|dc.description.abstract||I investigated desert mule deer (Odocoileus hemionus crooki) survival and habitat use at Elephant Mountain Wildlife Management Area (EMWMA) and surrounding properties in southwest Texas during 1990-1993. I estimated survival for 121 adult (> 33 month-old), 61 sub-adult (21-33 month-old), and 77 young (7-20 month-old) deer. Survival differed among years, age/gender classes, and seasons (P < 0.01). Subadults had highest survival, followed by adult females, adult males, and young deer. Hunting and natural stressors caused most adult buck deaths, while stressors and predation claimed most adult does and young. Subadult deaths were too few to identify mortality causes.
I applied a population model to the EMWMA herd; evaluating model performance, and simulating various management scenarios. Simulated trends agreed with observed data, while simulated survival estimates did not agree closely with observed rates. The model underestimated yearling and adult male survival, but overestimated fawn and adult female survival. However, most simulated survival estimates were included in 90% confidence intervals for observed rates. Harvest of females and longer seasons depressed populations. Severe drought overwhelmed harvest effects. Populations in which females were hunted declined more quickly, and to lower levels, during drought, than populations in which no female harvest occurred. Based upon survival and modeling results, doe harvest was not recommended.
I partitioned the area into 11 habitat layers and observed habitat use of adult deer. I observed no annual effects on habitat selection (P > 0.10). Males selected home ranges with low elevations and slope, south and west aspects, that were close to water, had high productivity but relatively few preferred forages, and were typified by woody and topological diversity. Females showed generalized selection among lower elevation and slope classes, choosing home ranges with slightly more preferred forages and lower total productivity than males chose. Males used sites that were further from water sources than sites selected by females. Habitat selection models showed potentially used areas that were difficult to identify using any single habitat variable.||