Search

Tropical land-use intensification is rapidly increasing in regions that harbour high levels of biodiversity, thus posing a serious threat to the stability and resilience of tropical ecosystems and the important ecosystem services that they provide. We compared functional group richness and functional dispersion in litter-invertebrate communities among four different land-use systems, ranging in intensity from primary degraded lowland forest to oil-palm agriculture in two landscapes on Sumatra, Indonesia. We then investigated the consequences for functional stability and community resilience by calculating functional redundancy and response diversity of sampled communities. From primary degraded forest to intensively managed oil-palm systems, we found a 46% decrease in species richness and a 48% reduction in density, but weaker effects on functional group richness and an increase in functional dispersion. Although we detected no significant alteration of response diversity, functional redundancy of litter-invertebrate communities decreased clearly by losing 37% of functionally redundant species due to land-use change. Our results indicate that land-use change, from tropical rainforest to oil-palm agriculture, can alter both taxonomic and functional diversity of litter-invertebrate communities, resulting in the loss of functional redundancy and thus functional stability of these ecosystems. However, we also show that land-use systems of intermediate management intensity, such as jungle-rubber agroforestry, could serve as reservoirs of functional diversity and stability in monoculture-dominated production landscapes.

External threats from agricultural intensification, fire encroachment, species invasion and illegal harvesting present major conservation challenges in isolated tropical forest remnants. These processes can greatly exacerbate the magnitude of edge effects as the degree of patch to matrix contrast increases. Theory suggests that mitigation of these effects should be possible through conservation strategies that remove external threats and restore adjacent matrix structure, but this has not been tested experimentally. In the rapidly-dwindling Afromontane rainforests of Nigeria, where nature reserves have the least protection of all African conservation areas, we created an experimental matrix restoration treatment in which we excluded livestock by fencing, maintained a fire-exclusion break, and passively revegetated a 200 m buffer zone in the surrounding matrix at replicated edges. After three years, dung beetle communities in remnant forests showed a 53% increase in abundance at sites adjacent to the restored matrix. Over 90% of the common dung beetle species differed in the magnitude of their edge responses between forest-to-restored versus forest-to-degraded matrix sites. Moreover, a significant difference in species richness across the forest-to-degraded matrix edge became non-significant following matrix restoration, and there was also a significant decrease in community dissimilarity across the edge gradient in these regenerating sites. Just three years after excluding threatening processes from comparatively small areas of matrix habitat, we found that these efforts not only reduced edge effects, but also (1) enhanced dung beetle populations in the adjacent reserve, (2) led to an increase in dung beetle capture rates in the regenerating matrix, and (3) facilitated re-establishment of species that were absent due to matrix degradation. Therefore, regenerating buffers can substantially increase effective reserve size and restore invertebrate communities in landscape mosaics where remnant habitats are embedded within anthropogenic landscapes.