover of 9.5% and mean shrub height of 30.5 em characterized nocturnal sites chosen by females in the Axial Basin and Danforth Hills. Night-roost habitat characteristics existing for other ground-roosting upland game birds support our results. Northern bobwhite quail (Colinus virginianus) preferred open canopy, short herbaceous cover (30-90 em) and limited litter at night-roost sites (Klimstra and Ziccardi 1963, Barnes et al. 1995), while ring-necked pheasants (Phasianus colchicus) chose herbaceous cover over sagebrush for winter night-roosts in southern Idaho (Leptich 1992). Scaled quail (Callipepla squamata) selected nocturnal areas with low shrub cover (Stormer 1984) and Montezuma quail (Cyrtonyx montezumae) selected grassy slopes within • an oak savanna habitat (Stromberg 1990) . Wind significantly affects greater sage-grouse metabolism (Sherfy and Perkins 1995) and, in instances where greater sage-grouse roosted in shrub-steppe cover type, the birds seemed to be there as a result of inclement weather. The shrub-steppe microclimate effectively reduces exposure to wind (Sherfy and Perkins 1995). Bobwhite quail also exhibited shifts in night-roost locations to more protective cover associated with wind (Klimstra and Ziccardi 1963). Two potential explanations exist for the selection of open cover types for nightroosting during the summer. First, night-roost locations have higher forb cover than random locations and perhaps provide forage opportunities in the crepuscular periods. Savage (1969) found broods fed in early mornings and evenings and loafed in sagebrush midday. Dunn and Braun (1986) found that greater sage-grouse broods fed in open cover types in the morning while Sveum et al. (1998) reported 52% of afternoon locations occurred in grassland cover types. Potentially, females roost in an open cover type in order to exploit the area for l 82 foraging in the crepuscular periods, particularly for broods, which exhibit high energetic ~ I I • demands. Secondly, open cover types may be chosen in order to provide grouse with greater escape potential from predators. Their large body size and short legs make it laborious for them' to initiate flight (Patterson 1952). Open habitat not only provides greater potential of detecting approaching mammalian predators but also minimal interference when flushed, thus facilitating rapid escape. Selection for individual forb species at nocturnal use sites may support the hypothesis that open habitat types are selected for feeding. Common salsify has been reported as a major forage item for greater sage-grouse, and occurred with twice the frequency at use locations than at random. Other major forb species, common dandelion, lupine, and yarrow, ·occurred more frequently at brood sites than random, suggesting selection for these forage ·items. The species that was most strongly avoided at night-roost locations was big sagebrush, confirming the hypothesis that greater sage-grouse select open habitat types in order to night-roost during the summer. Flock Size · The formation of flocks is thought to decrease vulnerability, as there are effectively more individuals to detect predators (Reynolds et al. 1988). Individual greater sage-grouse roosting sites are in close proximity to other birds during the breeding season, fall and winter (Patterson 1952). With a mean of 6 individuals per roost sites my results indicate that communal roosting is also practiced during brood-rearing. This could be when and where brood-mixing occurs. Bergerud (1988) suggested communal roosting occurred in late · summer for open-dwelling grouse populations. Patterson (1952) speculated this behavior served as a means of avoiding night attack by natural enemies. Potential night predators 83 included red foxes (Vulpes vulpes), coyotes (Canis latrans), and great homed owls (Bubo r· .. virginianus). Of the variables examined, visual obstruction readings were the factors related to flock size; flock size decreased with increased visual obstruction read close to the flock (2.5 m) at ground level (0-0.5 m) and increased with increased visual obstruction from the furthest reading distance (10m) and between the heights of 0.5-1.0 m. This indicates greater sage-grouse required either ground level cover at night-roost locations or larger flock size for increased vigilance from mammalian predators. Increased cover above ground level decreases flock size perhaps indicating that nocturnal avian predation is not a factor • influencing vigilance. Examination of carcasses retrieved within 5 days (mean = 2.4 days, n = 17) of death supports this hypothesis because I attributed 73o/o (n = 8) of known mortalities to mammalian predation and 9 % (n = 1) of known mortalities to great homed owls (Chapter IV). Nocturnal Versus Diurnal Habitat Selection Research on greater sage-grouse nocturnal roost sites in the winter has produced contradictory results. Beck (1977) followed greater sage-grouse to their nocturnal sites on their winter range in Jackson County, Colorado, and found that nocturnal roosting sites had the same vegetal and physical characteristics as feeding sites. However, Dalke et al. (1963) suggested that daytime loafing and feeding cover types were not the same as roosting sites. Authors have speculated that during the spring and summer greater sage-grouse select open cover types in which to breed and feed and seek out denser habitat in order to roost (Klebenow 1972, Blus et al. 1989, Nelle et al. 2000). My results indicated that night-roost locations were different from daily-use sites during brood-rearing in terms of shrub cover. 84 > However, contradicting previous speculation, females with broods selected less dense habitat in which to night roost. Mean shrub cover and shrub height at night roost locations was less dense (9% vs. 22%) and shorter (31 vs. 58 em) than at diurnal brood-sites. Sagebrush cover was also less dense (4% vs. 11 %) at nocturnal use sites. Additionally, the visual obstruction readings were all greater at diurnal sites (Table 3.4). Patterson (1952) noted a tendency for · sage-grouse to select taller sagebrush in draws and gullies in order to roost in the daytime, in • contrast to the smaller and more open sagebrush areas selected at night. MANAGEMENT IMPLICATIONS Habitat management guidelines recommend> 10% forb cover for brood-rearing (Connelly et al. 2000). Forb cover of 8%, which falls into the cover category of 3-9%, was selected by females for brood-rearing in the Axial Basin and Danforth Hills and falls short of this recommendation. However, the influence of this on survival of chicks is unknown. Changes in availability of forbs greatly influence grouse distribution and habitat selection. In Oregon, greater sage-grouse reportedly had larger home range sizes in areas having low forb abundance (Drut et al. 1994). The affinity of greater sage-grouse broods for areas of greater forb cover offers management opportunities. Practices such as seeding with exotic grasses, • herbicide treatments and widespread burning should be avoided. Greater sage-grouse broods are more reliant on invertebrate foods than adult birds (W allestad et al. 197 5). Management • practices that improve habitat for broods may be extremely beneficial as low brood productivity may be a factor limiting this greater sage-grouse population. Despite extensive research into the habitat requirements of greater sage-grouse, nightroosting habitat received little attention. Current habitat management guidelines apply only 85 to diurnal habitat needs and should be viewed cautiously (Beyer and Haufler 1994). I recommend that managers consider night-roosting habitat characteristics of broods a critical component of greater sage-grouse habitat requirements. Present literature on diurnal habitat use by broods does not recognize the importance of openings in the shrub-steppe community specifically for night roosting. Habitat guidelines recommend sagebrush cover ranging between 10-25% (Connelly et al. 2000). Fischer et al. (1996b) concluded that the short-term effect of fire does not serve to enhance brood habitat in southeastern Idaho. However, their results are based upon sampling procedures that occurred diurnally. Therefore, they may have missed greater sage-grouse nocturnal use of burnt areas. My results indicate greater sage-grouse seek out openings with < 5% sagebrush cover and < 10% shrub cover such as recently burnt shrub-steppe ·communities in order to roost. Therefore, I would agree with Klebenow (1972) and Gates (1983) who suggested a mosaic of burnt areas may benefit greater sage-grouse brood habitat. My results indicated the importance of a range of sagebrush canopy cover from 5 -12% for brood-rearing. More research is required to estimate the size of openings in sagebrush and • • • .. their frequency across the landscape . 86 ,. LITERATURE CITED ~·. r • , I ,. Aberg, J., G. Jansson, J.E. Swenson, and G. Mikusinski. 2000. Difficulties in detecting habitat selection by animals in generally suitable areas. Wildlife Biology 6:89-99. Apa, A.D. 1998. Habitat use and movements of sympatric sage and Columbian sharp-tailed grouse in southeastern Idaho. Ph.D. Dissertation, University of Idaho, Moscow, Idaho. Autenrieth, R.E. 1981. Sage grouse management in Idaho. Wildlife Bulletin 9. Idaho Department of Fish and Game, Boise, Idaho. 1 .. Barnes, G.B., L.A. Madison, J.D. Sole, and M.J. Lacki. 1995. An assessment of habitat quality for northern bobwhite quail in tall fescue-dominated fields. Wildlife Society Bulletin 23 :231-23 7. .. Beck, T.D.I. 1977. Sage grouse flock characteristics and habitat selection in winter. Journal ofWildlife Management 41:18-26. Bergerud, A.T. 1988. Population ecology ofNorth American grouse. Pages 578-648 in A.T. Bergurud and M.W. Gratson, eds. Adaptive strategies and population ecology of northern grouse. University of Minnesota Press, Minneapolis. Beyer, D.E., and J.B. Haufler. 1994. Diurnal versus 24-hour sampling ofhabitat use. Journal ofWildlife Management 58:178-180. ... Blus, L.J., C.S. Staley, C.J. Renny, G.W. Pendleton, T.H. Craig, E.H. Craig, and D.K. Halford. 1989. Effects of organophosphorus insecticides on sage grouse in southeastern Idaho. Journal of Wildlife Management 53:1139-1146 . Connelly, J.W., M.A. Schroeder, A.R. Sands, and C.E. Braun. 2000. Guidelines to manage sage grouse populations and their habitats. Wildlife Society Bulletin 28:967-985. Conover, W.J. 1999. Practical nonparametric statistics, Third Edition. John Wiley and Sons, Inc., New York. 584 pp. . Dalke, P.D., D.B. Pyrah, D.C. Stanton, J.E. Crawford, and E.F. Schlatterer. 1963. Ecology, productivity, and management of sage grouse in Idaho. Journal of Wildlife Management 27:811-841. Drut, M.S., J.A. Crawford, and M.A. Gregg. 1994. Brood habitat use by sage grouse in Oregon. Great Basin Naturalist 54:170-176. Dunn, P.O., and C.E. Braun. 1986. Summer habitat use by adult female and juvenile sage grouse. Journal of Wildlife Management 50:228-235. 87 Emmons, S.R. and C.E. Braun. 1979. Evaluation of the effects of changes in hunting regulations on sage grouse populations: evaluation of censuses of males. Colorado Game and Fish. Project W-37-R-35. Work Plan 3. Job 9a. Job Completion report. Fischer, R.A, K.P. Reese, and J.W. Connelly. 1996a. Influence of vegetal moisture content and nest fate on timing of female sage grouse migration. Condor 98:868-872. Fischer, R.A, K.P. Reese, and J.W. Connelly. 1996b. An investigation on fire effects within xeric sage grouse habitat. Journal of Range Management 49:194-198. Gates, J.G. 1993. Sage grouse, lagomorph and pronghorn use of sage-brush-grassland burn site on the Idaho National Engineering Laboratory. M.S. Thesis, Montana State University. Bozeman, Montana. Green, R.E. 1984. The feeding ecology and survival of partridge chicks (Alectoris rufa and Perdix perdix) on arable farmland in East Anglia. Journal of Applied Ecology 21:817830. Jamison, B.E., R.J. Robel, J.S. Pontius, and R.D. Applegate. 2002. Invertebrate biomass: associations with lesser prairie-chicken habitat use and sand sagebrush density in southwestern Kansas. Wildlife Society Bulletin 30:517-526. Johnson, G.D., and M.S. Boyce. 1990. Feeding trials with insects in the diet of sage grouse chicks. Journal of Wildlife Management 54:88-91. Johnson, R.A., and D.W. Wichern. 2002. Applied multivariate statistical analysis, fifth edition. Prentice-Hall Inc., Upper Saddle River, NJ. 767pp. Kershaw, L., A. MacKinnon, and J. Pojar. 1998. Plants of the Rocky Mountains. Lone Pine Publishing, Edmonton. 384pp. Klebenow, D.A. 1969. Sage grouse nesting and brood habitat in Idaho. Journal of Wildlife Management 33:649-662. ... Klebenow, D.A. 1972. The habitat requirements of sage grouse and the role of fire in management. Proceedings from Tall Timbers Fire Ecology Conference 12:305-315 . Klebenow, D.A., and G.M. Gray. 1968. Food habits of juvenile sage grouse. Journal of Range Management 21:80-83. Klimstra, W.D., and V.C. Ziccardi. 1963. Night roosting habitat ofbobwhites. Journal of Wildlife Management 27:202-214. Knowlton, G.F., and H. F. Thomley. 1942. Insect food of sage grouse. Journal ofEconomic Entomology 35:107-108. 88 Leptich, D.J. 1992. Winter habitat use by hen pheasants in southern Idaho. Journal of Wildlife Management 55:376-380. Nelle, P.J, K.P. Reese, and J.W. Connelly. 2000. Long-term effects of fire on sage grouse habitat. Journal of Range Management 53:586-59. Ott, R.L., and M. Longnecker. 2001. An introduction to statistical methods and data analysis, fifth edition. Duxbury, Pacific Grove, CA. 1152pp. Patterson, R.L. 1952. The sage grouse in Wyoming. Sage Books, Inc., Denver, CO. 341pp. Peterson, J.G. 1970. The food habits and summer distribution of juvenile sage grouse in central Montana. Journal of Wildlife Management 34:147-155. Porter, W.F., and K.E. Church. 1987. Effects of environmental pattern on habitat preference analysis. Journal of Wildlife Management 51:681-685. ,. Reynolds, C.J., P. Angelstam, and S. Redpath. 1988. Predators, their ecology and impact on gamebird populations. Pages 72-97 in P.J. Hudson and M.R.W. Rands (eds.), Ecology and Management of Gamebirds. BSP Professional Books, Oxford, England. 263 pp. I~ I SAS Institute Inc. 1999. Version 8.2. SAS Institute, Carey, North Carolina, USA. Savage, D.E. 1969. The relationship of sage grouse to upland meadows in Nevada. M.S. thesis. University of Nevada, Reno. 101 pp. Schoenberg, T.L. 1982. Potential impacts of strip mining on sage grouse movement. Colorado Game and Fish. Project W-37-R-35. Work Plan 3. Job 12. Job Completion Report. 59-142. Sherfy, M.H., and P.J. Perkins. 1995. Influence of wind speed on sage grouse metabolism. Canadian Journal of Zoology 73:749-754. Stormer, F.A. · 1984. Night-roosting habitat of scaled quail. Journal of Wildlife Management 48: 191-197. Stromberg, M.R. 1990. Habitat, movements and roost characteristics of Montezuma quail in southeastern Arizona. Condor 92:229-236. Sveum, C.M, J.A Crawford, and W.D. Edge. 1998. Use and selection ofbrood-rearing habitat by sage grouse in south central Washington. Great Basin Naturalist 58:344-351. .. Wakkinen, W.L. 1990. Nest site characteristics and spring-summer movements of migratory sage grouse in southeastern Idaho. M.S. Thesis. University of Idaho, Moscow, Idaho. ~· 89 Wakkinen, W.L., K.P. Reese, J.W. Connelly, and R.A Fischer. 1992. An improved spotlighting technique for capturing sage grouse. Wildlife Society Bulletin 20:425-426. Wallestad, R.O. 1971. Summer movements and habitat use by sage grouse broods in central Montana. Journal ofWildlife Management 35:129-136. Wallestad, R.O., J.G. Peterson, and R.L. Eng. 1975. Foods of adult sage grouse in central Montana. Journal ofWildlife Management 39:628-630 . • .. ' ~ , .. t F' .. • ----,--- --- ...- ~ 100 80 - ~ 1': .: ::r- II Brood sites -~ - - - - I ; C1) (,) cC1) 1: f . 1-- !. i-- (,) 't- 0 ..., 1-- - i i~ 0. - I ~· 40 - - cC1) (,) .... C1) - •,: I'· 1-- (,) 0 D Random sites ~ 1·1: I::' ........ :::s 60 f--- - - r- ~- 1-- :: 1-- ' .. 1:, L 20 - - , - - r- . r- - 1-- 1-- i li - r- ~ - - r- 1- 1- lj! -- ~ ---; i: I• ~~ 1 :~ " ., 0 I E CJ) cu 0. cu <( cu 0. 0 (.) 0::: I- 0:: <( (J,) t .c 0 .c a.. E :J <( (J,) CJ) 0 0. E ::J .E ·c I- ::J "0 L.. I- 0::: I- 0::: <( I ::J (J,) E :J cu 0:: CJ) _J 0 cu I- I I I L.. 0 (.) "0 0. 0 0. ''•: I E :J c 0 \ :I ! 1: I• 1-~I :;, [: I• ' I-- r:- h - I I 0 (.) 0. Cf) I .E CJ) 0 c (.) (J,) L.. <( 0 (J,) w cu . <( 'C" .c (f) ::::> (!) ~ ~ ~ <( 0. I 0 z (.) 0) >. 0 a.. Plant Species Figure 3 .1. The 25 most common plant species at greater sage-grouse brood use locations and random sites in Axial Basin and Danforth Hills, Colorado, 2001. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A. "" 0 i' f ,.., ., r , .. t-· 1 • - • .. ... _T ~ 'l "' "'I' ; 120 • Night-roost Sites 100 Q) (.) s:: Q) ''- 80 ::s (.) D Random Sites r- - - ~ J: t ~ - .:. 1:·· .1 ~ 1::; - (.) 0 '1- 60 - (.) 40 'Q) .':· _ ;, I ~~ D.. 20 - ~ ., :1\ ,- ~~~ " 1- I! _ ,, li,; ·to - I~· I' - - I• i'' 1- C'O 0 <( en a. ::J E ro ~ 0 ' t~~ I ' 1- - - . 1- - 1- .c a. E en ro a. t .c (IJ a. ::J "0 (.) r- 0 L- ~ r- ~ <( 0 0 ~ 0 r· en ~ - <1> <( 0 - 1- ~ - 1- .:. _ ··'~ ,, l i: t:~ (]) - l z 0 I E ::::J <( "0 E ::J .2 ·c r- '+- 0 C'O r- ~ 0 0_J I 0 (.) a. (/) · I E ::::J c 0 0> ~ 0 a. ~ 0 0 ~ I I <( (/) :J (9 a. en (]) c ·a. I I (.) > .E :c ro 0 0 (]) en ::::J _J ::J _J Plant Species Figure 3 .2. The 25 most common plant species at greater sage-grouse night-roost locations and random sites in Axial Basin and Danforth Hills, Colorado, 2001. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A. \0 ....... "- -""- " ,.., ... ,., ' .. . • 100 • Brood sites 1-:- Q) (.) 80 c: I!1.~: ...~ D Random sites ,.,, I"'" l'i I'• ~ 60 - I I"'" r- - f-- I"'" (.) 0 ..... 0 c 40 Q) (.) ... ij· I·· ~. I ~' 20 - ~~ r-~J ~: _ I ·~ ;;~; ~ :!' :' _ :. I 1::' !t I E en ro 0. I 0:: I- 0:: <( I (1) a.. 0 0. :c 11~·: w en I ctl 0. () 1- 0 ; I··· ' f-- n fi·· I (1) en ::J _J 0:: I- 0:: <( "'C ! f-- r- r- i ~ ::J "'0 L... II i•• '.' - f-- I 0 ,I I• i[ !r 1:: 0 rr ;:.,. I,! •f [1. r-; ~ :t: - I"; - i; -;:; ~. [? 1 r- ., r - / 'Lt f~ f';'; T [r Q) a.. i:i'' ctl <( E ::J c 0 >. 0 0') I :~ (1) t ..c h l ::J '+- E .E ctl ctl 0:: 0 1- E ::J 0 (.) <( '+- ·c nl ~ H f}~ I 0') 0') <( 0 () ·:; c 0 > I () I L. (.) 0 ~ 0 (.) 0. (/) n 0:: 0 >(/) [Ji -, I E ::J I L... ctl 0. 0 0. 0 a.. <( 1- a.. Plant Species Figure 3.3 . The 25 most common plant species at greater sage-grouse brood use locations and random sites in Axial Basin and Danforth Hills, Colorado, 2002. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A. \0 N ...,.- - --,.- .- ,- ~ ..,. ~ ~ "<( ~ -:- "~ ..,... 100 ~ Q) u r::: Q) ........ ~ 60 ...... 40 - r::: D Random sites 11~ - q-- I-- t- ·. t- .... •f. n 1:: 1- I'"· u .... 20 - I-- I-- I-- t- ,,, 1-- 1---~ 1- . I' 1- ,., ~~ f:' p I-- I-- : ,• Q) 1- 1- tI-- I •/ Q) c. c i.li fl! u u 0 ....0 • Night-roost sites r------ 80 - .... I• I ~. I-- ~ ~----- 1---- ~' t--- t- I-- t- r- 1---- 1- I-- I-- 1-- 1: I-- I 1- I !. !L 0 I E en ro c. I 0 :c 0.. I a> t .c I 0::: 0::: 1<( "0 '•! I -ro <( w 1;·: ~~ '. I ro 0 a. I L- (.) 0 ~ I 0::: 0::: 1<( n 1:' I c a> L- w I a> en :J _J I E :J s ·c a> ~ en ,. I 0 (.) <( c. C/) c. C/) :J 0 (.9 .E (.) <( 1- 0 u If· . I I ro c. :~ ' :J M D I :J E "E ro 0::: 1- I ·:; c 0 I C) C) <( I' I· I 0 c. c. 0 I 0 .I::. .I::. 0.. D tJ I E ...... c 1- :J 0 C) 0 ro (., I X ~ ro u >. 1- 0 0.. Plant Species Figure 3.4. The 25 most common plant species at greater sage-grouse night-roost locations and random sites in Axial Basin and Danforth Hills, Colorado, 2002. Shrubs are 4-letter genus/species code in capitals, grasses are 4-letter genus/species codes in all lower case, and forbs are 4-letter genus/species code with the first letter capitalized. Scientific names, common names and native status are found in Appendix 1-A. \.0 w 94 Table 3 .1. Results of the multiple regression model using structural habitat variables to predict flock size of greater sage-grouse females and broods at night-roost locations during summer in Moffat County, Colorado, 2001/2002. Parameter Estimate Regression Coefficient (SE) F-value P-value Intercept . 6.5 0.9 52.7 <0.001 Visual obstruction 0- 0.5 m read from a height of 0.5 m and a distance of 5.0 m -3.6 1.5 5.6 0.021 0.06 Visual obstruction 0.5- 1.0 m read from a height of 0.5 m and a distance of 10.0 m 4.7 1.5 9.6 0.003 0.09 Dependant Variable Partial ~ - ~ ~ -.~ .... \ ~ T -~ . "" "" ~ T • ~ ""' . .. -- -, - ... Table 3.2. Results of chi-squared tests among physical macrohabitat characteristics (means and standard error) measured at greater sage-grouse brood, nocturnal and random sites in northwest Colorado, 2001 /2002. Distance to nearest visible structure (m) Distance to nearest twotrack (m) Habitat Variable Slope Year Pooled 2001 2002 2001 2002 Pooled Pooled Random (n=93) 4.7 (0.5) 180 (15) 215 (20) 2008 (14) 1978 (17) 410.4 (123.7) 179.2 (15.5) Brood (n=92) 5.3 (0.6) 147.0(18.6) 188.0 (25.3) 2072 (19) 2117 (23) 306.5 (128.8) 163.0 (14.1) P-value 0.077 0.001 0.097 0.002 0.373 0.862 0.112 Years Pooled Pooled Pooled Pooled Pooled Random (n=93) 4.7 (0.5) 193 (12) 1997 (11) 410.4 (123.7) 179.2 (15.5) Nocturnal (n=58) 4.9 (0.6) 219 (17) . 2110 (16) 383.1 (27.5) N/a P-value 0.183 < 0.001 0.930 N/a Years Pooled 2001 2002 2001 2002 Pooled Pooled Brood (n=92) 5.3 (0.6) 147 (19) 188 (25) 2072 (19) 2117 (23) 306.5 (128.8) 163.0 (14.1) Nocturnal (n=58) 4.9 (0.6) 192 (27) 245 (20) 2113 (22) 2108 (24) 383.1 (27.5) N/a P-value 0.941 0.185 0.231 0.008 0.996 0.994 Aspect Elevation (m) ) 0.007 N/a \0 Vl 96 ) Table 3.3. Microhabitat characteristics varying between years at random sites (mean and standard error) in Axial Basin and Danforth Hills, Colorado, 2001/2002. Habitat variable 2001 2002 F-value P-value Species richness 18.2 (0.7) 11.7(1.0) 29.7 <0.001 Percent forb cover 4.6 (0.3) 2.5 (0.4) 16.6 <0.001 Percent exotic cover 38.1 (2.7) 26.7 (3.8) 4.4 0.038 Percent bare ground 24.7 (3.0) 15.6 (2.8) 4.1 0.047 Visual obstruction from 5 m 18.5 (2.0) 11.5 (1.9) 5.8 0.018 Visual obstruction from 10 m 29.2 (2.1) 17.4 (2.1) 14.1 <0.001 Visual obstruction from 2.5 m (0.5-1.0 m) 9.7 (1.6) 5.9 (1.8) 4.2 0.045 97 Table 3 .4. Pattern of 4 principal components derived from analysis of 15 microhabitat variables from 94 random plots in Axial Basin and Danforth Hills, Colorado, 2001/2002. Structural Feature PC1 Percent shrub cover PC2 PC3 PC4 0.254 -0.250 0.368 Percent sagebrush cover 0.252 -0.344 Shrub height 0.255 Forb height 0.581 Grass height 0.266 Percent grass cover 0.514 Percent litter cover -0.343 0.314 0.507 VISOBa11 a 0.278 -0.309 VISOBa22 a 0.311 0.302 VISOBa23 a 0.306 VISOBb11 a 0.282 VISOBb12 a 0.306 -0.266 VISOBb13 a 0.292 -0.287 VISOBb21 a 0.238 0.271 VISOBb22 a 0.305 0.364 VISOBb23 a 0.304 0.295 Eigenvalue 7.8 2.1 1.4 1.1 Percent of variation explained 48.7 13.4 8.5 7.2 Cumulative percent 48.7 62.1 70.1 77.8 -0.404 -0.441 a Visual obstruction readings where the letter responds to reading height (a= 0.5 m, b = 1.5 m), the first number responds to the 50cm demarcations on the pole (1 = 0 m- 0.5 m, 2 = 0.5 m- 1.0 m) and the second number responds to reading distance from the pole (1 = 2.5 m, 2 = 5. 0 m, 3 = 10 m ). 98 J;.- Table 3.5. Logistic regression models of habitat variables predicting the binary variables for random and greater sage-grouse brood or night-roost sites and diurnal and nocturnal sites in northwest Colorado 2001/2002. Dependant Variable Parameters Regression Coefficient (SE) Wald)( df p Percent forb cover 31.7 (7.0t 20.5 1 < 0.0001 Percent bare ground -3.7 (1.3) a 8.6 1 0.0033 VISOBa12 b 0.8 (0.3) 9.9 1 0.0016 VISOBa13 b -0.7 (0.2) 7.6 1 0.0058 Percent forb cover 14.5 (6.9) a 4.5 1 0.0343 Percent bare ground -2.9(1.1)a 6.6 1 . 0.0105 VISOBa12 b -0.9 (0.4) 4.5 1 0.0332 VISOBa13 b -0.4 (0.2) 2.8 1 0.0966 PC 1 -0.7 (0.2) 15.8 1 < 0.0001 Percent exotics -2.2 (1.1)a 4.4 1 0.0355 VISOBa12 b -1.5 (0.3) 21.1 1 < 0.0001 Brood/ Random I~ "' Night-Roost/ Random Nocturnal/ Diurnal a Arcsin transformed values b Visual obstruction readings where the letter responds to reading height (a= 0.5 m, b = 1.5 m), the first number responds to the 50cm demarcations on the pole (1 = 0 m- 0.5 m, 2 = 0.5 m- 1.0 m) and the second number responds to reading distance from the pole (1 = 2.5 m, 2 = 5.0m,3=10m). ( 99 f ~ I· Table 3.6. Mean and standard error (SE) for habitat variables measured at greater sagegrouse brood, night-roost and random sites in Axial Basin and Danforth Hills, Colorado, 2001/2002. SE Night Roost (n=58) SE Random (n=94) SE 22.4 1.7 9.5 1.3 20.9 1.6 Percent sagebrush cover a 10.6 1.2 4.4 1.0 14.4 1.4 Shrub height (em) 58.1 4.1 30.5 2.7 44.0 2.3 Forb height (em) 11.2 0.7 7.6 0.4 8.2 0.5 Grass height (em) 20.3 0.8 15.2 0.6 17.1 0.7 Percent grass cover a . 6.5 0.6 4.9 0.4 5.9 0.8 Percent forb cover a 8.0 0.7 5.4 0.6 3.8 0.3 Percent exotics a 37.4 2.4 27.0 2.6 33.9 2.3 Percent bare ground a 10.2 1.2 9.1 2.1 21.3 2.2 Litter a 84.8 1.6 80.7 2.2 75.8 2.4 VISOBa11 (em) b 16.3 1.2 3.1 0.9 12.0 1.3 VISOBa12 (em) b 22.2 1.3 4.0 0.9 15.9 1.4 VISOBa13 (em) b 27.9 1.4 9.6 1.6 24.8 1.6 VISOBa22 (em) b 8.8 1.1 1.4 0.6 5.4 1.1 VISOBa23 (em) b 12.5 1.2 2.1 0.9 8.3 1.2 VISOBb11 (em) b 11.5 1.0 0.6 0.3 7.7 1.2 VISOBb12 (em) b 18.5 1.3 2.2 0.7- . 13.5 1.4 VISOBb13 (em) b 26.4 1.4 6.3 1.3 20.2 1.5 VISOBb21 (em) b 3.0 0.6 0.5 0.3 2.0 1.5 VISOBb22 (em) b 5.7 0.8 0.3 0.2 3.3 0.8 VISOBb23 (em) b 9.5 1.2 1.1 0.7 5.1 1.0 Brood (n=92) Percent shrub cover a • r )> 'II II" Variable a Arcsin transformed for MANOV A, untransformed mean and SE are reported bVisual obstruction readings where the letter responds to reading height (a= 0.5 m, b = 1.5 m), the first number responds to the 50cm demarcations on the pole (1 = 0 m- 0.5 m, 2 = 0.5 m- 1.0 m) and the second number responds to reading distance from the pole (1 = 2.5 m, 2 = 5.0 m, 3 = 10 m). 100 • ~ • .., Table 3. 7. Microhabitat characteristics varying among early and late brood-rearing sites (mean and standard error) of greater sage-grouse in Axial Basin and Danforth Hills, Colorado, 2001. Brood-rearing periods were estimated as the time in which 90% of movements had been made to alternative cover types or long distances movements had ceased. In 2001 this date was 7 July. SE Within Canonical 2.6 Late broodreanng (n=28) 21.5 2.7 0.094 12.7 1.8 8.4 1.4 0.271 Shrub height (em) 53.9 5.8 53.0 4.0 0.146 Forb height (em) 11.1 0.8 11.3 0.9 0.122 Grass height (em) 21.7 1.0 20.0 1.1 -0.018 Percent grass cover a 5.8 0.5 9.1 1.8 0.149 Percent forb cover a 7.5 0.7 8.9 1.0 -0.214 Percent exotics a 40.9 4.6 37.7 3.1 0.118 Percent bare ground a 14.0 2.8 9.9 1.5 0.003 Litter a 77.9 4.2 87.6 1.2 -0.121 VISOBal1 (em) b 16.5 2.0 16.1 2.6 0.116 VISOBal3 (em) b 32.3 2.2 31.4 2.3 0.133 VISOBa21 (em) b 3.7 1.5 5.3 1.5 0.091 VISOBa23 (em) b 12.5 2.5 15.4 2.1 0.192 VISOBb11 (em) b 12.6 1.8 11.5 1.8 0.233 VISOBbl3 (em) b 28.1 2.0 26.8 2.4 0.200 VISOBb21 (em) b 3.0 1.2 2.6 0.9 0.176 VISOBb23 (em) b 8.3 2.1 10.3 2.0 0.209 Early broodrearing (n=31) SE Percent shrub cover a 20.3 Percent sagebrush cover a Variable • • a Arcsin transformed for MANOVA, untransformed mean and SE are reported b Visual obstruction readings where the letter responds to reading height (a= 0.5 m, b = 1.5 m), the first number responds to the 50cm demarcations on the pole (1 = 0 m- 0.5 m, 2 = 0.5 m- 1.0 m) and the second number responds to reading distance from the pole (1 = 2.5 m, 2 = 5.0 m, 3 = 10 m). 101 • CHAPTER 4. SEASONAL AND ANNUAL MOVEMENTS AND SURVIVAL OF GREATER SAGE-GROUSE IN NORTHWEST COLORADO • INTRODUCTION The life history of greater sage-grouse has been relatively well studied (Schroeder et al. 1999) and many studies have monitored radio-marked individuals. Despite this, there are shortcomings in the reports of annual and seasonal survival rates and home ranges across the range of greater sage-grouse. Large-scale survival studies have been limited to markrecapture data from banded individuals or have not provided seasonal estimates, confidence limits or annual comparisons (Connelly et al. 1994, Zablan et al. 2003). Home range analysis in the literature has been limited to a few radio-marked individuals describing brood home range, and no data are available estimating annual home range sizes. Home range is an area used by an individual in its normal daily activities (Burt 1943). Brood home range is the area maintained by a brood from hatch until10 weeks, when brood identity is lost and chicks are considered independent (Wallestad 1971, N. Burkepile personal communication). Home range size and movements are widely used in wildlife biology to assess animal-habitat relationships and survival (Thompson and Fritze111989, White and Garrott 1990, Drut et al. 1994, Chamberlain and Leopold 2000). However, home range estimation can be influenced by autocorrelated data (Otis and White 1999), sample size and the home range estimator used (Girard et al. 2002). The validity and utility of applying results of home range analysis to long-term wildlife management objectives also depends on the assumption that populations remain faithful to such ranges over time (Van Dyke et al. 102 • 1998). Greater sage-grouse females exhibit site fidelity to leks (Dunn and Braun 1985), nest sites (Fisher et al. 1993) and winter areas (Berry and Eng 1985) . • Understanding the dynamics of a wild population requires precise estimates of annual survival (Zablan et al. 2003). My objectives were to (1) estimate seasonal and annual movements, (2) estimate monthly, seasonal and annual survival rates by age class and (3) to develop predictive seasonal and annual survival models of females using years, seasons, age and trend as covariates. METHODS Movements were separated into 4 periods for analysis: breeding (16 March -15 June), brood-rearing (15 June- 31 August), fall (1 September- 15 November) and winter (16 November- 15 March), based on grouse behavior. The lek of capture served as the point of origin for all grouse captured on lek sites and the initial nest site was the ending point for breeding season movement, thus analysis was restricted to females that initiated nests. Distances were estimated using ARCVIEW. Mean distances to nests following a nest success between years and a failure in the same year were compared using a student's t-test. Lek to nest distance was related to female age, location of lek, and year using ANOVA. Mean distances moved by yearling and adult females seasonally and annually were compared using an ANOVA (Ott and Longnecker 2001). Seasonal Movements For fall and winter seasons, I estimated the mean movement of individual grouse from the lek or site of capture and then derived the overall mean movement per season from the averages to take into account discrepancy in number of locations per bird. Grouse with < 103 3 locations/ season were not included in the analysis. Fall and winter movements were log transformed to meet the assumption of normality and then assessed for variation between • ages, and years using ANOVA. Movement directions and altitudinal migration was estimated using ARCVIEW. Daily Movements Seasonal home range and median daily movement were restricted to the broodrearing season, as consecutive locations in the fall and winter were too far apart in time (Samuel and Fuller 1996). Linear distances were calculated between subsequent locations and divided by the number of days elapsed between locations. A regression analysis was used to estimate the influence of daily movement on number of chicks/female at 6 weeks post-hatch. Daily movements did not meet the assumption of normality therefore I used a non-parametric Kruskal-Wallis ANOVA to evaluate differences between ages, years and brood status (successful vs. unsuccessful females). Median distances moved to night-roost sites were calculated for females whose diurnal location was< 12 hrs before or after the night-roost site. Home range An ARCVIEW extension, Animal Movement (Hooge and Eichenlaub 1997) was used to estimate home range sizes. Independent locations (2: 24 hrs apart) were used in the calculations. A 95% fixed kernel estimate of home range was used with least squares crossvalidation to choose the width of kernel bands. This estimator is considered accurate even with < 20 locations, and is not biased by an unequal number of locations (Seaman et al. 1999). Brood home ranges (hatch to 10 weeks) from 2001 and 2002 were pooled because samples from each year were too small for statistical analysis. I used a non-parametric 104 Kruskal-Wallis ANOVA to evaluate differences in brood home range size with respect to nest location and age. Annual home range estimates were evaluated for females that survived annually and had > 20 independent locations (range 21- 40) using a non-parametric ANOVA to estimate differences due to year or age. Survival Female survival was analyzed using program MARK (White and Burnham 1999), which incorporated Kaplan-Meier product limit method with staggered entry (Pollock et al. 1989). Monthly, seasonal and annual survival rates for yearlings and adults were calculated. The annual survival interval commenced 1 April. Grouse classified as yearlings at the time of capture were re-classified as adults the following April. After building the global model, the data were analyzed for overdispersion, a phenomena which reflects the lack of independence or heterogeneity among individuals (Anderson et al. 1994). A c > 1 indicates that the data are heterogeneous and dependent. I calculated c = 2.43 for the annual data, indicating that the gregarious nature of greater sage-grouse year-round caused individuals to be interpreted as units. Overdispersion was not seen in the seasonal analysis (c < 0.77). The annual data set was corrected for overdispersion and models were selected using the small sample bias corrected quasi-Akaike's Information Criteria (QAICc). The small sample-bias corrected Akaike's Information Criteria (AICc) was used to select among seasonal models. The principle of parsimony was used to select the best model among alternative parametric models (Burnham and Anderson 1998). Models were selected to determine whether survival rates differed among season, between years or age classes (Pollock et al. 1989). Individuals were recovered as soon as possible after a mortality signal was heard. Cause-specific 105 mortality was classified as avian, mammalian, or unknown through examination of the carcass and surrounding area (S. Boutin, personal communication). RESULTS Seasonal Movements Mean lek to nest movements for 2001 (4.00 km, SE = 0.56, n = 41) did not differ (F= 0.17, 1 df, P = 0.682) from 2002 (4.01 km, SE = 0.46, n = 60). Mean lek to nest movements for females attending leks in the Axial Basin (4.58 km, SE = 0.45, n = 72) was larger (F = .. 6.74, 1 df, P = 0.011) than that of females attending leks in the Danforth Hills (2.59 km, SE = 0.43, n = 29). Average lek to nest movements for adults (3.90 km, SE = 0.42, n = 72) was similar (F = 0.26, 1 df, P = 0.613) to that of yearlings (4.31 km, SE = 0.69, n = 26). The average distance adult females moved between consecutive nests either between years or to renest was 0.45 km (n = 25). The mean distance females moved to consecutive nests following a successful nesting attempt was 0.32 km, SE = 0.09, n = 13, which did not differ (t = 1.34, P = 0.190) from the distance moved following a nest failure to renest or between years (0.58 km, SE = 0.18, n = 12). Mean female movement in the fall from the lek of capture was 7.8 km (range: 0.421.5 km, n = 70) and did not vary between ages (F= 0.03, P = 0.874) or years (F= 0.00, P = 0.955). Mean movement from winter locations to leks was 9.9 km (range: 0.8 - 30.6 km, n = 76) and also did not vary between ages (F = 0.02, P = 0.893) or years (F = 0.27, P = 0.608). Seasonal movements tended to be elevational in nature. Directional movements of females was northwest to winter sites < 1,978 m in the Axial Basin. Some females moved further northwest to winter home ranges between Lay and Sunbeam, Colorado. Female movement was from mid-October to mid-November and then grouse remained at low elevations until 1.4. li?" .. i 106 mid-March. Movements to leks occurred from mid-March to early April. Elevation of females during the breeding s~ason ranged from 2,045-2,090 m. Another movement upward in elevation(> 2,100 m) was observed in July and August (Figure 4.1). Daily Movements I collected 1,103 greater sage-grouse locations over the brood-rearing period. Of these locations, 92% were collected within 10 days of the previous location. Individuals with < 3 locations and location> 10 days apart were excluded from analysis (Apa 1998). For females with broods (n = 23), the greatest daily movements were made in weeks 6 and 7 post-hatch. I detected no difference in daily movement by year(/= 1.82, P = 0.179) or age of female(/ = 0.67, P = 0.412). A difference was detected, however, for brood status. Females with broods (median: 184 rnlday, range= 15 - 520, n =50) moved further(/= 25.53, P < 0.001) than those without (median: 103 rnlday, range= 46- 566, n = 47). Brood size at 6 weeks post-hatch was not associated with daily movements (F = 0.06, P = 0.813, l = 0.001, n =50). Median movement from the most recent day locations(< 12 hrs) to night roost locations was 397m, range: 77 - 4,462 m. Home range The distributions of brood and annual home range sizes were skewed so the median was used to determine the central tendency of the data (Ott and Longnecker 2001 ). An outlier, with home range of8,176 ha was removed from analysis ofbrood home range. No differences were detected between brood home range sizes of adults and yearlings (/ = 0.13, 1 df, P = 0.721). However, home range sizes of brood females that nested in the Axial Basin (1,151 ha) were larger(/= 3.69, 1 df, P = 0.055) than those that nested in the Danforth Hills (439 ha) (Table 4.1 ). Annual home range estimates did not vary between years(/= 0.04, 1 107 df, P = 0.850). Median annual home range of yearling females was 8,574 ha, n = 26 and did not differ(/= 0.0001, 1 df, P = 0.990) from that of adult females (6,556 ha, n = 43) (Table 4.2). Survival Annual survival rates for adult females were 0.65, and 0.48 for 2001 and 2002, respectively (Table 4.3). Annual survival rates were 0.71 and 0.78 for yearling females. Annual survival rates pooled over 2 years were 0.75 for yearling females and 0.57 for adult females. Seasonal survival rates were lowest in the brood-rearing period in 2001 and the breeding period for 2002. Survival for the brood-rearing period (16 June- 31 August) in 2001 was 0.84 for adult females and 0.85 for yearling females, while survival for the breeding period (1 April- 15 June) in 2002 was 0.75 for adult and 0.84 for yearling females (Table 4.3). Monthly survival rates in 2001 were lowest for both adult (0.88) and yearling (0.88) females in July (Tables 4.4 and 4.5). In 2002, however, monthly survival rates were lowest in May (0.85) and October (0.85) for adult females and May, June and September and October (0.93) for yearling females. November through March were periods of low mortality for females in both years (Tables 4.4 and 4.5). Models were constructed to determine whether seasonal survival rates varied with year or age. In model building, {S (t)} denotes variation due to year and {S (a)} denotes variation due to age. The model {S (a+ t)} denotes variation due to both .parameters. Models within a !J. AICc of :S 2 showed insufficient evidence to be excluded as the most credible model. AICc > 2 !J. but :S 4, provided weak evidence that the model was not the best. Models with !J. AICc > 4 but :S 7, exhibited strong evidence that the model was not the best fit 108 for the data (Burnham and Anderson 1998). The model that best fit the data for breedingperiod survival was the model {S (a+ t)}, which indicated that survival varied between ages and years (Table 4.6). For the brood-rearing, fall, and winter seasons the model that best fit the data was the null model, indicating that survival did not vary by age or year (Table 4.6). The annual survival models selected were the null model and survival that varied with age, year and seasonal trend (Figure 4.2). Based on the principle of parsimony, the null model was considered the best of the competing models (Table 4. 7). Length of time between the last alive signal and carcass retrieval for greater sagegrouse averaged 10.46 days. Through examination of carcasses retrieved within 5 days ( x = 2.5, n = 18), I attributed 39% of mortalities to mammalian predation (n = 7), 39% unknown (n = 7), 17% to avian predation (n = 3), and 6% to reptilian predation (n = 1). Of the avian mortalities 1 was likely a great homed owl and the other 2 were attributed to golden eagles (Aquila chrysaetos). The reptile death was attributed to a Western rattlesnake bite (Crotalus viridis). This mortality also occurred while the female was incubating. Two puncture wounds were found on the lower abdomen and were perhaps inflicted as the female attempted to protect her eggs. Her carcass was found a short distance from the nest and her eggs were covered in blood. Of the avian-suspected depredations retrieved within 5 days (n = 3), 1 was attributed to a great homed owl. The carcass was recovered at midnight during a night-roost sampling session. The transmitter had not switched to mortality, the carcass was still warm and the female had died as a result of impact to the head. The microhabitat was a wet meadow. The other 2 mortalities were attributed to avian causes because of characteristics of the kill. One had a severed head present and bones dislocated with clean breaks at the joints. Intestines 109 were pulled out and smeared on the remains. The microhabitat was an go, north-facing slope in 5-l 0% sagebrush canopy cover. The other avian-suspected mortality had streaks of whitewash present at the site of the collar. No other remains were found. The microclimate was a 20° south-facing slope at the edge of a rock cliff. Of the mammalian-suspected depredations retrieved within 5 days (n = 7), none had whitewash, pellets, scat, raptor feathers or tracks present at the kills site. Four carcasses had blood present indicating that the kill was made at the site. Of those, 2 were females incubating nests, 1 was killed in a mountain shrub community and 1 in 5-10% sagebrush canopy cover. The head was severed at 1 kill site, and legs remained at 2 sites. Seventy-one percent of mammalian-suspected depredations had skin remaining on tufts of feathers and 43% exhibited feathers broken at the quills and the cecum present in 1 piece. Two sites had broken bones, cracked between the joints. Feathers were scattered from 1-3m2 and wings were remaining intact at 43% of kill sites. The mammalian depredations could not be distinguished to species, due to confounding evidence at all sites. • DISCUSSION Seasonal Movements Greater sage-grouse populations have been described as migratory or non-migratory (Eng and Schladweiler 1972, Wallestad 1975, Schoenberg 1982, Connelly et al. 1988, Leonard et al. 2000). Migration can either occur between winter/nest areas and summer areas, winter and nest/summer areas or by a combination of movements between winter, nest and summer areas (Schroeder et al. 1999). 110 Nesting tends to be associated with sagebrush habitat surrounding lek sites (Connelly et al. 2000b), but has also occurred > 20 km from breeding grounds (Wakkinen et al. 1992). Distances traveled from breeding grounds to nest sites in central Montana averaged 2.5 km and 2.8 km for adults and yearlings, respectively (Wallestad and Pyrah 1974). Alternatively, a study by Schoenberg (1982), in North Park, Jackson County, Colorado indicated that adult females traveled further to nests than did yearlings. Lek to nest distance averaged 2. 7 km. Lek to nest movements of 4.0 km in my study is similar to those found in Idaho, which ranged from 3.4 to 4.9 km (Wakkinen 1990, Apa 1998, Fischer 1994). Lek to nest movements may be an indicator of suitable nesting habitat. Habitat in the Axial Basin was less suitable for nesting than that in the Danforth Hills (Chapter 1), thus females were required to move further from breeding grounds to nesting sites in the Axial Basin. The mean distance to consecutive nests was 0.45 km (n = 25), which is less than the 0.70 km (n = 22) reported in Idaho (Fischer et al. 1993). Dunn and Braun (1985) speculated that nest fidelity might be related to nest success. My results failed to confirm this, as movements to successive nests did not differ based on nest fate. Greater sage-grouse in northwestern Moffat County migrated up to 30.3 km to wintering areas in late November as reported by Dunn and Braun (1986). This is consistent with what I observed; the greatest distances moved from brood-rearing sites to fall and winter areas (30.6 km) occurred in November and December. Mean distances from the lek of capture in the fall and winter was 7.8 and 9.9 km, respectively. Dargon et al. (1942) noted elevational migration of greater sage-grouse near Craig, Colorado and in the Axial Basin. In this study movements to below 2,073 m began in mid-November and were completed by the start of December. A second migration upward in elevation began in late February and 111 continued on through March. I observed mean elevation of females below 2,000 m from November-March. Movement upward in elevation began in late March and continued on through August. Movements of migratory greater sage-grouse in Wyoming, Idaho and . Colorado ranged from 20-82 km (Schoenberg 1982, Berry and Eng 1985, Connelly et al. 1988) between summer, winter and breeding areas. The mean movement of females between •• seasons from the lek of capture was < 10 km in my study, which indicates this population is non-migratory. Daily Movements Median daily brood movements (184 m/day) were similar to those reported by Apa (1998) (144 rnlday, n = 9), but less than those reported by Autenrieth (1981). Previous research indicated that greater sage-grouse chicks changed their diet and subsequently their habitat use at 6 weeks of age (Martin 1970, Peterson 1970, Drut et al. 1994). This corresponded with the greatest movements being made at 6 and 7 weeks post-hatch. There was also a peak in movement of broods at 3 and 4 weeks post-hatch. This was associated with females that hatched nests late, likely re-nesters. Broods that hatched late in the season were forced to move earlier in search of forbs and insects. Females with broods (n =50) moved a distance of 80 rnlday further than females without broods (n = 47). Broods have higher energetic demands and may require greater search times for specific food items. Contrastingly, Apa (1988) observed that females with broods (median: 144m/day, n = 9) moved less than females without broods (median: 284 m/day, n = 4). While brood size in my study was not associated with daily movements, a study on gray partridge broods in the United Kingdom showed that survival rates were strongly negatively correlated with distances moved (Green 1984, Rands 1985). Similarly, 112 research on greater sage-grouse broods in Oregon related larger brood home ranges with less suitable habitat availability, and conjectured that the differences were associated with reduced reproductive potential (Drut et al. 1994). Median movements from the most recent day locations to night-roost locations were 397m. Similarly, Beck (1977) reported movements of individuals to winter night-roost locations from their last observed diurnal locations of> 250 m. Given that the median daily movement for brood females was only 184 m/day, and movements to night-roost locations were 2 times further, indicated that females with broods moved significant distances in order to find suitable night-roosting habitat. Additionally, this suggested that literature values of daily movements were underestimated as researchers failed to include nocturnal forays into their movement data. Home range Home range sizes may reflect the energy expended by an animal in searching for its basic requirements (Burt 1943). Chamberlain and Leopold (2000) found that movement in turkeys (Meleagris galapavo) increased with decreasing habitat quality and increasing habitat variability. Likewise, survival rates of male ruffed grouse (Bonasa umbellus) in central Missouri and female ring-necked pheasant in southern Wisconsin were inversely related to seasonal home range size and mean daily movement (Gatti et al. 1989, Thompson and Fritzen 1989). Alternatively, Hubbard et al. (1999) found that the risk of mortality for female turkeys decreased by 2.0% for every 10 ha increase in home range size. Home range sizes for greater sage-grouse vary in the literature, although few data are available. Three yearling females with broods occupied a summer home range of 94 ha in southeast Idaho (Connelly and Markham 1983). Wallestad (1971) reported brood home 113 range sizes in central Montana to be 233 ha (n = 13). In Oregon, brood home range sizes were associated with less suitable habitat (Drut et al. 1994). Home range sizes from hatch to 12 weeks were 900 ha and 7,200 ha for good and poor habitat, respectively. Using this standard, our median brood home range size for 10 weeks of548 ha (n = 31) implied that the habitat available was good. However, home range sizes varied significantly with nesting location in my study area. The implication is that birds that nested in the Danforth Hills (median home range= 439 ha) were required to move less in order to meet their basic needs than birds that nested in the Axial Basin (median home range= 1,151 ha). Habitat in the • Axial Basin had less forb cover and more bare ground than in the Danforth Hills, attributes which made it less suitable for brood-rearing (Chapter 1). Birds breeding in the Danforth Hills typically showed an altitudinal migration in the late brood-rearing period while females breeding in the Axial Basin either moved into the Danforth Hills or remained in riparian/wet meadows in the Axial Basin. The movement of birds in late summer in my study area is consistent with Wallestad (1971), Schoenberg (1982), and W akkinen (1990) who noted shifts in late brood-rearing habitat to more mesic sites as forbs began to desiccate. Fischer et al. (1996) related this movement to vegetal moisture content of :S 60% water. Most females in the study area migrated southward and upward in elevation to mountain shrub communities in the Danforth Hills (Savage 1969). However, some birds in the Axial Basin remained in wet meadow or riparian areas following a gradient of forbs (Klebenow 1969). Leonard et al. (2000) described a 3-way movement of a migratory population of greater-sage grouse on the Upper Snake River plain of Idaho. Movements were from breeding/summer use areas to fall areas and fmally to wintering sites, and encompassed 114 27 6,400 ha. Median annual home range estimates for yearlings and adults in the Axial Basin and Danforth Hills was 8,574 and 6,556 ha, respectively. No estimates are available in the literature describing annual home range size of individual greater sage-grouse, however, the annual home range sizes and seasonal distances traveled from the lek of capture indicated that the population of greater sage-grouse residing in the Axial Basin and Danforth Hills is I I. I~ non-migratory. Survival The annual survival models selected suggested that survival varied with age, year and seasonal trend. Additionally, seasonal survival models selected indicated that survival rates varied between years and ages for the breeding period and did not vary by age or year for brood-rearing, fall or winter seasons. Mean annual survival rates for adult and yearling females in the study area were 57% and 75%, respectively. Previous estimates in Moffat County, Colorado indicated female survival rates to be 60%. In Idaho, annual survival rates for females as reported by Connelly et al. (1994) ranged from 68-85%. Similarly, Zablan et al. (2003) estimated annual survival rates of adults (59%) to be less than that of yearling females (77%) from a 17-year band-recovery data set in North Park, Colorado. They indicated, however, that models incorporating time effects for each year did not fit the data as well as those that incorporated a trend over multiple years. For the breeding period, the model incorporating annual time effects was likely selected for my data set due to the disparity in precipitation over the 2-year study (Appendix 1-B). In the second year, the lowest seasonal survival rates were during the breeding period, and were greater for adult than yearling females. Of 13 mortalities in May and June 2002, 62% (n = 8) of these 115 occurred as females incubated potentially as a result of reduced herbaceous cover at the nest due to low precipitation in the spring (Chapter 1). The survival rates and the competing annual models indicated that there was an upward trend in seasonal survival from the breeding period to the winter. This is consistent with populations in Idaho where 52% percent of adult female mortality occurred during the spring and summer and 46% occurred in September and October (Connelly et al. 2000a). This data, however, was collected on exploited greater sage-grouse populations where 42% of deaths of adult females were attributed to hunting. Additionally, Wik (2002) reported that the lowest seasonal female survival rates in southwestern Idaho occurred in the fall and associated these mortalities with harvest. Seasonal and monthly survival rates did not indicate hunting to be a factor affecting mortality of adult females in my study population. I had no documented band returns over 2 seasons of 109 radio-marked and 23 banded grouse alive in September. Birds resided in both the Danforth Hills, which is closed to hunting, and on BLM and private land in the Axial Basin during the hunting season in Moffat County. Although high monthly mortality rates were observed among yearling females in September and October, examination of carcasses indicated predation to be the main cause of mortality. Bowman and Harris (1980) indicated cover strongly influenced prey vulnerability. Typical canopy cover for nesting females is less than that at brood-rearing sites (Schoenberg 1982). A biological explanation for low survival during the breeding season, particularly in a drought year, was that incubating females were immobile and had poor concealment due to stunted herbaceous growth. Grass heights were 6-7 em shorter and there was 30% less vertical and 25% less horizontal cover at the nest bowl in 2002 (Chapter 1). Additionally, low survival rates occurred in July, during the brood-rearing period. In this period, as the 116 vegetation desiccated, grouse made their greatest movements related to vegetal moisture content (Fischer et al. 1996). Additionally, female survival was low in October, the time when females were making their movements to winter sites. Females in the Axial Basin and Danforth Hills underwent the same trends as those in southern Idaho, and experienced low mortality over the winter (Connelly et al. 2000a). Cause-specific mortality was primarily mammalian, although, using marks on the transmitter as a diagnostic tool, I may have over-estimated this cause of mortality. Bumann and Stauffer (2002) argued that survival studies reporting cause-specific mortality typically overestimated mammalian-caused mortality due to scavenging of the carcass. By analyzing remote cameras at ruffed grouse carcasses in Virginia, they estimated 100% of carcasses ~ I were scavenged in > 5 days. They indicated that marks to the transmitter can serve as a diagnostic when determining cause of death, as mammalian predators tended to focus on the head and neck when killing avian prey. Using this time criteria, 71% (n = 5) of mortalities attributed to mammalian causes may be false as there were no marks on the transmitter. ., I MANAGEMENT IMPLICATIONS Survival rates for the population of greater sage-grouse in the Axial Basin and Danforth Hills appeared to be consistent with or higher than literature values in years of normal precipitation. Cause-specific mortality appeared to be primarily mammalian, although a high proportion of these may have been false due to misinterpretation of remains due to scavenging. In order for a cause-specific mortality study to be of greater value, the time from death to carcass recovery needs to be reduced and the rate of scavenging of greater sage-grouse carcasses in a shrub-steppe community needs to be estimated through the use of .. 117 remote cameras (Bumann and Stauffer 2002). Because nest success appeared high, chick survival (Chapter 3) and juvenile survival needs to be examined as mechanisms limiting this population (Connelly and Braun 1997, Connelly et al. 2000b). Nest site fidelity in this population appears to be higher than reported in other studies. Nest fidelity can result in reduced nesting if large tracts of nest habitat are destroyed (Fischer et al. 1993), thus emphasizing the importance of nest habitat identification and conservation. The exaggerated movements females made to night-roost habitat within the context of daily brood movements may have indicated that this habitat type may have been limited during the summer. Habitat features are involved in determining the vulnerability of prey and are exaggerated under drought conditions (Reynolds et al. 1988). Habitat conditions I I~ I r. appeared to be inversely related to movement and directly associated with survival. Further research is needed to explore how different land use practices might improve habitat conditions for nesting and brood-rearing in the Axial Basin and whether increasing the frequency of burned areas on the landscape would decrease movements to nocturnal use sites. 118 LITERATURE CITED Anderson, D.R., K.P. Burnham, and G. C. White. 1994. AIC model selection in overdispersed capture-recapture data. Ecology 75:1780-1793. Apa, A.D. 1998. Habitat use and movements ofsympatric sage and Columbian sharp-tailed grouse in southeastern Idaho. Ph.D. Dissertation, University of Idaho, Moscow, Idaho. Autenrieth, R.E. 1981. Sage grouse management in Idaho. Wildlife Bulletin 9. Idaho Department ofFish and Game, Boise, Idaho. ,, r, Beck, T.D.I. 1977. Sage grouse flock characteristics and habitat selection in winter. Journal ofWildlife Management 41:18-26. Berry, J.D., and R.L. Eng. 1985. Interseasonal movements and fidelity to seasonal use areas by female sage grouse. Journal of Wildlife Management 49:237-240. Bowman, G.B., and L.D. Harris. 1980. Effect of spatial heterogeneity on ground nest depredation. Journal of Wildlife Management 44:806-813. Bumann, G.B., and D.F. Stauffer. 2002. Scavenging of ruffed grouse in the Appalachians: influences and implications. Wildlife Society Bulletin 30:853-860. Burnham, K.P., and D.R. Anderson. 1998. Model selection and inference- a practical information-theoretic approach. Springer-Verlag New York Inc., New York. 353pp. Burt, W.H. 1943. Territoriality and home range concepts as applied to mammals. Journal of Mammalogy 24:346-352. Chamberlain, M.J., and B.D. Leopold. 2000. Habitat sampling by female wild turkeys during preincubation. Wilson Bulletin 112:326-331. Connelly, J.W., and C.E. Braun. 1997. Long-term changes in sage grouse Centrocercus urophasianus populations in western North America. Wildlife Biology 3:229-234. Connelly, J.W., and O.D. Markham. 1983. Movements and radionucleolide concentrations of sage grouse in southeastern Idaho. Journal of Wildlife Management 4 7: 169-177. Connelly, J.W., H.W. Browers, and R.J. Gates. 1988. Seasonal movements of sage grouse in southeastern Idaho. Journal of Wildlife Management 52:116-122. Connelly, J.W., M.A. Schroeder, A.R. Sands, and C.E. Braun. 2000b. Guidelines to manage sage grouse populations and their habitats. Wildlife Society Bulletin 28:967-985. 1 119_ Connelly, J.W., A.D. Apa, R.B. Smith, and K.P. Reese. 2000a. Effects of predation and hunting on adult sage grouse Centrocercus urophasianus in Idaho. Wildlife Biology 6:227-232. Connelly, J.W., K.P. Reese, W.L. Wakkinen, M.D. Robertson, and R.A. Fischer. 1994. Sage grouse ecology report. Idaho Department ofFish and Game Job Completion Report. W160-R-19. Subproject 9. Boise. . . I Dargon, L.M., R.J. Keller, H.R. Shepard and R.N. Randall. 1942. Survey of 1941-1942: Food studies, parasite relations, habitat requirements, including preliminary data on sharp-tail grouse in Moffat and Routt Counties. Colorado Game and Fish Commission, Sage Grouse Survey 4: 10pp. Drut, M.S., J.A. Crawford, and M.A. Gregg. 1994. Brood habitat use by sage grouse in Oregon. Great Basin Naturalist 54:170-176. r • I~ ,..... I Dunn, P.O., and C.E. Braun. 1985. Natal dispersal and lek fidelity of sage grouse. Auk 102:621-627. Dunn, P.O., and C.E. Braun. 1986. Late summer-spring movements of juvenile sage grouse. Wilson Bulletin 98:83-92. Eng, R.L., and P. Schladweiler. 1972. Sage grouse winter movements and habitat use in central Montana. Journal ofWildlife Management 36:141-146. · Fischer, R.A. 1994. The effects of prescribed fire on the ecology of migratory sage grouse in southeastern Idaho. Ph.D. Dissertation, University of Idaho, Moscow, Idaho. Fischer, R.A, A.D. Apa, W.L. Wakkinen, K.P. Reese, and J.W. Connelly. 1993. Nestingarea fidelity of sage grouse in southeastern Idaho. Condor 95:1038-1041. Fischer, R.A, K.P. Reese, and J.W. Connelly. 1996. Influence ofvegeta1 moisture content and nest fate on timing of female sage grouse migration. Condor 98:868-872. Gatti, R.C., R.T. Dumke, and C.M. Pils. 1989. Habitat use and movements of female ringnecked pheasants during fall and winter. Journal of Wildlife Management 53:462-475. r Girard, 1., J.P. Ouellet, R. Courtois, C. Dussault, and L. Breton. 2002. Effects of sampling effort based on GPS telemetry on home-range size estimations. Journal of Wildlife Management 66: 1290-1300. Green, R.E. 1984. The feeding ecology and survival of partridge chicks (Alectoris rufa and Perdix perdix) on arable farmland in East Anglia. Journal of Applied Ecology 21:817830. 120 Hooge, P.N., and B. Eichenlaub. 1997. Animal movement extension to arcview. Ver. 1.1. Alaska Biological Science Center, U.S. Geological Survey, Anchorage, AK, USA. Hubbard, M.W., D.L. Gamer, and E.E. Klaas. 1999. Factors influencing wild turkey female survival in southcentral Iowa. Journal ofWildlife Management 63:731-738. Klebenow, D.A. 1969. Sage grouse nesting and brood habitat in Idaho. Journal of Wildlife Management 33:649-662. Leonard, K.M., K.P. Reese, and J.W. Connelly. 2000. Distribution, movements and habitats of sage grouse Centrocercus urophasianus on the Upper Snake River Plain of Idaho: changes from the 1950s to the 1990s. Wildlife Biology 6:281-286. Martin, N.S. 1970. Sagebrush control related to habitat and sage grouse occurrence. Journal of Wildlife Management 34:313-320. Otis, D.L., and G. C. White. 1999. Autocorrelation of location estimates and the analysis of radiotracking data. Journal of Wildlife Management 63:1039-1044. Ott, R.L., and M. Longnecker. 2001. An introduction to statistical methods and data analysis, fifth edition. Duxbury, Pacific Grove, CA. 1152pp. Peterson, J.G. 1970. The food habits and summer distribution of juvenile sage grouse in central Montana. Journal of Wildlife Management 34:147-155. Pollock, K.H., S.R. Winterstein, C.M. Bunck, and P.D. Curtis. 1989. Survival analysis in telemetry studies: the staggered entry design. Journal of Wildlife Management 53:7-15. Rands, M.R.W. 1985. The survival of gamebird (Galliformes) chicks in relation to pesticide use on cereals. Ibis 128:57-64. Reynolds, C.J., P. Angelstam, and S. Redpath. 1988. Predators, their ecology and impact on gamebird populations. Pages 72-97 in P.J. Hudson and M.R.W. Rands (eds.), Ecology and Management of Gamebirds. BSP Professional Books, Oxford, England. 263 pp. Samuel, M.D., and M.R. Fuller. 1996. Wildlife radiotelemetry. Pages 370-418 in T.A. Bookout, ed. Research and management techniques for wildlife and habitats, 5th ed., revised. The Wildlife So~iety, Bethesda, MD. Savage, D.E. 1969. The relationship of sage grouse to upland meadows in Nevada. M.S. thesis. University of Nevada, Reno. 101 pp. Schoenberg, T .L. 1982. Potential impacts of strip mining on sage grouse movement. Colorado Game and Fish. Project W-37-R-35. Work Plan 3. Job 12. Job Completion Report. 59-142. 121 Schroeder, M.A., J.R. Young, and C.E. Braun. 1999. Sage grouse, Centrocercus urophasianus. The Birds ofNorth America. 425:27pp. Seaman, D.E., J.J. Millspaugh, B.J. Kemohan, G. C. Brundige, K.J. Raedeke, and R.A. Gitzen 1999. Effects of sample size on kemal home range estimates. J oumal of Wildlife Management 63:739-747. Thompson, P.R., ill, and E.K. Fritzen. 1989. Habitat use, home range, and survival of territorial male ruffed grouse. Journal ofWildlife Management 53:15-21. . Van Dyke, F.G., W.C. Klein, and S.T. Stewart. 1998. Long-term range fidelity in Rocky Mountain elk. Journal of Wildlife Management 62:1020-1035 . Wakkinen, W.L. 1990. Nest site characteristics and spring-summer movements of migratory sage grouse in southeastern Idaho. M.S. Thesis. University of Idaho, Moscow, Idaho. Wakkinen, W.L., K.P. Reese, and J.W. Connelly. 1992. Sage grouse nest locations in relation to leks. Journal of Wildlife Management 56:381-383. Wallestad, R.O. 1971. Summer movements and habitat use by sage grouse broods in central Montana. Journal ofWildlife Management 35:129-136. • Wallestad, R.O. 1975. Male sage grouse responses to sagebrush treatment. Journal of Wildlife Management 39:482-484 . Wallestad, R.O., and D. Pyrah. 1974. Movement and nesting of sage grouse females in central Montana. Journal of Wildlife Management 35:129-136. White, G. C., and K. P. Burnham. 1999. Program MARK: Survival estimation from populations of marked animals. Bird Study 46 Supplement, 120-138. White, G.C., and R.A. Garrott. 1990. Analysis of wildlife radio-tracking data. Academic Press, New York City, New York, USA . • Wik, P .A. 2002. Ecology of greater sage-grouse in south-central Owyhee County, Idaho. M.S. Thesis, University of Idaho, Moscow, Idaho. Zablan, M.A., C.E. Braun, and G.C. White. 2003. Estimates of greater sage-grouse survival in North Park, Colorado. Journal of Wildlife Management 67:144-154 . • • ' • ~ ... .... 2150.-------------------------------------------------------------~ 2100 2050 a '--" ~ ·-ta 2000 0 > Q) ...... ~ 1950 1900 1850 i:>~<;\ \~~ ~~<;\ ~~ )]> +~~ ·~ ~~~ +~ ((.) \v~ \$> .N~~ ~v"'o ~· ~'\. ~ ~~ s((.)<:i, ~"' ~(55 ov ~"' ~ ~~ ..::;.o ~"' ~ CJ~ ~~ Figure 4.1. Mean monthly elevation and standard error of radio-marked greater sage-grouse females in the Axial Basin and Danforth Hills, Colorado in 2001 and 2002. ~ N N 'p I . ' C: .. ~ f."' 'I' fW' .. ~ • ,... • ' .. , ..,. ""' 0.94 0.92 0.90 ·~ ;::::$ C/1 0.84 0.82 0.80 Breeding Brood-rearing Fall Winter Season Figure 4.2. Annual survival model {S (T)}, time trend in seasonal survival of greater sage-grouse females in the Axial Basin and Danforth Hills, Colorado in 2001 and 2002. Survival rates and standard errors reported. ,_. N w 124 ,._ • Table 4.1. Median home-range estimates (95% fixed kernel, with least squares crossvalidation) and 95% confidence intervals around the median of greater sage-grouse broodrearing females in northwest Colorado, 2001-2002 . Category n Median (ha) Lower CI (95%) Upper CI (95%) Range Axial Basin females 15 1,151 199 3,334 470-4,186 Danforth females 14 439 160 1,129 124- 1,581 Yearlings 7 470 * * 124-4,186 Adults 22 543 199 1,379 135-3,343 * No confidence interval at this n exists .... >t .. 125 Table 4.2. Median annual home-range estimates (95% fixed kernel, with least-squares cross validation) and 95% confidence intervals around the median of greater sage-grouse females in northwest Colorado, 2001/2002 and 2002/2003. w ~ n Median (ha) Lower CI (95%) Upper CI (95%) Range Yearling 26 8,574 3,129 12,959 64821,533 Adult 43 6,556 4,277 11,554 1,24961,544 Category ~ .... .... . + ~ 126 ... Table 4.3. Annual and seasonal survival rates of radio-marked greater sage-grouse females in northwest Colorado 2001 and 2002. f. ~ Survival Rates Standard Error Breeding 0.91 0.05 Brood-rearing 0.84 0.06 Fall 0.89 0.06 Winter 0.96 0.04 Annual 0.65 0.08 Breeding 0.96 0.04 Brood-rearing 0.85 0.07 " Fall 0.95 0.04 .; Winter 0.90 0.06 Annual 0.71 0.09 Breeding 0.75 0.06 Brood-rearing 0.93 0.04 Fall 0.90 0.05 Winter 0.82 0.07 Annual 0.48 0.07 Breeding 0.84 0.08 Brood-rearing 0.94 0.06 Fall 0.87 0.09 Winter 1.00 0.00 Annual 0.78 0.10 Season ..,. 2001/2002 Adult (n = 35) ..., ~ ..... ... .... Yearling (n = 27) 2002/2003 Adult (n = 59) ~ IP' Ar II'" • Yearling (n = 18) .. 'I" . ~ • • r • ~ • ~ , .. .,. ~ ,. ~ ., l ' Table 4.4. Monthly survival estimates of adult greater sage-grouse females, 2001 and 2002, in northwest Colorado. 2001 (n = 35) Survival Estimate Standard Error April 0.97 May 2002 (n =59) 95% Confidence Interval Lower Upper Survival Estimate 0.03 0.82 1.00 0.92 0.94 .0.04 0.79 0.99 June 1.00 0.00 1.00 July 0.88 0.06 August 0.96 September Month Standard Error 95% Confidence Interval Lower Upper 0.04 0.81 0.96 0.85 0.05 0.73 0.92 1.00 0.93 0.04 0.81 0.98 0.71 . 0.95 0.95 0.03 0.83 0.99 0.04 0.79 1.00 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 0.95 0.04 0.82 0.99 October 0.89 0.06 0.71 0.96 0.85 0.04 0.81 0.99 November 1.00 0.00 1.00 1.00 0.97 0.03 0.81 1.00 December 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 January 1.00 0.00 1.00 1.00 0.88 0.06 0.71 0.95 February 0.96 0.04 0.75 0.99 0.91 0.06 0.70 0.98 March 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 'N """"' -.....l Table 4.5. Monthly survival estimates of yearling greater sage-grouse females, 2001 and 2002, in northwest Colorado. 2001 (n = 27) Survival Estimate Standard Error April 1.00 May 2002 (n = 18) 95% Confidence Interval Lower Upper Survival Estimate 0.00 1.00 1.00 1.00 1.00 0.00 1.00 1.00 June 0.92 0.05 0.74 July 0.88 0.07 August 1.00 September Month Standard Error 95% Confidence Interval Lower Upper 0.00 1.00 1.00 0.94 0.05 0.69 0.99 0.98 0.94 0.06 0.68 0.99 0.68 0.96 1.00 0.00 1.00 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 0.94 0.06 0.66 0.99 October 0.95 0.04 0.74 0.99 0.93 0.06 0.65 0.99 November 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 December 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 January 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 February 0.90 0.06 0.69 0.98 1.00 0.00 1.00 1.00 March 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 ~ N 00 129 Table 4.6. Breeding, brood-rearing, fall and winter survival models varying by year, age and year+ age for greater sage-grouse females in Moffat County, Colorado, 2001/2002. Models with 11 AICc ~ 2 are considered competing models. Season Model AICc 11 AICc # Parameters Deviance {S(a+t)}c 110.5 0.0 3 104.4 {S (t)} a 112.4 1.9 2 108.4 {S (a)} b 112.6 2.1 2 108.5 {S (.)} d 116.5 6.0 1 114.5 {S (t)} a 83.6 0.0 2 79.5 {S (.)} d 83.9 0.3 1 81.9 {S(a+t)}c 85.6 2.1 3 79.5 {S (a)} b 85.9 2.4 2 81.8 {S (.)} 68.1 0.0 1 66.0 {S (t)} a 69.9 1.8 2 65.8 {S (a)} b 70.0 1.9 2 65.9 {S(a+t)}c 71.9 3.8 3 65.7 {S (.)} a 65.5 0.0 1 63.5 {S (a)} b 66.0 0.5 2 61.9 f {S (t)} a 67.2 1.7 2 63.1 l {S (a+ t)} c 70.0 2.5 3 61.7 Breeding ..._ ,.., Brood-rearing ... i "'!( Fall t r Winter I ~ '• a variation due to year b Variation due to age c Variation due to year and age d No variation due to year or age 131 CHAPTER 5. FITNESS OF GREATER SAGE-GROUSE FEMALES ASSOCIATED WITH DISEASE AND PRE-NESTING BODY CONDITION IN NORTHWEST COLORADO INTRODUCTION Body Condition Condition is a term that describes the fitness of an individual to cope with its present and future needs (Owen and Cook 1977). Condition indices are used to adjust body weights ofbirds with structural size differences and should reflect the relative fitness of the individual (Bailey 1979). Individuals in good condition (those with high energy reserves) should exhibit greater probabilities of survival and reproduction than individuals in poor condition (Hepp et al. 1986, Reynolds et al. 1988). The proximate controls of reproduction operate through the energy available to reproducing birds thus the role of female condition is critical in determining reproductive effort (Krebs 2001). Remington and Braun (1988) investigated carcass composition of greater sage-grouse during the winter in North Park, Colorado. Greater sage-grouse gained weight over the winter, presumably for breeding and nesting activities in spring. Barnett and Crawford • (1994) linked maternal nutrition to productivity in greater sage-grouse, however they provided no specifics on blood parameters or body condition indices of females in good • .. condition. Beck and Braun (1978) conjectured that differential survival in greater sagegrouse might be weight-related. I assessed body condition of greater sage-grouse in northwestern Colorado by using a body condition index that adjusted for body size and a 132 biochemistry panel. These parameters were assessed for their influence on reproduction and survival. Biochemistry Few studies dealing with plasma biochemistry and physiology in wild birds can be found in the scientific literature. Most research papers have dealt with domestic or captive birds kept in zoos, rehabilitation facilities, or research centers (Sturkie 1976, Gee et al. 1981, Redig 1993). Normal reference values of blood chemical constituents are known only for 5% of bird species and have been studied mostly in captive situations (Ferrer 1993). Currently there are no published data indicating reference values for greater sage-grouse, thus this study and research being conducted in Oregon and Nevada (M. Dunbar, personal communications) will be important in defining these reference ranges for pre-laying females. The knowledge of normal reference values in plasma for wild species is important as it may provide insights into the health of individuals, quality of diet and suitability of habitat (Gavett and Wakeley 1986). Normal blood parameter reference values for endangered species involved in reintroduction or restoration programs are important in order to better understand the physiological status of the released birds. Additionally, serum chemistry is routinely used for the detection of organ disease in domestic mammals and is becoming more • common for avian patients (Campbell and Coles 1986). A biochemistry panel can help identify the occurrence and severity of diseases, or to confirm disease diagnosis (Ritchie et • al. 1994). Factors such as age (Sturkie 1976), sex (Perry et al. 1986), captivity (Dobado-Berrios ,. et al. 1998), diet (Gavett and Wakeley 1986, Ferrer 1993) and reproductive status (Beckerton and Middleton 1982, Cain et al. 1982, White et al. 1987) can influence the total variation 133 found in metabolites, electrolytes and other organic molecules. However, presently these factors are poorly understood due to difficulty of gathering information on different ageclasses in wild species. Research has also indicated that the physiological condition of birds can influence serum chemical constituents (Dolnik 1973, Driver 1981). Therefore, the biochemistry panel of female greater sage-grouse entering the reproductive period may be an indicator of physiological condition and thus may influence productivity or survival rates. Diseases Serology tests examine contagious diseases transmissible to and from domestic fowl. Little is known how or if these domestic diseases affect wild avian populations. Avi an influenza can cause anorexia, labored breathing, diarrhea, swelling and death in game birds (Wobeser 1997, Keramas 2002). Salmonella spp. can cause emaciation, respiratory problems, loss of coordination and death in birds (Steele and Galton 1971 ). Three important Mycoplama species known to poultry are M gallesepticum, M synoviae, and M meleagridis. • Transmission of Mycoplasma spp. occurs through direct contact or airborne droplets and can be transferred through the eggs from female to progeny (Wobeser 1997). M gallesepticum is a known pathogen of upland game birds raised in captivity and causes airsacculitus in chickens (Gallus domesticus) and sinusitis in turkeys (Yoder 1991). M synoviae causes respiratory disease and synovitis in domestic poultry (Kleven et al. 1991 ). M meleagridis has been know to affect the respiration and growth in juvenile turkeys (Yamamoto 1991). M. meleagridis occurs in wild turkeys, however, reports of occurrence in other upland game birds have not been confirmed. It has been speculated that Mycoplasma spp. could cause major die-offs and impaired reproduction and survival in sage-grouse (Braun et al. 1994). 134 As available habitat decreases in size and animals are forced into areas of high densities, disease outbreak may become more prevalent (Braun et al. 1994). Additionally, the gregarious nature of greater sage-grouse could facilitate disease transmission. A reduced body condition may make a host slower, weaker and thus more vulnerable to other mortality factors (Atkinson and Van Riper ill 1991, Hudson and Dobson 1991). Diseases, body condition and serum chemistry of greater sage-grouse in northwest Colorado are unknown. My objectives were to (1) investigate pre-breeding season body condition through assessment of a condition index and a biochemistry panel, (2) estimate disease prevalence, and (3) relate body condition and disease prevalence to reproductive parameters and survival rates during breeding and brood-rearing . • METHODS Females were captured from 27 March to 26 April in 2001 and from 25 March to 12 April2002 using spot-light techniques and a CODA net-launcher. At the time of capture, 2 ml ofblood were taken from 81 females (43 adults and 38 yearlings) and 34 males. Bleeding was done from the cutaneous ulnar vein using 22 and 26-gauge needles attached to a 3 cc syringe (Campbell and Coles 1986). Prior to puncture, alcohol was used to clean the skin. Blood s3mples were placed into sterile tubes; red-topped tubes without anticoagulant and lithium-heparin tubes. Serum for the biochemical analysis was centrifuged at 3000 rpm for 10 minutes and frozen until analysis at Washington State Clinical Pathology Laboratory. The blood chemistry profile indicated aspartate amino transferase (AST), alkaline phosphatase .. (AP), creatinine phosphokinase (CK), cholesterol, glucose, total protein, creatinine, phosphorous, sodium, potassium, chloride, and uric acid. Calcium was measured only in 135 2002. The serum for disease analysis was sent both as fresh and frozen samples to the Colorado Veterinary Diagnostic Laboratory at Colorado State University to determine whether storage methods would alter disease results. Diseases were screened using a plate agglutination test. Additionally, 15 samples were sent to the Animal Health -Laboratory of the Idaho Department of Agriculture, Boise. Both male and female grouse were tested for avian influenza, S. pullorum/ S. typhoid, M gallisepticum, M synoviae, and M meleagridis using the plate agglutination tests. Morphological characteristics of females were measured at time of capture and included mass and keel length. Mass was measured using 5000g electronic scales. Keel bone length was measured using calipers. I monitored females until nesting was evident (Chapter 1). When a nest was found • mid-laying, I determined the date of clutch initiation by backdating from the hatch date and using a laying rate of 2 eggs per 3 days (Schroeder et al. 1999). A hand-held Y agi antenna, attached to a receiver/scanner, was used to monitor the radio-tagged grouse. Each female was monitored 1-2 times a week throughout the spring and summer to 31 August to estimate breeding and brood-rearing survival. Reproductive Parameters Body Condition Index. No linear relationship existed between body mass and days to initiate egg-laying (F = 0.40, P = 0.529, r 2 = 0.01) therefore, to estimate pre-laying body condition, I used the equation body mass*keel-3, defined by Johnson and Boyce (1990). The influence ofbody condition on nest initiation and nest success was estimated using ANOVA. Analysis of the relationship between body condition indices and initiation date, clutch size, and hatch size 136 -r I (number eggs which hatched) was estimated by linear regression (McClave and Dietrich 1994). Biochemistry. Simple linear regression was used to estimate whether there was a trend in biochemical values as hens approached nest initiation (range: 3-43 days). Serum chemical I~ constituents were compared between ages (yearling or adult) 1 years, capture technique and I ~- ' I~ the interactions among variables using a 3-way factorial multivariate analysis of variance (MANOVA). Missing values were replaced by the mean value for that variable (Ott and I ._ Longnecker 2001). Calcium was not included in this analysis, as values had only been r collected in 2002. Evaluation of univariate ANOVA and canonical analysis was used to I r~ determine which variables contributed to overall differences. Protein, phosphorous, AP, creatinine, calcium and cholesterol have been associated with egg-laying in birds (Ritchie et al. 1994). Forward step-wise logistic regression was used to model the importance of these serum chemical constituents in predicting each of 2 binary variables; nest initiation and nest success (Beier 1989). Multiple regression models were used to predict the importance of the biochemistry panel in predicting clutch initiation date, clutch size and hatch size. A significance level of P _::: 0.10 was used to determine which variables entered and remained in the model. Calcium was analyzed separately as only 1 year of data was available. Nest initiation and success were analyzed using ANOVA and simple linear regression was used to analyze clutch initiation and clutch size. All analyses were performed in SAS 8.2 (SAS Institute 1999). I~ 137 L Diseases. Relationship of prevalence of disease and nest initiation and nest success was determined through a chi-squared goodness of fit test (McClave and Dietrich 1994). Associations between nest initiation date and clutch size were determined through t-tests 1 ,~ (Snedecor and Cochrane 1980). Forward step-wise logistic regression was used to model the importance of the body condition index or serum chemical constituents in predicting disease presence or absence. Age and year were included as class variables. Survival Body Condition. Female survival was analyzed for breeding and nesting (1 April to 15 June) and brood-rearing (16 June to 31 August). To assess whether variation in survival coincided with biochemical values, body condition indices, or disease prevalence, I used individual encounter histories with covariates and incorporated them into the design matrices in • program MARK (White and Burnham 1999). Missing individual covariate values were assigned the mean of the variable (Ott and Longnecker 2001). The QAICc statistic for small populations was used to select the best model among alternative parametric models (Burnham and Anderson 1998). Assessment of the data showed evidence of overdispersion (c = 1.38) (Anderson et al. 1994). A model was first selected to determine whether survival ratios differed among ages, seasons, years, or none of these parameters. Next, a set of a priori models were developed under the guidance of curren~ avian nutrition literature models. They were constructed upon the best of the aforementioned models and incorporated the body condition index and blood parameters as covariates in design matrices. 138 The variables that were included in the models included the body condition index, total protein, uric acid, cholesterol, glucose, potassium, sodium and phosphorous. Serum protein levels for bird species range between 2-6 grams/ 100ml (Campbell and Coles 1986, Redig 1993, Lumeij 1997). Low protein levels can indicate chronic disease, malnutrition, malabsorption, chronic blood loss and starvation. Elevated protein levels indicate dehydration, shock, chronic disease, trauma or infection (Pesek 1996). Mean protein value for greater sage-grouse hens in Oregon and Nevada was 5.3 g/dl (M. Dunbar, personal communications). Age and diet may influence uric acid concentration in the blood, a product of nitrogen and protein catabolism (Ritchie et al. 1994). Cholesterol concentrations vary with diet, particularly protein uptake (Ritchie et al. 1994). Potassium is necessary for muscle activity, and protein synthesis (Redig 1993). Elevated glucose levels can indicate malnutrition, bacterial infections, hormonal problems, stress, and diabetes (Pesek 1996). Both sodium and potassium help maintain the body's fluid balance (Ritchie et al. 1994). Mean phosphorous level for pre-laying hens in Oregon was 7.3 mg/dl (n = 144) (M. Dunbar, personal communication). Elevated levels can indicate renal disease, and starvation (Campbell and Coles 1986). Survival rates were estimated using program MARK (Chapter 4) for the first month after capture for individuals captured by the CODA net-launcher and night spotlighting. Models were analyzed to determine whether capture technique influenced survival. AICc was used to select the best model as no overdispersion was detected in this data set (c < 1.00). r f~ I I~ 139 Diseases Survival models were also estimated using program MARK for those individual greater sage-grouse tested forM synoviae. The association of disease prevalence with survival was tested against the model in which survival did not vary with disease prevalence. QAICc was used to select the best model as the data were adjusted for overdispersion (c = 1.33). The model with the lowest QAICc value was chosen as the model which best fit the data. RESULTS Reproductive Parameters Body Condition Index. . The mean body condition index in 2001 was 0.0015, SD = 0.003 (n = 43), which did not differ (F = 0.76, P = 0.388) from 2002 (0.0015, SD = 0.002, n = 38). The mean body condition index for adults was 0.0015, SD = 0.0003 (n = 47), and did not differ (F= 2.16, P = 0.147) from that for yearling females (0.0014, SD = 0.0002, n = 34). Body condition did not influence nest initiation rates (F = 0.51, P = 0.477, n = 73) or nest success (F = 0.42, P = 0.520, n = 60). Additionally, clutch size, hatch size and nest initiation dates were not related to the body condition index (Table 5.1 ). •· )- Biochemistry. Simple linear regression indicated no trend in serum biochemistry as females neared nest initiation (r2 < 0.40). Therefore, females were pooled, regardless of breeding stage. No differences in the biochemistry panel were detected for year* age* capture (Wilk's A-= 0.89; F = 0.73; 12,68 df; P = 0.717), year* age (Wilk's 'A= 0.79; F= 1.48; 12,68 df; P = 0.153), year*capture (Wilk's A= 0.58; F= 0.87; 12,68 df; P = 0.577), age*capture (Wilk's A= 0.91; 140 y [.._ I,. • F = 0.57; 12,68 df; P = 0.856) or age (Wilk's A= 0.87; F = 0.84; 12,68 df; P = 0.605). There was, however, a difference detected between years (Wilk's A= 0.43; F = 7.25; 12,68 df; P < . 0.001). Creatinine (F= 49.49, P < 0.001), total protein (F=24.88, P < 0.001), cholesterol (F = 16.63, P < 0.001), and AP (F = 5.70, P = 0.019) were the factors driving the difference between years (Table 5.2). All values were greater in 2002 than in 2001. Additionally, analysis revealed differences due to capture techniques (Wilk's A= 0.65; F = 3.08; 12,68 df; P = 0.002). Trapping with the CODA net-launcher resulted in elevated levels of CK (F = 7.01, P = 0.010) and chloride (F = 12.05, P = 0.001) (Table 5.3). None of the biochemical constituents entered the logistic regression equations as significant predictors of nest initiation or success. Similarly, none of the variables entered the multiple regression equations as significant predictors of initiation date, clutch size or hatch size. Calcium was not associated with nest initiation (F = 1. 07, P = 0.311, n = 26), clutch initiation date (F = 0.03, P = 0.85, / = 0.002, n = 19), nest success (F = 0.46, P = 0.803, n = 22), or clutch size (F = 1.14, P = 0.30, r2 = 0.081, n = 15). Diseases. There was no difference (P = 1.000) in fresh an~ frozen sample results forM. gallesepticum (n = 22), M meleagridis (n = 22), S. pullorum/ S. typhoid (n = 20), or avian influenza (n = 25). There was 89% consistency for results of fresh and frozen samples tested forM synoviae (n =18). Serum samples were divided and sent to 2 separate labs to determine the consistency in results. There was 100% consistency forM gallesepticum (n = · 15), M. meleagridis (n = 15), S. pullorum/ S. typhoid (n = 15) and avian influenza (n = 15) and a 93% consistency forM synoviae (n = 15) . 141 The serology analysis in 2001 indicated a 55% occurrence of M synoviae for females (n = 31) and 92% presence for males (n = 12). In 2002, the occurrence of M synoviae was only 12% (n = 33) and 6% (n = 18) for females and males, respectively. There was no difference between prevalence in adult females (45%, n = 38) and yearling females (67%, n = 21) (t = -0.73, P = 0.392) over the 2 years. The tests were negative forM gallesepticum (n = .. 54), M meleagridis (n =56), S. pullorum/ S. typhoid (n = 47), and avian influenza (n =52) in females over 2 years. Similarly, males tested negative forM gallesepticum (n = 24), M meleagridis (n = 24), S. pullorum/ S. typhoid (n = 26), and avian influenza (n = 29). Nest initiation, nest success, clutch initiation date and clutch size were not influenced by the presence of M synoviae (Table 5.4). Forward stepwise logistic regression found cholesterol, sodium and year as significant predictors of positive disease results. The logistic regression equation that resulted was Logit (Y) = 33.36 - 0.02 (cholesterol)- 0.20 (sodium)+ 0.91 (year)+ e (Table 5.5). Cholesterol and sodium was greater in grouse negative forM. synoviae (163.06 mgldl, SE = 5.98; 155.37 mEq/1, SE = 0.71, n = 30) than those that tested positive for the disease (143.88 mgldl, SE = 10.36; 154.14 mEq/1, SE = 0.82, n = 14). Survival Body Condition. The model which best fit the data from the simple models of survival varying with age, season, year or none of these, was the model varying with age (Table 5.6). Therefore, subsequent complex models were built upon the age parameter. There were several competing models which best fit the data(~ QAICc ~ 2). These included survival varying with age + uric acid, survival varying with age + body condition index, and survival varying 142 . with age+ protein (Table 5.6). Survival rates in the first month post-capture were 0.90, SE = 0.03, n = 116 for grouse captured at night and 0.96, SE = 0.04, n = 26 for grouse captured by the CODA net-launcher. Capture effects did not explain the data(~ AICc 2: 2) better than the.null model. Diseases. The model which best fit the data for females tested forM synoviae was the null model, survival that did not vary with disease prevalence. The difference in QAICc was just slightly > 2, indicating there is weak evidence that the 2 models are not competing (Burnham and Anderson 1998) (Table 5.7). DISCUSSION • Body Condition Index My results detected no relationships between body condition indices, nest initiation rate, nest success, initiation date, clutch size or hatch size. These may not have been detectable as the range for body condition indices for yearling (0.0011 - 0.0018) and adult females (0.0010- 0.0020) was narrow. Similarly, no correlation was evident between body I~ . con9ition (adjusted for reproductive status and body size) and clutch initiation dates and clutch size for willow ptarmigan (Lagapus lagapus) populations in British Columbia and Manitoba (Robb et al. 1992). Additionally, no associations were made between clutch size and hen weights for ruffed grouse or prairie chickens (Tymphanuchus cupido) in Minnesota (Maxson 1974, Bergerud 1988). Alternatively, Breitenbach et al. (1963) showed concomitance between female weight and clutch initiation date and clutch size for ring-necked pheasants. Winter weight of 143 pheasants in Ohio was also associated with productivity (Edwards et al. 1964). Likewise, research by Porter et al. (1983) on turkeys in Minnesota indicated the weight of females to be highly related to productivity. A study on white-tailed ptarmigan (Lagopus leucurus) indicated female body condition was associated with the length and frequency of incubation breaks, presumably associated with nest success (Weibe and Martin 1997). Studies have indicated, and my research supports the notion (Chapter 4), that greater sage-grouse suffer the highest natural mortality in the spring and summer (Wallestad 1975, Connelly et al. 1994, Connelly et al. 2000). Beck and Braun (1978) showed that greater sage-grouse females gained weight from September to April. My data illustrated that the spring condition of greater sage-grouse females, based on a body condition index, did not limit reproduction in • 2001 or 2002. Pre-breeding body condition index was a predictor of survival through the breeding and brood-rearing periods. Kabat et al. (1956) observed a direct correlation between fat stores and body weight and the ability of female ring-necked pheasants to resist stress. Similarly, heavier female pheasants and turkeys showed higher survival rates (Edwards et al. 1964, Porter et al. 1983). Robb et al. (1992) observed no relationship between willow ptarmigan female body condition and reproductive parameters but noted females in poor condition exhibited lowered survival probabilities. Similarly, greater sage-grouse in poor condition pre-breeding appear to invest fully in reproduction, at the expense of reduced .. - survival during breeding and brood-rearing. Biochemistry • The biochemistry panel for females in northwest Colorado was comparable to values collected in Oregon and Nevada except for AST and uric acid levels (M. Dunbar, personal 144 communication) (Table 5.2). Uric acid levels for my population of greater sage-grouse fell within the range of mean uric acid levels for 12 species of captive birds at the Patuxent Wildlife Research Center (6.9- 12.4 mg/dl) (Gee et al. 1981). AP, cholesterol, creatinine, . sodium, potassium and chloride are elements with which no reference values exist for greater sage-grouse. I compared these values to the range of mean values acquired for the 12 captive species (Gee et al. 1981). All my results fell within the range of means reported except chloride. The levels of chloride for the 12 captive species ranged from 99- 112 mEq/1, while the mean value for greater sage-grouse females was 119 mEq/1. Chloride levels fall within a narrow range in avian species (Ritchie et al. 1994) and reasons for the elevated levels are unknown. My results indicated that biochemistry constituents did not vary between ages, but • showed disparity due to year and capture technique. Creatinine, cholesterol, protein and AP exhibited higher levels in 2002 than in 2001. Creatinine is minimally affected by dietary or tissue proteins and is insensitive as a diagnostic test (Ritchie et al. 1994). Elevated creatinine levels have been described in connection with egg-laying (Ritchie et al. 1994). Age, heredity, nutrition and diseases affect cholesterol levels in avian blood, which range from 100 to 200 mg/dl in healthy birds • (Campbell and Coles 1986, Redig 1993). Cholesterol concentrations vary with a bird's diet and are linked to diets higher in protein sources (Ritchie et a~. 1994). Low protein levels can indicate chronic disease, malnutrition, malabsorption, chronic blood loss and starvation. Total protein concentrations in female birds increase just prior to egg-laying (Lumeij 1997). In mourning doves (Zenaida macroura) cholesterol and protein concentrations were associated with attainment of puberty and reproductive potential (White et al. 1987). In birds, increased AP activity is associated with increased osteoblastic activity such as skeletal 145 .. growth and impending ovulation (Campbell and Coles 1986, Lumeiij 1997). AP levels may be elevated due to irritation of cells in different tissues yet have no specific importance (Ritchie et al. 1994). The elevated biochemical levels in 2002 may be associated with egg-laying or diet. January- March of2002 experienced 1/3 less precipitation in the form of snow than 2001 (Appendix 1-B). Thus vegetation and invertebrates were earlier to emerge in 2002. For this reason, although sampling dates were relatively consistent among years, biochemistry values may have increased in association with forbs and invertebrates in the diet. Additionally, clutch initiation was earlier in 2002 (Chapter I) and thus females may have been at a later ,. reproductive stage although they were sampled at equivalent dates. Differences related to capture technique may be due to variation associated with diurnal changes in blood (Dolnik 1973) or differential stress. Creatinine kinase (CK) and chloride levels were greater for females captured by the CODA net-launcher than those captured by spot-light trapping. The physiological influence of serum chloride is poorly understood in birds, however ranges are typically narrow (Ritchie et al. 1994). Chloride levels may not have any biological relevance to capture for greater sage-grouse, however, 1: elevated chloride levels due to the stress of capture have been reported in grizzly bears ( Ursos arctos) (Brannon 1985). Elevated levels of creatinine kinase may be as a result of the excitement of handling (Lumeij 1997) or due to elevated exercise prior to handling. Plasma -. r.- CK levels in turkeys were sensitive to handling and stress (Lumeij 1997). It is doubtful that CK levels are associated with the time of capture as elevations are typically seen due to muscle cell damage (Ritchie et al. 1994). The CODA net-launcher was deployed on leks within the first hours of daylight and was set to capture females surrounding the dominant - - - - - - - - - - - -- - -- - - - - -- 146 11"' male. Mean number of individuals captured was 3.1 (range: 1-10, n = 8), thus individual processing time was lengthened and may have produced increased stress in grouse. Despite, the apparent increase in stress due to use of the CODA net-launcher, the use of this technique did not influence survival rates within 1 month post-capture. None of the biochemistry elements measured were significant predictors of any of the reproductive parameters measured. For black grouse (Tetrao tetrix) losses during incubation are most likely attributed to predation, while factors such as climate, maternal condition, or nest site availability are of minor importance (Angelstam 1984). Contrastingly, Wittenberger (1978) argued that viability of grouse eggs and chicks are correlated with condition of .... incubating hens, presumably associated with spring forage conditions. Greater sage-grouse in the Axial Basin and Danforth Hills exhibited above average nest success and comparable clutch sizes to other populations of greater sage-grouse across the range of the species (Chapter 1). Given these results, it does not appear that pre-breeding body condition is limiting reproduction. Survival models with the biochemistry panel as a gauge of pre-breeding condition • I~ may give an indication of the diet and habitat available to females. The models indicated that protein and uric acid were important parameters in greater sage-grouse survival. Protein l._ levels are greatest in forbs (Peterson 1970) and invertebrates and have been used to diagnose malnutrition (Perry et al. 1986). Protein deficiencies resulted in decreases in body mass of northern bobwhite and scaled quail (Giuliano et al. 1996). Plasma uric acid is the major nitrogen excretory product in birds and is an index of protein catabolism. Serum concentrations are coupled with protein utilization and long-term fasting in avian species (Cherel et al. 1987, Boismenu et al. 1992). Although I did not directly study the diet of the 147 greater sage-grouse population in the Axial Basin and the Danforth Hills, it was observed that total serum protein and uric acid were important in modeling survival. Uric acid and protein levels may be associated with forbs and invertebrates in the diet and could potentially be influencing survival of greater sage-grouse females during breeding and brood-rearing. Diseases There was an 89% consistency for results of fresh and frozen samples tested forM synoviae and a 100% consistency for results of the other diseases tested. To validate laboratory results, serum samples were divided and sent to 2 separate laboratories. There was a 93% consistency forM synoviae and a 100% consistency for results of the other diseases tested. The serology analysis in 2001 indicated a 55% occurrence of M. synoviae for females and 92% presence for males. In 2002, the laboratories were asked to test initially forM synoviae and then test for the remaining pathogens. The occurrence of M. synoviae was only 12% and 6% for females and males, respectively. Three females were re-captured in 2002, and all were consistently negative forM synoviae between years. No other pathogens tested for were positive over the duration of the study. Both laboratories used a plate agglutination screening test provided by the same company. This test is not as accurate as ·a hemoagglutination inhibition test (K. Eyre, personal communication). The assessment works on the premise of antigen compliment binding with the disease and has 85% specificity and 91% sensitivity. M synoviae is smaller than the other Mycoplasma spp., thus making it more susceptible to false positives when the antigen creates a clumped lattice with red-blood cells, rather than the pathogen. Additional testing (hemoagglutination inhibition test) and culturing samples from individuals testing positive for the disease would be required in order to accurately determine disease prevalence. -------------------------------------------------------------------------- -- 148 Braun et al. (1994) speculated that Mycoplasma could potentially cause impaired reproduction and survival in sage-grouse. My results failed to confirm this, although the statistics were based on questionable disease incidence. I found no relationship between M synoviae and nest initiation, clutch initiation date, clutch size or nest success. Ritchie et al. (1994) found that the majority of the reproductive system in chickens function similarly in Mycoplasma infected birds as those without Mycoplasma . . Survival of greater sage-grouse during the breeding and brood-rearing period also did not vary with occurrence of M synovia e. 1>- I~ , .. Blood biochemistry values were used to understand the physiological status of the grouse infected with M synoviae. Sodium and cholesterol levels were the only predictors of positive disease results and the levels were less in grouse testing positive forM. synoviae. Sodium values are maintained within narrow limits in birds and thus differences may indicate a pathological condition, however this is rarely seen in avian patients (Ritchie et al. 1994). Cholesterol concentrations are usually lower in stressed mammals but have been reported as elevated in stressed canvasbacks (Ayathya valisineria) (Perry et al. 1986). Assuming that individuals positive forM synoviae had greater physiological stress, they exhibited the same trends in cholesterol as mammals. However, the diagnostic value of sodium and cholesterol in birds is poor (Ritchie et al. 1994), and these statistics are based on inconclusive evidence of disease, thus these results should be viewed cautiously. No clinical symptoms of M synoviae were detected during handling. Hoffinan et al. (1997) were also not able to confirm the presence of M gallesepticum or M. synoviae through culturing for wild turkeys in west-central Colorado and concluded that false positives had been reported. Forty-three percent of their birds tested positive for either of these 149 pathogens (n =51). Given there was 0 occurrence for avian influenza, S. pullorum/ S. typhoid, M gallisepticum, and M meleagridis, these diseases appear not to be a major threat to the greater sage-grouse populations in the area. Possible explanation for the inconsistency in M synoviae from 2001 to 2002 could be an indication that the tests were giving a higher percentage of false positives in 2001 due to different sampling protocol, or the disease was present at a higher level in 2001. The absence of clinical signs, large variation in disease prevalence between years, inconsistency in results due to freezing and between labs, the variable nature of the plate agglutination test, and no influence on reproductive and survival parameters suggests incidence of M synoviae in greater sage-grouse females is inconclusive. Analysis of biochemical values indicated that birds positive forM synoviae might have experienced physiological affects, although sodium and cholesterol have poor diagnostic value and these statistics are based on uncertain results. Due to these confounding results, I . recommend that future research examine the physiological response of greater sage-grouse females experimentally infected with M synoviae. MANAGEMENT IMPLICATIONS Reference serum chemical ranges of greater sage-grouse are important for predicting physiological and pathological condition of birds. Future research needs to estimate serum chemical ranges for chicks, males and females year-round. Measures of body condition did not appear related to reproductive parameters but were associated with survival. Particularly important predictors of survival were the body condition index, uric acid and protein levels, factors that may fluctuate with diet. Invertebrate abundance varies with herbaceous cover (Jamison et al. 2002) and precipitation (Crawford 1981), thus limitations of invertebrate 1._ 150 biomass due to drought or habitat quality may lead to reduced survival of greater-sage grouse during the reproductive period. This re-emphasizes the importance of maintaining native forb species in pre-breeding and breeding habitats (Chapter 1, Chapter 3). With increasing attention on handling potentially threatened or endangered sagegrouse species, biologists need to be aware of acute stress on trapped birds, particularly those considered for translocation or detained for study. Use of the CODA net-launcher as a capture method appears to be more stressful than night spotlighting on individuals although effects may be short-term as there did not appear to be an association with capture technique and survival of individuals 1 month post-capture. However, influences of such stress on the survival and success of translocated grouse is unknown. The population of greater sage-grouse in the Axial Basin and Danforth Hills does not appear to be infected with avian influenza, Salmonella pullorum/ S. typhoid, Mycoplasma gallisepticum, or M melgridis. Although individuals tested positive forM synoviae, I am unable to say that they were infected with the disease. To determine any impacts of the disease, individuals need to be experimentally injected with the pathogen and then associated with body condition, serum chemical constituents and fitness measures. 151 LITERATURE CITED Anderson, D.R., K.P. Burnham, and G.C. White. 1994. AIC model selection in overdispersed capture-recapture data. Ecology 75:1780-1793. Angelstam, P. 1984. Sexual and seasonal differences in mortality in the black grouse (Tetrao tetrix) in boreal Sweden. Ornis Scandinavica 15:123-134. Atkinson, C.T., and C. Van Riper Ill. 1991. Pathogenicity and epizootiology ofhaematazoa. Pages 19-48 in J. E. Loyle and M. Zuk, eds. Behaviour and Evolution of Bird-Parasite Interactions, Oxford: Oxford University Press. Bailey, R.O. 1979. Methods of estimating total lipid content in the redhead duck (Aythya american) and an evaluation of condition indices. Canadian Journal of Zoology 57:18301833. Barnett, J.K., and J.A. Crawford. 1994. Pre-laying nutrition of sage grouse hens in Oregon. Journal ofRange Management 47:114-118. Beck, T.D.I., and C.E. Braun. 1978. Weights of Colorado sage grouse. Condor 80:241-243. Beckerton, P.R., and A.L.A. Middleton. 1982. Effects of dietary protein levels on ruffed grouse reproduction. Journal of Wildlife Management 46:569-579. Beier, P. 1989. Use of habitat by mountain beaver in the Sierra Nevada. Journal of Wildlife Management 53:649-654. Bergerud, A.T. 1988. Population ecology of North American grouse. Pages 578-648 in A.T. Bergurud and M.W. Gratson, eds. Adaptive strategies and population ecology of northern grouse. University of Minnesota Press, Minneapolis. Boismenu, C., G. Gauthier, and J. Larochelle. 1992. Physiology of prolonged fasting in greater snow geese (Chen caerulescens atlantica). Auk 109:511-521. Brannon, R.D. 1985. Serum chemistry of central and northern Alaska grizzly ~ears. Journal of Wildlife Management 49:893-900. • .. Braun, C.E, M.F. Baker, T.E. Remington, and W.D. Snyder. 1994. Upland bird management analysis guide, 1994-1998. Colorado Division of Wildlife, Fort Collins . Breitenbach, R.P., C.L. Nagra, and R.K. Meyer. 1963. Effect of limited food-intake on cyclic annual changes in ring-necked pheasant hens. Journal of Wildlife Management 27:24-36. 152 Burnham, K.P., and D.R. Anderson. 1998. Model selection and inference- a practical information-theoretic approach. Springer-Verlag New York Inc., New York. 353pp. Cain, J.R., S.L. Bearsom, L.O. Rowland, and L.D. Rowe. 1982. The effects of varying dietary phosphorus on breeding bobwhites. Journal ofWildlife Management 46:10611065. .. Campbell, T.W., and E.H. Coles. 1986. Avian clinical pathology. Pages 279-301 in Coles, E.H, ed. Veterinary clinical pathology 4th edition, W.B. Saunders Company, Philadelphia. Cherel, Y., J.C. Stahl, and Y.L. Maho. 1987. Ecology and physiology of fasting in king penguin chicks. Auk 104:254-262 Connelly, J.W., M.A. Schroeder, A.R. Sands, and C.E. Braun. 2000. Guidelines to manage sage grouse populations and their habitats. Wildlife Society Bulletin 28:967-985. Connelly, J.W., K.P. Reese, W.L Wakkinen, M.D. Robertson, and R.A. Fischer. 1994. Sage grouse ecology report. Idaho Department ofFish and Game Job Completion Report. W160-R -19. Subproject 9. Boise. Crawford, C.S. 1981. Biology of desert invertebrates. Springer-Verlag, New York, N.Y. 314 pp. Dobado-Berrios, P.M., J.L. Tella, 0. Ceballos, and J.A. Donazar. 1998. Effects of age and captivity on plasma chemistry values of the Egyptian vulture. Condor 100:719-725. Dolnik, T.V. 1973. Diurnal and seasonal cycles of the blood sugar in sedentarty and migrating birds. Zoologicheskii Zhumal. 52:94-103. 1.... Driver, E.A. 1981. Hematological and blood chemistry values of mallard, Anas p. platyrhynchos, drakes before, during and after remige moult. Journal of Wildlife Diseases 17:413-421. Edwards, W.R., P.J. Mikolaj, and E.A. Leite. 1964. Implications from winter-spring weights of pheasants. Journal ofWildlife Management 28:270-279. 1 .. Ferrer, M. 1993. Blood chemistry studies in birds: some applications to ecological problems. Pages 1031-1044 in Trends in Comparative Biochemistry & Physiology. Council of Scientific Research Integration, Trivandrum. India. ... • • Gavett, A.P., and J.S. Wakeley. 1986. Blood constituents and their relation to diet in urban and rural house sparrows. Condor 88:279-284. 153 Gee, G.F., J.W. Carpenter, and G.L. Hensler. 1981. Species differences in hematological values of captive cranes, geese, raptors and quail. Journal of Wildlife Management 45:463-483 . Giuliano, W.M., R.S. Lutz, and R. Patino. 1996. Reproductive responses of adult female northern bobwhite and scaled quail to nutritional stress. Journal of Wildlife Management 60:302-309. Hepp, G.R., R.J. Blohm, R.E. Reynolds, J.E. Hines, and J.D. Nichols. 1986. Physiological condition of autumn-banded mallards and its relationship to hunting vulnerability. Journal ofWildlife Management 50:177-183. Hoffman, R.W., M.P. Luttrell, W.R. Davidson, and D.H. Ley. 1997. Mycoplasmas in wild turkeys living in association with domestic fowl. Journal of Wildlife Diseases 33:526535. Hudson, P.J., and A.P. Dobson. 1991. The direct and indirect effects of the caecal nematode Trichostygongylus tenuis on red grouse. Pages 49-68 in J. E. Loyle and M. Zuk, eds. Ecology, behaviour and evolution of bird-parasite interactions, Oxford Oxford University Press. Jamison, B.E., R.J. Robel, J.S. Pontius, and R.D. Applegate. 2002. Invertebrate biomass: associations with lesser prairie-chicken habitat use and sand sagebrush density in southwestern Kansas. Wildlife Society Bulletin 30:517-526. Johnson, G.D., and M.S. Boyce. 1990. Feeding trials with insects in the diet of sage grouse chicks. Journal of Wildlife Management 54:88-91. Kabat, C.R., R.K. Meyer, K.G. Flakas, and R.L. Hines. 1956. Seasonal variation in stress resistance and survival in the hen pheasant. Wisconson Conservation Deptartment. Technical Wildlife Bulletin 2, Madison. 48pp. Keramas, A. 2002. Natural ~ases and an experimental study ofH9N2 avian influenza in commercial broiler chickens of Iran. Avian Pathology. 31:247-252. ., .... • Kleven, S.H., G.N. Rowland, and N.O. Olson. 1991. Mycoplasma synoviae infection. Pages 223-231 In Diseases ofPoultry, 9th ed., B.W. Calnek, H.J. Barnes, C. W. Beard, W. M. Reid, and H. W. Yoder, Jr. (eds.). Iowa State University Press, Ames, Iowa. Krebs, C.J. 2001. Ecology: The experimental analysis of distribution and abundance. 5th ed. Benjamin Cummings, San Francisco, California. Lumeij, J.T. 1997. Avian clinical biochemistry. Pages 857-883 in J. J. Kaneko, ed. Clinical biochemistry of domestic animals. Fifth edition, Academic Press, New York. 154 • • Maxson, S.J. 1974. Activity, home range, and habitat usage of female ruffed grouse during the egg-laying, incubation, and early brood periods as determined by radio telemetry. M.S. thesis. University of Minnesota, St. Paul. McClave, J.T., and F.H. Dietrich II. 1994. Statistics, Sixth edition. MacMillan College Publishing Company, New York, NY. Ott, R.L., and M. Longnecker. 2001. An introduction to statistical methods and data analysis. Fifth edition. Duxbury, Pacific Grove, CA. 1152pp. l. Owen, M., and W.A. Cook. 1977. Variations in body weight, wing length and condition of mallard Anas platyrynchos and their relationship to environmental changes. Journal of Zoology 183:377-395. Perry, M.C., H.H. Obrecht, B.K. Williams, and W.J. Kuenzel. 1986. Blood chemistry and hematocrit of captive wild canvasbacks. Journal of Wildlife Management 50:435-441. Pesek, L. 1996. "The Aviary." [http://theaviary.com/s1295-52.shtml]. Peterson, J.G. 1970. The food habits and summer distribution of juvenile sage grouse in central Montana. Journal of Wildlife Management 34:147-155. Porter, W.F., G. C. Nelson, and K. Mattson. 1983. Effects of winter conditions on reproduction in a northern wild turkey population. Journal of Wildlife Management 47: 281-290. Redig, P. 1993. Medical management ofbirds of prey. The Raptor Center, St. Paul, MN, 173 pp. Remington, T.E., and C.E. Braun. 1988. Carcass composition and energy reserves of sage grouse during winter. Condor 90:15-19. Reynolds, C.J., P. Angelstam, and S. Redpath. 1988. Predators, their ecology and impact on gamebird populations. Pages 72-97 in P.J. Hudson and M.R.W. Rands, eds., Ecology and Management of Gamebirds. BSP Professional Books, Oxford, England. 263 pp. -t Ritchie, B.W., G.J. Harrison, and L.R. Harrison. 1994. Avian medicine: principles and application. Wingers Publishing, Inc., Florida. Robb, L.A., K. Martin, and S.J. Hannon. 1992. Spring body condition, fecundity and survival in female willow ptarmigan. Journal of Animal Ecology 61:215-223 . •• .. SAS Institute Inc., 1999. Version 8.2. SAS Institute, Carey, North Carolina, USA . 155 Schroeder, M.A., J.R. Young, and C.E. Braun. 1999. Sage grouse (Centrocercus urophasianus) in A. Poole and F. Gills (eds). The Birds ofNorth America, No. 425. The Birds of North America, Inc., Philadelphia, PA . ... Snedecor, G.W., and W.G. Cochrane. 1980. Statistical methods. ih Edition. Iowa State University Press, Ames Iowa. 507 pp. Steele, J.H., and M.M. Galton. 1971. Salmonellosis. Pages 51-58 in Infectious and parasitic diseases of wild birds, J.W. Davie, R.C. Anderson, L. Karstad, and D.D. Trainer. (eds). Iowa State University Press, Ame·s, Iowa. Sturkie, P.D. 1976. Blood: physical characteristics, formed elements, hemoglobin, and coagulation. 53-75 in P.D. Sturkie, ed. Avian Physiology. 3rd edition. Springer-Verlag, New York, N.Y. Wallestad, R.O. 1975. Male sage grouse responses to sagebrush treatment. Journal of Wildlife Management 39:482-484. White, G.C., and K.P. Burnham. 1999. Program MARK: Survival estimation from populations of marked animals. Bird Study 46 Supplement, 120-138. White, L.M., R.E. Mirarchi, and M.E. Lisano. 1987. Reproductive capability of wild hatching-year mourning doves in Alabama. Journal of Wildlife Management 51:204-211. Wiebe, K.L., and K. Martin. 1997. Effects of predation, body condition and temperature on incubation rhythms of white-tailed ptarmigan Lagopus leucurus. Wildlife Biology 3:219227. Wittenberger, J.P. 1978. The evolution of mating systems in grouse. Condor 80:126-137. . .. .. Wobeser, G. 1997. Diseases of wild waterfowl, 2nd ed. Plenum Press, New York, N.Y . 324pp. Yamamoto, R. 1991. Mycoplasma meleagridis infection. Pages 212-223 in Diseases of Poultry, 9th ed., B.W. Calnek, H.J. Barnes, C. W. Beard, W. M. Reid, and H. W. Yoder, Jr. (eds. ). Iowa State University Press, Ames, Iowa. Yoder, H.S. 1991. Mycoplasma gallesepticum infection. Pages 198-212 in Diseases of Poultry, 9th ed., B.W. Calnek, H.J. Barnes, C.W. Beard. W. M. Reid, and H. W. Yoder, Jr. (eds.). Iowa State University Press, Ames, Iowa. ... 156 Table 5.1. Coefficient of determination (r) for impact of spring body condition on reproductive parameters of greater sage-grouse females in Axial Basin and Danforth Hills, Colorado, 2001/2002. • • .. • Parameter n F p ; Clutch initiation date 43 0.68 0.415 0.016 Clutch size 52 0.01 0.938 0.001 Hatch size 32 0.02 0.883 0.001 ._ ... ~ 157 Table 5.2. Blood biochemistry values for adult and yearling greater sage-grouse females captured in March and April, Axial Basin and Danforth Hills, Colorado, 2001/2002 and for Oregon and Nevada captured 1999-2001. ~ " Axial Basin and Danforth Hills Blood Parameter (units) • ~ !-+ y i -r.~ ~ ·~ ..... • Oregon and Nevada (Dunbar 2002, personal communication) Combined years and ages 2001 X 2002 SD X SD X SD AST (U/L) a 353.8 (n=50) 39.7 334.2 (n= 38) 49.9 430 (n=144) 77 CKb (U/L) 2563.4 (n=50) 2440.9 1992.0 (n=38) 1533.4 2396 (n=143) 1017 Glucose (mg/dl) 352.1 (n=50) 49.3 339.7 (n=37) 48.9 320 (n=144) 35 Total protein (mg/dl) 4.2 (n=47) 0.6 4.9 (n=38) 0.5 5.3 (n=139) 1.2 Phosphorous (mg/dl) 6.4 (n=46) 1.8 7.2 (n=38) 2.5 7.3 (n=144) 2.4 Uric acid (mg/dl) 10.9 . (n=49) 3.5 10.9 (n=38) 6.0 5.4 (n=143) 2.2 25.6 (n=38) 6.9 21.4 (n=141) 7.4 Not tested Calcium (mg/dl) APe (U/L) 483.9 (n=52) 224.7 609.3 (n=38) 245.6 Not tested Cholesterol (mg/dl) 135.3 (n=49) 34.3 174.5 (n=38) 40.8 Not tested Creatinine (mg/dl) 0.4 (n=44) 0.1 0.8 (n=38) 0.2 Not tested Sodium (mEq/1) 155.0 (n=42) 2.7 154.9 (n=31) 3.9 Not tested Potassium (mEq/1) 4.6 (n=44) 1.6 4.6 (n=35) 1.2 Not tested 119.0 6.5 Chloride (n=42) (mEq/1) a AST =aspartate amino transferase b CK = creatinine phosphokinase cAP = alkaline phosphatase 119.1 (n=31) 6.5 Not tested 158 Table 5.3. Biochemistry values for 2 capture methods (mean and standard error) in pre-laying greater sage-grouse females in northwest Colorado, 2001/2002. t .. CODA Net Launcher Biochemistry panel CKa Chloride Mean SE Mean SE F-value P-value 3642.6 (n = 22) 780.8 1874.6 (n = 66) 110.9 11.52 0.001 0.9 19.18 < 0.001 124.1 (n = 19) a creatinine phosphokinase .,... ... .,.. Spotlight trapping 0.6 117.3 (n =54) .. ~ r~ 159 Table 5.4. The influence of M synoviae on reproductive parameters (mean and standard error) of greater sage-grouse females in the Axial Basin and Danforth Hills, Colorado, 2001/2002. ~ Nest success Mean clutch initiation date SE Mean clutch SIZe SE 75% (n = 16) 69% (n = 13) - 24 April (n = 26) 1.4 6.5 (n = 10) 0.8 NegativeM synoviae 85% (n = 40) 57% (n = 35) 20 April (n =9) 1.7 6.9 (n = 22) 0.3 Test statistic x 2 = 0.78 1=0.20 t = 1.39 t = 0.52 0.377 0.658 0.175 0.607 t Nest initiation rate PositiveM synoviae rY • ~ ~ ~ P-value I ....... r l 160 Table 5.5. Logistic regression model of body condition and biochemical variables predicting the prevalence of M synoviae in pre-breeding greater sage-grouse females in northwest Colorado, 200112002. Dependant Variable Presence of M synoviae Parameters Regression Coefficient .(SE) F-value. Wald's Chi-square P-value Cholesterol -0.02 0.01 3.12 0.077 Sodium -0.20 0.13 2.47 0.116 Year 0.91 0.49 3.40 0.065 162 Table 5. 7. Breeding and brood-rearing survival models of greater sage-grouse females tested forM synoviae in northwest Colorado, 2001/2002. Models with 11 QAICc .:S 2 are considered competing models. QAICc 11 QAICc Parameters Deviance {S (.)} 52.97 0.00 1 50.93 {S (prevalence)} 55.04 2.07 2 50.91 Model f l