Epsilon Open Archive

Abstract

In West Africa policies for prescribed early fire, grazing and selective tree cutting in the savanna-woodlands are rarely based on long-term experimental studies. The purpose of this study was to provide scientific evidence based on field data from two case studies for an informed discussion on the long-term response of herbaceous abundance both at the community and individual species levels to fire, grazing, selective cutting and their interactions. A long-term factorial experiment was established in two State forests reserve in Burkina Faso, and mainly differing in their soil attributes. Community abundance data recorded from line intercept sampling over 13 years, were analyzed using a multivariate ordination technique known as Principal Response Curves (PRC). The results indicate that disturbance regimes, independently or interactively, influenced species abundance over time with inter-site specificity. The dynamics of these disturbance regimes exhibited temporal variation which could be related, to some extent, to inter-annual variation in annual rainfall. The PRC ordination accounted for 38% and 34% of the variation within the data set for sites with deep and shallow soils, respectively. At the site with deep soils, more than one PRC axis was needed to summarize the community response sufficiently, suggesting that the species reacted in different ways to disturbances. The PRC method approach to the analysis of disturbance dynamics allowed us to distil the complexity of the community responses to those of individual species and to identify species that can serve as indicators of certain disturbance regimes.