Modeling survival of Columbian sharp-tailed grouse on tribal landscapes in north central Washington

Richard Patrick Whitney

Columbian Sharp-tailed grouse (Tympanuchus phasianellus columbianus) are an important cultural and subsistence species for regional Native American Tribes in Washington. However, the once historically large populations of this species have rapidly declined over the last 110 years and they are currently a state-threatened species. We studied survival of 139 radio-collared grouse from 8 breeding grounds (lek sites) on landscapes managed by the Colville Confederated Tribes (CCT), including 65 grouse introduced from 3 out-state source populations from 2002 - 2009. We measured a habitat suitability index (HSI) on four lek sites to help evaluate three working hypotheses that survival would be: 1) lower for females than males, 2) higher for resident grouse than for translocated grouse, and 3) higher for grouse residing on lek sites with higher habitat quality (indexed by HSI rank at lek sites). We compared survival models resulting from traditional survival analyses (e.g., Cox Proportional Hazard models; parametric models) with those derived from adaptations of survival trees (i.e., regression trees; boosted regression trees). Grouse iv survival rates varied by sex, age class, cohort (year captured), season (month captured), breeding site (lek), source of grouse (resident vs. translocated), and HSI rank of lek sites. Survival rates in a variety of exploratory models generally were lower for introduced vs. resident grouse, for females, and for grouse occupying lesser quality habitats as measured by HSI rank indices measured around lek sites. Overall, sharptailed grouse experienced the highest mortality rates from late spring to early summer seasons (breeding through brood rearing). Our survival models demonstrate that conclusions drawn from traditional survival models may be extended and refined using non-parametric, survival tree models that are better able to detect and quantify complex ecological interactions in unbalanced data sets collected across multiple years, seasons, and landscapes. We suggest that persistence of local populations of sharp-tailed grouse on landscapes managed by the CCT likely could be improved through careful management of grazing practices and landscape-level restoration of critical native grassland habitats adjacent to existing lek sites in an effort to increase regional grouse survival rates and support viable metapopulations.

Modeling survival of Columbian sharp-tailed grouse on tribal landscapes in north central Washington

Files and links (1)

Abstract

Metrics

Details