The Groenvaly grassland fragmentation experiment: design and initiation

Abstract

An experiment designed to study the effects of habitat fragmentation on biological diversity in the montane grasslands of Southern Africa was launched in September 1994. The Groenvaly experiment is positioned on the western slopes of the Drakensberg escarpment in Mpumalanga, South Africa. The hypotheses being tested are: (1) that fragmentation reduces biological diversity; (2) that fragmentation alters community composition (invasion); (3) that the reduction in diversity is dependent on fragment size; (4) that species viability is differentially affected by distances between fragments; and (5) that afforestation causes irreversible community change. In addition, the experiment evaluates the relevance of mainland–island, metapopulation and source–sink models for describing the population characteristics of the grassland taxa. The experimental design comprises two fragment sizes (0.25 and 1.5 ha) replicated 12 times. Six replicates were retained as grassland fragments surrounded by newly established Pinus elliottii plantations. A further six replicates served as control fragments in an adjacent designated grassland conservation area. Sampling to monitor community composition (vegetation, Coleoptera, Orthoptera, arachnida, small mammals) and the population dynamics (mark–recapture) of selected taxa (small mammals to date) was conducted. Monitoring of the population dynamics of selected Coleoptera and Orthoptera taxa, identified as suitable models for further study, and the soil micro-fauna in fragments will commence in 1997. Two years following plantation establishment effective fragmentation has not yet been achieved. However, the analysis of pre-treatment data is used to evaluate the effectiveness of the experimental design. The analyses suggest that there is a sound basis for future interpretation of experimental results because no significant differences between the experimental and control plots were detected. The presence of some variation within experimental and control plots underscores the importance of adequate replication in ecological field experiments, a strength of the experimental design employed here.