543Ð550 McClanahan

Corals were bleached throughout Kenya and most of the western Indian Ocean in the warm season of 1998 and coral mortality occurred over an extended period from March to July 1998 (1–3). Nearly all studied Kenyan coral reefs had coral cover of around 10% after the coral bleaching (1, 3). The greatest levels of mortality were experienced in marine protected areas that originally had high coral cover of around 45%, where mortality was estimated at 70% and the density of genera per site decreased by 30%. Unprotected reefs originally had coral cover of around 20%, experienced 45% coral mortality and a 44% loss of coral genera per site (1). There was also an immediate increase in the abundance of fleshy algae in about half the reefs following the loss of hard coral (1). An initial analysis of the fishery one year after the bleaching indicated that there was little detectable effect on the fishery, both in terms of the catch biomass and catch composition nor was there evidence that the socioeconomic position of the fisher communities was effected (4). Coral reef scientists commonly evoke competing hypotheses of the effects of coral loss on the benthos and fisheries. One is that the loss of coral cover will increase space for fast-growing turf algae and that the associated increase in organic production will result in elevated abundance of herbivores and perhaps other trophic groups influenced by this increased production (5, 6). The second hypothesis is that the loss of coral will lead to a loss of reef complexity, refuge for coral reef fishes and a consequent loss of fish production (7, 8). A third alternative is that the loss of coral will open up space for colonization by fleshy algae (1, 9) and result in lower benthic production, less palatable algae and this will reduce the abundance and diversity of fish (10, 11). It is likely that all of these processes occur simultaneously with a loss of coral cover, but empirical field studies are required to test whether these changes are, in fact, the outcome of coral loss and which process dominates. We examine aspects of these hypotheses by studying the changes in the reef benthos and fisheries 3 years before and after the 1998 coral mortality event. We also examine the effects of these changes on the economics of the artisanal fisheries.