Turnover of plant trait hierarchies in simulated community assembly in response to fertility and disturbance

Plant ecologists have placed increasing emphasis on a functional understanding of vegetation. One way to gain insight into the assemblage of vegetation communities is to investigate plant trait responses to environmental gradients or experimental treatments. We present simulations of responses of suites of traits to treatments differing in soil resources and disturbance intensity, in order to construct a functional response hierarchy of traits. We focus on the traits specific leaf area (SLA), plant height, seed mass and life cycle. Though only four traits are varied, these traits are connected to other traits either through trade-offs (e.g. SLA with leaf life span and relative growth rate, seed mass with seed number) or allometric rules (e.g. above-ground biomass scales positively with below-ground biomass). Thus a wide range of plant life history is represented in the simulations. We simulated the assemblage of plant types composed of these traits at two fertility levels and four disturbance treatments, i.e. every 7 years, annually, or monthly mown, and annually ploughed. We present the results of a simulation using LEGOMODEL, an individual-based, spatially explicit, ecological field model and develop a novel method to construct a functional response hierarchy of traits. A competitive ranking of plant types is constructed and subsequently translated into a functional response hierarchy of traits on the basis of the variability of trait states within the plant type ranking. The competitive superior trait states as well as the functional hierarchy change between different treatments. Except for deeply disturbed conditions, perennial plants are always superior. Tall canopy height is superior on fertile soil, as long as it is not mown monthly. High SLA is advantageous at fertile or ploughed sites. Life cycle and canopy height are always at the first two ranks of the functional hierarchy of traits. Our results concerning the advantageous trait states are in line with published field studies, except for the trait seed mass. Here LEGOMODEL predicts competitive advantage of small seeds under any treatment. The functional response hierarchy of traits presents a hypothesis to be refined by further field research as we know of no field study addressing this question. (C) 2006 Elsevier B.V. All rights reserved