Modelling the effects of fire and competition on the dynamics of neotropical savannas: from plants to community

Abstract

Savannas are one of the most extensive biomes of the world and are subject to intense pressures due to large human populations using this ecosystem, even though it is characterized by low crop yields and poor nutritional quality of natural pastures. Conversion to agriculture and fragmentation of Neotropical savannas, and increasing population pressures are promoting the extent and use of anthropogenic fire. In this paper we develop a mathematical model to account for spatial dynamics of the savanna grass ommunity. This model is based on a matrix model of growth of three dominant grasses of the Neotropical savanna (Elyonurus adustus, Leptocoryphium lanatum and Andropogon semiberbis). The model integrates plant architecture and competition to simulate the dynamics of this system. Our model focuses on fire as an essential event in the dynamics of the savanna community, especially frequency of occurrence. We study the current frequency of one fire a year and how changes in fire frequency tend to destabilize the system. The phenology of the species is a key point to understand the behaviour of these species on relation to fire. Late growth species (i.e. A. semiberbis) are more adapted to the current frequency than early ones, showing a higher competitive ability. Our results suggest that if there is a change of the actual fire frequency this system could undergo a decrease of phenological and plant form diversity.