Abstract
Although classic ecological models often have assumed functional equivalence among individuals in a population, ecologists now recognize that individual variation can modulate ecological processes across levels of organization. Nevertheless, current models disproportionately emphasize variation among cohorts, and considerable uncertainty remains over the mechanisms that generate within-cohort variation and the downstream consequences for population and community dynamics. State-dependent life-history theory provides a useful framework for predicting the causes and consequences of within-cohort variation. Behavioral or physiological adjustments made by individuals in response to their underlying state (e.g., energy reserves or disease status) can influence fitness and, by extension, population performance. For iteroparous animals, resource allocation by maternal females often is state dependent; however, the population-level consequences of this strategy and the mechanisms that govern them remain largely unresolved. To explore individual variation in reproductive effort and its associated outcomes, we developed and empirically parameterized a state-dependent, individual-based model of maternal resource allocation for a long-lived, iteroparous mammal, the North American elk (Cervus canadensis). Females were allowed to adjust their investment in gestation versus lactation in response to their nutritional condition in spring. We tested the prediction that females in poor condition could increase fitness by delaying parturition and increasing investment in gestation, giving birth to correspondingly larger neonates that had a greater chance of surviving their first month of life (when mortality is generally highest) and subsequently reducing investment in lactation to help rebuild somatic reserves. We predicted that population growth would be faster when resource allocation was state dependent than when gestation length was decoupled from female condition and adjustment of reproductive investment was largely post-natal. Our results supported this prediction: state-dependent resource allocation by maternal females increased population growth by an average of 4%, leading to larger population sizes after 30 years. Population growth was consistent across a range of winter severities, suggesting that state-dependent resource allocation also could help buffer populations against climatic variation. Our results reveal a potentially general mechanism underpinning intraspecific variation in life-history strategies and suggest that such variation at the individual level can influence performance outcomes at the population level.