SATCHMO: A spatial simulation model of growth, competition, and mortality in cycling savanna patches

Abstract

Many mechanisms have been suggested to explain the coexistence of woody species and grasses in savannas, yet, evidence from field studies and simulation models has been mixed. Shrub encroachment is an ecological and economic problem in savannas worldwide which generally is attributed to overgrazing. Patch-dynamics is a new mechanism explaining tree-grass coexistence and the natural occurrence of shrub encroachment in savannas. A patch-dynamic savanna consists of patches in which cyclical succession between grassy and woody dominance proceeds spatially asynchronously. The spatially explicit, individual-based patch-scale simulation model SATCHMO was developed to investigate cyclical succession in the paradigm of patch-dynamics for arid and semi-arid savannas. SATCHMO is designed to capture within-patch shrub population dynamics based on a grid of 51 m side length and a resolution of 10 cm. The model shrub characteristics were derived from Acacia mellifera, the main encroaching species in African savannas. The aim of SATCHMO is to give a detailed small-scale understanding of above- and belowground growth, competition, and mortality of savanna woody plants and the influence of precipitation and fire on patch transition frequencies, shrub growth rates, and shrub size frequencies. With SATCHMO, we want to identify the conditions leading to cyclical successions in general and shrub encroachment in particular. Soil moisture is the most important parameter in SATCHMO influencing growth, reproduction, and mortality of shrubs and grass tufts, and that mediates competition. To acknowledge the importance of belowground interactions in savannas, shrub root growth and competition are modelled spatially explicitly. The model output was successfully validated with morphometrical and spatial data from the field site in the South African Kalahari thornveld and with recent literature data on savanna woody species cover. Global sensitivity analysis with Latin hypercube sampling shows that soil moisture is the most important driver of shrub cover dynamics in semi-arid savannas.