An integrative trait-based framework to infer resource budgets and life-histories of long-lived plants

Abstract

A fundamental assumption of functional ecology is that functional traits determine life-histories. Yet correlations between traits and life-history components are often weak, especially for long lived plants. This is because trade-offs, constraints, dynamic resource budgets and the scaling from single organs to entire plants cause complex relationships between traits and life-history. To elucidate these relationships, we present an integrated Trait-Resource-Life-History (TRL) framework that infers how functional traits affect organ-level costs and benefits of different life history components, how these costs and benefits shape the dynamics of whole-plant resource acquisition and allocation, and how these dynamics translate into life history. We illustrate this framework by developing a TRL model for a functionally diverse group of woody plants (22 species of the genus Protea from the South African Greater Cape Floristic Region). Using hierarchical Bayesian latent state-space modelling, we statistically parameterise this model from data on year-to-year variation in growth, reproduction and maternal care (serotiny) for 600 individuals. The parameterised model reveals that higher resource acquisition translates into both larger absolute resource pools and greater proportional resource allocation to reproduction. Accordingly, specific leaf area, a key trait increasing resource acquisition, is associated with larger resource pools, an earlier age of maturity as well as increased vegetative and reproductive performance at young to intermediate ages. In contrast, seed nitrogen content has opposing effects on the benefits of different organs and thus only shows weak correlations with life-history components. Importantly, the TRL model identifies trait and resource-mediated trade-offs at the level of organs, whole-plant resource budgets and life-histories. It can thus quantify key components of life-history theory that are so far largely inaccessible for long-lived plants. This permits novel insights into ecological and evolutionary mechanisms shaping life-histories. Application of the proposed framework to a broad range of plant systems should be facilitated by the increasing availability of trait and demographic data, whole-plant phenotyping and high resolution remote sensing. The integration of the TRL framework with models of biotic interactions further holds promise for a resource-based understanding of community dynamics across trophic levels and a closer integration of functional ecology, evolutionary ecology, community ecology and ecosystem science.