A Precipitation-Habitat Model for Northern Bobwhites on Semiarid Rangeland

Abstract

Models that predict wildlife responses to natural and human perturbations are useful in research, extension, and management. Thus, we developed a priori and tested a precipitation-habitat model for predicting autumn density of northern bobwhites (Colinus virginianus) on semiarid rangeland. Variables in the model included the proportion of an area usable based on dispersion of woody cover, Thornwaite's index of precipitation effectiveness, proportion of sand-sized particles in the soil, proportional coverage of woody cover and forbs, and proportional exposure of bare ground. Observed densities were a linear function of modeled densities based on pooled data from the Rio Grande Plains (n = 16) and Gulf Coast Prairies, Texas (n = 6) (r2 = 0.34, 20 df, a = 0.0021). The model differed from species-habitat models by using an environmental correlate of bobwhite abundance (precipitation effectiveness), which allowed incorporation of quail population history into model output. The model provides an analytical framework for identifying habitat deficiencies on semiarid rangeland and for predicting bobwhite population response to remedial management.