Analytical expressions of variability in ecosystem structure and function obtained from three-dimensional stochastic vegetation modelling

Source abstract

Whole ecosystem exchange of water, carbon and energy is predominately determined by complex leaf-level processes occurring at individual plants. Interaction between individuals results in a distribution of environmental conditions that drive a variety of nonlinear response functions such as transpiration and photosynthesis. The nonlinearity of biophysical processes requires higher-order statistical descriptions of micro-environment distributions in order to accurately determine the landscape-scale mean functional response. We present a mathematical framework for describing vegetation structure based on the density, dispersion, size distribution and allometry of individuals within a landscape. Using three-dimensional stochastic vegetation modelling, we develop analytic expressions of the second-order statistics of vegetation canopies, namely the mean and variance of leaf area density and leaf area index with height. These expressions also allow for the approximation of the distribution of beam penetration and sunfleck statistics through the canopy as a function of height. Finally, we demonstrate how landscape-scale fluxes are strongly affected by the variability in canopy micro-environments, and how stochastic vegetation modelling improves flux estimates relative to traditional homogeneous canopy models.