Sensitivity of primary production to precipitation variation is much higher in some ecosystems than others. Explanations for this variation have focused on how physiology and resource co-limitation influence plant growth. But plants do much more than grow. Plants evolve distinct life history strategies in order to cope with uncertainty and variability in their environment. Differences in life history strategy reflect differences in how organisms allocate energy to growth, survival, and fecundity. Variation in life history strategies could translate to variation in the sensitivity of plant growth to resource fluctuations. However, theory linking life history strategies to ecosystem-level dynamics is immature. We extended existing theory by using an annual plant population model to address the question; under what conditions do differences in life history strategies lead to differences in ecosystem response to fluctuating resources? We first varied environmental predictability (correlation between germination and fecundity) and resource variability (variation in fecundity) to find evolutionary stable strategies of germination fraction and germination variance. We then analyzed how variation in these strategies influenced the sensitivity of seed production to fluctuations in resource supply.
Results/Conclusions
Differing combinations of predictability and variability in the environment produced distinct life history strategies. Highly variable and predictable environments led to plasticity of seed production, whereas highly variable, unpredictable environments led to conservative bet-hedging. Unexpectedly, we found that the influence of life history on the sensitivity of seed production to resource fluctuations depended on the shape of density dependence. Rapidly saturating density-dependence masked the influence of life history strategy. When density dependence saturated slowly, plastic strategies translated to high sensitivity, while bet-hedging reduced sensitivity. Our results offer clear predictions about when and where life history should influence ecosystem sensitivity to resource fluctuations, and some of these predictions may be testable using existing data.