Disturbance and succession on the Krakatau Islands, Indonesia
This thesis set out to investigate the influence of disturbance on the succession of the Krakatau islands (Rakata, Sertung, Panjang). The hierarchical model of succession by S. Pickett and colleagues (1987) was adopted as a research framework, and provided the basis for an alternative model of succession on Krakatau that focuses on processes rather than successional pathways. Investigations were conducted on (i) the meso-scale, and (ii) the patch-scale, (i) quantified the recent disturbance regime, and inter- and intra-island differences in diversity, (ii) compared sapling performance (growth, mortality and recruitment), and species compositional patterning in space and time for saplings and the seed bank with respect to island, gap size and severity of disturbance. Multivariate techniques were used, and amongst other attempts at characterising the light environment, hemispherical photography was employed. For the first time the effect of a continuous period of volcanic activity (1992-1995) of Anak Krakatau could be directly quantified and compared between Panjang and Sertung (ash-affected) and Rakata (receiving no ash). Increased rates of gap formation in the volcanically active period in comparison to the previous decade were found for all islands. This supports the disturbance-driven model of Whittaker and colleagues. However, an extension is required, because, contrary to expectation, Rakata also experienced more disturbance. This increase is argued to be a result of more severe weather conditions, and an increased number of earth tremors, during times of volcanic activity. The disturbance factors of extreme climatic events (e.g. ENSO events) and human impact are also proposed for inclusion in the alternative model. Drought associated with the 1994 El Niño is of relevance to short-term and potential long-term impact on regeneration dynamics and succession. Attention was drawn to the local human influence of pumice mining on the coastal forests. Supporting previous findings on the plot- and whole island scales, data from species presence/absence transects established that species richness and beta-diversity on the ash- affected islands was also lower on the meso-scale. Panjang's canopy composition is less uniform, and locally more species-rich than Sertung's. More evidence of the suggested decline of the mono-dominant species Neonauclea calycina and Timonius compressicaulis was gathered. The third dominant, Dysoxylum gaudichaudianum, is expanding in the lowlands of all islands. This is aided by its ability to regenerate in moderate shade, to grow rapidly in gap environments, and its tolerance of ash-fall, drought and herbivory. However, on Rakata, it is not expected to become generally mono-dominant because a considerable number of other potential canopy species are present. Sapling performance and species composition and its changes were in general strongly affected by ash-fall and drought. These factors tended to override effects of gap size and severity of disturbance. Advance regeneration, and the composition of the local forest type were identified as important factors influencing the composition of the early stages of gap-fill. The local forest type also seemed to contribute most to seed bank composition. As rarer species tended to have clumped distributions, and 'safe sites' for regeneration seemed not to be limiting, dispersal constraints were argued to be the most likely factors slowing diversification, unless further severe volcanic disturbance leads to successional set-back. The latter also strongly limits the predictability of succession on Krakatau.