Concerns over declining mule deer (Odocoileus hemionus) populations during the 1990s prompted research efforts to identify and understand key limiting factors of deer. Similar to past deer declines, a top priority of state wildlife agencies was to evaluate the relative importance of habitat and predation. We therefore evaluated the effect of enhanced nutrition of deer during winter and spring on fecundity and survival rates using a life table response experiment involving free-ranging mule deer on the Uncompahgre Plateau in southwest Colorado, USA. The treatment represented an instantaneous increase in nutritional carrying capacity of a pinyon (Pinus edulis)−Utah juniper (Juniperus osteosperma) winter range and was intended to simulate optimum habitat quality. Prior studies on the Uncompahgre Plateau indicated predation and disease were the most common proximate causes of deer mortality. By manipulating nutrition and leaving natural predation unaltered, we determined whether habitat quality was ultimately a critical factor limiting the deer population. We measured annual survival and fecundity of adult females and survival of fawns, then estimated population rate of change as a function of enhanced nutrition. Pregnancy and fetal rates of adult females were high and did not vary in response to treatment. Fetal and neonatal survival rates increased in response to treatment, although the treatment effect on neonatal survival was marginal. Overwinter rates of fawn survival increased for treatment deer by 0.16−0.31 depending on year and fawn sex, and none of the 95% confidence intervals associated with the effects overlapped zero. Overwinter rates of fawn survival averaged 0.905 (SE = 0.026) for treatment deer and 0.684 (SE = 0.044) for control deer. Nutritional enhancement increased survival rates of fetuses to the yearling age class by 0.14−0.20 depending on year and fawn sex; 95% confidence intervals slightly overlapped zero. When averaging estimates across sexes and years, treatment caused fetal to yearling survival to increase by 0.177 (SE = 0.082, 95% CI: 0.016−0.337). Annual survival of adult females receiving treatment (Ŝ = 0.879, SE = 0.021) was higher than survival of control adult females (Ŝ = 0.833, SE = 0.025). Our estimate of the population rate of change (λ̂) was 1.165 (SE = 0.036) for treatment deer and 1.033 (SE = 0.038) for control deer. Increased production and survival of young (i.e., fetal, neonatal, and overwinter fawn survival) accounted for 64% of the overall increase in λ̂, whereas adult female survival accounted for 36% of the increase in λ̂. The effect of nutrition treatment on overwinter fawn survival alone accounted for 33% of the overall increase in λ̂.

We documented food limitation in the Uncompahgre deer population because survival of fawns and adult females increased considerably in response to enhanced nutrition. We found strong evidence that enhanced nutrition of deer reduced coyote (Canis latrans) and mountain lion (Puma concolor) predation rates of ≥6-month-old fawns and adult females. Our results demonstrate that observed coyote predation, by itself, is not useful for evaluating whether coyotes are negatively impacting a deer population. Our results also indicate that mountain lions may select for deer in poorer condition under some circumstances, suggesting that mountain lion predation may not always be an additive source of mortality. Disease mortality rates of adult females did not decline in response to enhanced nutrition. Winter-range habitat quality was a limiting factor of the Uncompahgre Plateau mule deer population. Therefore, we recommend evaluating habitat treatments for deer that are designed to set-back succession and increase productivity of late-sera