Modeling sage grouse: Progressive computational methods for linking a complex set of local, digital biodiversity and habitat data towards global conservation statements and decision-making systems

Abstract

Modern conservation management needs to link biological questions with computational approaches. As a global template, here we present such an approach from a local study on sage grouse breeding habitat, leks, in North Natrona County, Wyoming, using remote sensing imagery, digital datasets, spatial statistics, predictive modelling and a Geographic Information System (GIS). Four quantitative models that describe sage grouse breeding habitat selection were developed for multiple scales using logistic regression and multivariate adaptive regression splines (MARS-Salford Systems). Based on candidate models and AIC, important habitat predictor variables were elevation, distance to human development, slope, distance to roads, NDVI and distance to water, but not Sagebrush. Some predictors changed when using different scales and MARS. For the year 2011, a cumulative prediction index approach is presented on how the population viability of sage grouse can be assessed over time and space using Markov chain models for deriving future landscape scenarios and MARS for species predictions. © Springer-Verlag Berlin Heidelberg 2005.