|the support of Richards Bay Minerals, and we now maintain an extensive database on the chronosequential development of tree, understory vegetation, millipede, bird and small mammal assemblages of known ecological history. This enables us to 1) evaluate rehabilitation success and 2) develop and test hypotheses related to self-induced community development across temporal and spatial gradients. Studies underway in this outdoor laboratory include the following:
The restoration goal of this disturbed coastal dune forest is to re-create a forest that is similar to relatively intact forests in the region. Successful dune forest rehabilitation should predictably be associated with structural and functional development of both the biotic and abiotic components of the ecosystem towards a benchmark, in this instance an area of undisturbed forest. CERU has been monitoring ecological development in assemblages of plants and animals, as well as in soil variables (nutrient and fertility levels) that reflect ecosystem function. The Ecological Monitoring Program evaluates the trajectory of development in ecological variables, comparing them both to trajectories obtained during previous years, and to trajectories across a chronosequence of naturally regenerating forest patches in the region. The program also evaluates the predicted endpoints of development. The program is, therefore, an important tool for mining companies (such as Richards Bay Minerals) that need to meet the objective of ecological sustainability.
Principles of succession predict that disturbances, such as those induced by mining, may bring about the development of biotic communities through primary succession. This is characterised by local colonisations and extinctions that take place in stages and result in the natural regional and local species pools being indistinguishable. The decay of differences in these species pools is expected to follow negative exponential trends.
Rehabilitation does not always follow a deterministic trend. The existence of multiple stable states, chaotic or non-equilibrium dynamics, continuous disturbances and the vagaries of climate, to name a few, all prevent predictable change in groups of forest species. We are assessing the empirical evidence in support of current restoration models, using data accumulated during some 15 years of regular surveying of plant, millipede, bird and rodent assemblages.
Dispersal is a key process in the restoration of any ecosystem. Within dune forests, frugivorous birds consume a wide range of fruit - some 59 percent of all dune forest trees are bird dispersed. Effective restoration management therefore depends on birds as agents of tree dispersal. Our research focuses on questions such as, which bird species disperse which tree species? Does the geographic distribution of a tree species, even on a small scale, depend on the movements of birds? To determine whether the distribution of tree species is related to the distribution of bird species, we record, both "ends" of tree dispersal, the consumption of fruit of some forest tree species by birds and the deposition of tree seed in their faeces.
Eminent ecological theories dictate that landscape pattern influences colonisation and extinction dynamics and, therefore, determines biological diversity and species composition within a habitat patch. This is not just of theoretical importance, as habitat fragmentation continues to be a major threat to biological diversity. Since 1977, Richards Bay Minerals has been mining for minerals on the dunes to the north of Richards Bay Town. The mining process results in a mosaic of remnant forest, bare sand, commercial forestry, and regenerating indigenous forest. Restoration ecology addresses spatial and temporal influences on how plants and animals reach a disturbed site and survive there. Therefore, an understanding of the constraints to community assembly in a fragmentation metaphor is imperative. We are investigating the influence of the spatially and temporally changing landscape pattern of the mining lease area on colonisation and extinction dynamics in forest birds and trees and the consequences thereof for habitat restoration.
Reproductive characters of plants, including seed type, seed size, reproductive strategy, age at first reproduction and phenology, play a role in determining species-specific dispersal distances, dispersal rates and dispersal frequencies. Together, these factors influence the likelihood of a species dispersing to a new habitat and are therefore important for the outcome of a restoration effort. In this study on coastal dune forest trees, we quantify the relationship between life history characteristics that are closely associated with dispersal and the pattern of colonization in new habitats.
Human-induced fragmentation of landscapes disrupts the spatial arrangement of habitats. This may reduce biodiversity and disturb ecological processes. Fragmentation of potential source habitat may also reduce species richness and inhibit colonisation of rehabilitating habitats, thereby disrupting restoration programmes that aim to recover disturbed areas. However, there is a lack of consensus on the consequences of fragmentation for species in the literature. Following island biogeography theory, we anticipate that a reduction in dune-forest fragment size and increased distance from continuous dune forest will reduce millipede species richness and abundance in fragments and patches of rehabilitating dune vegetation. We also expect that due to varying responses of species to fragmentation the millipede community composition in these fragments will differ from that in a continuous dune forest. We are currently investigating the distribution of millipede species across the RBM lease area, in remnant and intact coastal dune forest, as well as regenerating patches of various ages. This project will give us insight into the consequences of fragmentation for the rehabilitation and eventual restoration of the coastal dune forest millipede community.