Individual variation in growth, reproduction, and mortality is widespread, even in asexual organisms. Analysis of this variation reveals the existence of strong positive and negative correlations among life history traits. These correlations often run counter to the predictions of simple theory. For example, a strong positive correlation between growth and reproduction is often found, even when individuals are raised in low-food environments, whereas theory would predict a negative correlation. In this study, we investigate whether dynamic energy budget (DEB) theory can provide a mechanistic understanding of such life history correlations by fitting DEB models to the growth, reproduction, and mortality trajectories of individual Daphnia pulicaria. We fit several thousand life history trajectories from thirteen different D. pulicaria genotypes across gradients of food quality and quantity, where total ingestion was controlled.
Results/Conclusions
We show that DEB models are flexible enough to fit individual growth and reproduction trajectories. Previous studies have tended to fit trajectories averaged across many individual Daphnia. We show that such average fits give a misleading impression of energy allocation and utilization. We show that the variation in best-fit parameter values among individuals within a genotype is large. This variation is larger than the variation we find in mean parameters among the genotypes. This is a striking finding, given that D. pulicaria reproduces clonally, and suggests that genotypes are not nearly as distinct (from a life history perspective) as expected. Moreover, we show that the empirically-observed positive and negative correlations among life history traits (such as the strong positive correlation observed between growth and egg production) arise from strong correlations in the underlying parameters governing energy allocation and utilization. This is a more nuanced view than that taken by current theory, which tends to assume that correlation arises because of individual variation in single parameters. This has important implications for how we model life history evolution, and we conclude by proposing a hierarchical model of evolution.