Greater prairie‐chicken brood‐site selection and survival in the Nebraska sandhills

Abstract

ABSTRACTThe greater prairie‐chicken (Tympanuchus cupido pinnatus) is a grassland bird species of conservation concern, although the region of the Nebraska Sandhills has the largest and most stable population in North America. The Sandhills region is semiarid and dominated by privately owned rangeland; vegetation in the region is typically less dense than the tall grass ecosystem. Most management recommendations for the greater prairie‐chicken are based on work in tall grass prairies, so our goal was to estimate the responses of brood‐site selection and survival to the unique vegetation characteristics of the Sandhills region. We studied prairie‐chickens on private rangelands in Rock and Brown Counties from 2009 to 2011. We radio‐marked 139 females and monitored females with broods from May through July using pickup‐mounted and handheld telemetry systems. We used discrete choice models to assess micro‐scale habitat selection within used pastures. Specifically, we investigated the effect of local heterogeneity of vegetation structure and composition on site selection. The contiguous grasslands of the Sandhills contain variation in soils, slope, and proximity to water table, so we also assessed selection of these ecological sites (the equivalent of land‐use categories in fragmented landscapes) for brood‐rearing within the broad landscape (macro‐scale). Last, we assessed variation in brood survival in response to weather and vegetation structure and composition. Prairie‐chicken females with broods selected sands ecological sites (upland, rolling hills sites) that had thicker vegetation with higher visual obstruction reading (VOR; = 7.7 cm, SE = 0.4), more variable VOR ( = 20.1, ±2.1), and greater litter depth ( = 0.07 cm, ±0.01) than at coupled random locations (VOR: = 6.6 cm, ±0.1; variance VOR: = 15.6, ±0.8; litter depth: = 0.05 cm, ±0.01). The selection function for brood sites peaked for sites with approximately 10 cm VOR, which was significantly less than the range of 20–30 cm of visual obstruction suggested for management of tallgrass prairies. In fact, only 5.0% of random sites and 5.4% of sites used by hens with broods had VOR >20 cm. However, VOR and litter depth did not explain variability in brood survival, and we found considerable uncertainty in our analysis of daily survival. Models with forb cover were ranked highest in the survival analyses (combined ωAIC c = 0.63), which provided some support for a positive effect of forb cover on brood survival (βforb cover = 0.327, SE = 0.337). We suggest that land managers in the Sandhills region provide brooding habitat in grazed, upland ecological sites that is characterized by minimal bare ground, a consistent distribution of litter, a forb component, and VOR levels of 4–11 cm in pastures in the vicinity of leks. © 2015 The Wildlife Society.