Search for a command to run...
The development of the vegetation and floras of the Krakatau Islands in the Sunda Straits, Indonesia, since their "sterilization" in 1883 is described. Key features of the post—1883 environment, such as pedogenesis, geomorphology (coastal change), human influence, and recent volcanic activity are detailed, and their possible influence on spatial and temporal patterns in both vegetation and flora is discussed. Field work conducted in 1979, 1983, and 1984 has enabled an assessment of the present state of vegetation development, aided by plot—based sampling and analysis of the arboreal component, employing numerical classification (by TWINSPAN) and ordination (by DCA). The coastal communities were established early and have remained little changed, although the distribution of the various components has changed according to the influence of a dynamic coastal geomorphology. Fifty years after forest closure, the forest of the interiors remain species—poor and composed of typical early—seral species. Rakata was dominated inland by Neonauclea calycina and Ficus pubinervis up to ≈550 m altitude, above which a mossy forest of Ficus spp. and Schefflera polybotrya scrub was recorded. The main inland forest types on Rakata Kecil and Sertung were of young, evenaged stands of Timonius compressicaulis (extensive) and older stands of Dysoxylum gaudichaudianum (often with a T. compressicaulis understory). The principal axes of variation within the data were found to be between extremes of the Rakata forest types, with Sertung and R. Kecil stands remaining undifferentiated until lower levels of the analyses. These patterns were ascribed to a mix of environmental variation (coastal and altitudinal factors), chance variations in colonization, and to volcanic action following the emergence of Anak Krakatau in 1927. Disturbance by volcanism (e.g., in 1930, 1934—1936, 1939, 1952—1953, 1961) has resulted in the deflection of vegetation succession on R. Kecil and Sertung into a different pathway from that followed on Rakata, which has remained unaffected by the activity. On Anak Krakatau, volcanic action has prevented successful colonization away from accreting coastal deposits, and has several times eliminated the entire flora. In addition, the major source of propagules for Anak Krakatau is from within the group, and for these several reasons the new island is shown to be a poor analogue for the early recovery phases of the other islands. Complete floral lists are given for each recorded survey for each island in the group, including data from surveys in 1979, 1982, and 1983. These data have been revised from previous publications on the basis of recent herbarium work and literature searches. The data are analyzed according to several different models: cumulative species totals, species totals for particular combinations of surveys, and totals calculated on the assumption of minimum turnover. The early beach spermatophyte assemblages of Rakata have undergone relatively few losses in comparison to the assemblages of the interior, while within the latter there has been a relatively high proportion of losses among the pioneering pteridophytes. The possession of major habitat types has been identified as critical in determining the shape of the overall colonization curve, through the passive sampling of different source pools. The diversity of the floras of the group as a whole and of Rakata and Anak Krakatau has continued to increase. The curves of species present on Sertung and R. Kecil have levelled and fallen respectively, as a result of the volcanic activity of Anak Krakatau. It is shown that the number of animal—transported species on the Krakatau group has continued to increase over the last 50 yr, and that this accounts for the majority of the increase in the size of the spermatophyte flora. The animal—transported species and the early sea—dispersed species appear to be species—stable groups. Later sea—dispersed species included ephemerals and species of temporary habitats, and have experienced a relatively high proportion of losses. Few beach species that have established on all of the three main islands have subsequently become extinct from the group. The number of pteridophytes on the islands has increased over the last 50 yr, mainly through the addition of forest species. A large proportion of plant species has been found only on Rakata, which samples an upland source pool not represented on the other islands. It is argued that the assumptions of "classical" island biogeography are inappropriate to these data and that the pattern in floral recolonization can best be understood as a successional process involving broad habitat and dispersal mechanism determinants. The implication of these findings is that community dynamics are highly significant in determining rates of immigration, colonization (i.e., successful immigration), and extinction, and that the probabilities of each vary among different groups of species and through time.