- Greater Sage-grouse (Centrocercus urophasianus) habitat research has historically focused on fine-scale (0.007 - 0.032 ha) vegetation structure and composition immediately surrounding sites selected by birds. However, little work has evaluated vegetation attributes important for Greater Sage-grouse at a landscape-scale or identified landscape attributes that influence habitat use patterns. Habitat use patterns by Greater Sage-grouse are complex and can occur across relatively large heterogeneous landscapes. This creates a major challenge for managers to interpret and predict habitat use patterns as well as to evaluate habitat suitability and prioritize habitats that are in need of ecological restoration. The goals of this research were to evaluate plot-level habitat characteristics found to be important in sustaining Greater Sage-grouse populations at a landscape-level and to identify landscape-level attributes associated with bird occurrence. Specific questions this research addressed were: 1) what is the variation in vegetation composition and structure at the plot versus landscape-level, 2) how does topography influence the distribution of vegetation composition and structure, and 3) what attributes at the landscape-level are most closely associated with Greater Sage-grouse habitat use? To address these questions we selected a 31,416 ha area in central Oregon surrounding a Greater Sage-grouse lek with a population that has been relatively stable since 1987. In February 2006, 50
Greater Sage-grouse were trapped, radio collared, and then tracked for two consecutive years. Four-hundred eighty bird UTM (Universal Transverse Mercator) coordinate location points were recorded for the entire population of birds during the duration of this study. Each collared Greater Sage-grouse was located on average every 15 ± 0.56 (mean ± SE) days, ranging from 1 to 154 days. Vegetation for the entire study area was mapped by cover types, which were defined by the dominant shrub species. When shrubs were not present in the plant community, cover types were separated by other surface characteristics such as bare ground, water, meadow, etc. A total of 23 cover types were delineated. Cover types were mapped using 0.5-m NAIP (National Agricultural Imagery Program) imagery. In addition to cover type, a set of biophysical predictor variables were created for the entire study area in a GIS (Geographic Information System) to evaluate the association with Greater Sage-grouse location points. These variables included elevation, slope, aspect, curvature, solar radiation, ruggedness index, northing, easting, and distance from roads, leks, and mesic habitats. A stratified random sample with cover types serving as the stratum was used to select random locations for sampling plot-level habitat variables. A total of 352 plots were sampled from 18 cover types across the study area with a minimum of 15 plots per cover type. Vegetation measurements collected were similar to those reported in the habitat guidelines developed by Connelly et al. (2000) and the Bureau of Land Management et al. (2000). Measurements included vegetation cover, height, and density of forbs recognized as important Greater Sage-grouse food species. Plot elevation, slope, aspect, curvature and landscape position were also recorded. Summary statistics were used to describe means and ranges within and between cover types. A combination of multiple linear regression and analysis of variance (ANOVA) were used to evaluate the effects of topographic attributes on the distribution of vegetation composition and structure. To address the third question, maximum entropy software was used to develop models that predict Greater Sage-grouse seasonal habitat use, generate maps from those models, and describe the shapes of the response curves as it relates Greater Sage-grouse habitat preference to individual landscape predictor variables.
Total shrub canopy cover across all cover types averaged 19.4%, ranging from 11.6 to 27.7%. Big (mountain and Wyoming) and low sagebrush canopy cover commonly varied between 2.6 and 16 fold within cover types. Deep-rooted perennial tussock grass cover averaged across all upland plots, was 26.7%, ranging from less than 1% to over 50%. Vegetation cover, Greater Sage-grouse food forb density, and sagebrush and grass height were significantly (P < 0.05) correlated with topographic attributes. Cover for the different plant life forms and food forb density increased with elevation. Cover for most of the herbaceous life forms was also greater on north than south aspects. Compared to Connelly et al. (2000) and the BLM et al. (2000) habitat guidelines, < 1% of the study area satisfied breeding and nesting guideline criteria, while < 31% satisfied the brood-rearing guideline criteria. Although most of the study area did not meet habitat recommendations presented in the guidelines, patches imbedded throughout the study area did and most areas satisfied many but not all of the guideline requirements. These results suggests that evaluating only mean values of community structure, both within and among cover types across the study area, limited the ability to fully identify patch variability and landscape heterogeneity as it relates to habitat suitability across large areas.
Maximum entropy results suggest Greater Sage-grouse habitat use during the breeding season increases near leks and within cover types of low sagebrush and low sagebrush/mountain big sagebrush complexes. Preferred summer habitat includes areas relatively high in elevation, distances that are close to leks, and within or a close proximity to habitats that harbor succulent vegetation through much of the summer. With Greater Sage-grouse utilizing resources within expansive landscapes, understanding the attributes that can be applied at a landscape-scale that attract disproportionate levels of habitat use can help managers predict where birds are likely to occur across the landscape. With the ability to discriminate between areas that Greater Sage-grouse are likely to use or avoid, managers can allocate limited resources to more effectively create, manipulate, and administer habitat conservation efforts where bird use is predicted and prioritize areas across the landscape in need of ecological restoration.