Modelling competitive ability of neotropical savanna grasses: Simulation of shading and drought impacts on biomass production

Abstract

In this paper, we develop a technique to model the spatial distribution of shoots along vertical and horizontal dimensions of a plant community. We use it to simulate the growth of a tropical savanna near the city of Barinas, Venezuela, to explore the responses of the peak biomass of a plant community to a range of 10-50% reduction of rainfall. We selected three dominant grass species: Elyonurus adustus, Leptocoryphium lanatum, and Andropogon semiberbis in a 4×7 m study plot. We estimate parameters values from data measured in the field. The number of shoots for each plant is obtained according to soil water availability and distributed vertically by 10 cm levels using a transition matrix. Convolution allows calculation of leaf area index for each cell and vertical level, which is then used to calculate light attenuation and thus the proportion of shaded shoots in each cell and level. With this information, maximum evapotranspiration is determined to calculate soil moisture using daily rain time series. Biomass is calculated for all species based on shoot biomass measured in the field and fire is simulated by removing a fraction of the shoots segments of all species.Modeled biomass fits reasonably well to field data.The model predicts significant differences in the response of each species to drought. A. semiberbis was the most drought resistant of the three species, whereas L. lanatum was the least, and E. adustus was intermediate, in agreement with observations in the literature. Our model suggests that drought resistance increases with the biomass/transpiration ratio for the species considered.