Empirical tests of life-history evolution theory using phylogenetic analysis of plant demography

Summary 1. A primary goal of evolutionary ecology is to understand factors selecting for the diversity of life histories. Life-history components, such as time-to-reproduction, adult survivorship and fecundity, might differ among species because of variation in direct and indirect benefits of these life histories in different environments or might have lower-than-expected variability because of phylogenetic constraints. Here, we present a phylogenetic examination of demography and life histories using a data base of 204 terrestrial plant species. 2. Overall, statistical models without phylogeny were preferred to models with phylogeny for vital rates and elasticities, suggesting that they lacked phylogenetic signal and are evolutionarily labile. However, the effect of phylogeny was significant in models including sensitivities, suggesting that sensitivities exhibit greater phylogenetic signal than vital rates or elasticities. 3. Species with a greater age at first reproduction had lower fecundity, consistent with a cost of delayed reproduction, but only in some habitats (e.g. grassland). We found no evidence for an indirect benefit of delayed reproduction via a decrease in variation in fecundity with age to first reproduction. 4. The greater sensitivity and lower variation in survival than in fecundity was consistent with buffering of more important vital rates, as others have also found. This suggests that studies of life-history evolution should include survival, rather than only fecundity, for the majority of species. 5. Synthesis. Demographic matrix models can provide informative tests of life-history theory because of their shared construction and outputs and their widespread use among plant ecologists. Our comparative analysis suggested that there is a cost of delayed reproduction and that more important vital rates exhibit lower variability. The absolute importance of vital rates to population growth rates (sensitivities) exhibited phylogenetic signal, suggesting that a thorough understanding of life-history evolution might require an understanding of the importance of vital rates, not just their means, and the role of phylogenetic history.