Title: Temporal changes in community structure of four coral reef areas off Mauritius

Information on temporal changes in the community structure of Mauritian coral reefs is lacking. In this study multivariate techniques are used to analyse changes in reef structures at four locations around the island with relevant bearings to the coastline (shore, back and fore reefs) and exposed to different levels of human activities: Trou aux Biches (TAB, tourism), Bambous Virieux (BV, agriculture), Bel Ombre (BELO, agriculture and tourism) and Ile aux Benitiers (IAB, control). No significant changes in community structure occurred at BELO (back reef), IAB (shore and fore reefs) and TAB (fore reef). Consistent changes, although not significant, were observed at the shore reefs of BELO and BV and back reefs of IAB and BV where the position of the reef structure in 2005, on MDS ordination, were furthest away from the initial reef structure. At BELO (shore reef), a significant shift from foliaceous non-Acropora to dead corals in the later years was noted. Branching Acropora was the most dominant lifeform category at the back reefs of BELO and BV. The main contributing lifeforms at the other reefs were: tabulate Acropora at IAB back reef and massive non-Acropora at TAB fore reef. Abiotic forms dominated the substrate of other coral reefs: dead coral at TAB back reef, rubbles at IAB shore reef and rock at IAB fore reef. Large scale analysis of percentage cover of lifeforms showed that the coral reefs off Mauritius could be assigned to four main types of coral reefs, with identical community structure: (i) the fore reefs at IAB and TAB (rock and massive non-Acropora), (ii) the shore reefs at BELO and BV (macroalgae and submassive non-Acropora), (iii) the back reefs at BELO and BV together with the shore reef at IAB (branching Acropora), (iv) the back reefs at IAB and TAB (dead coral and tabulate Acropora). The coral reefs around the island could be divided into two major areas, along a diagonal southwest-northeast axis, with proper reefs off the south and east coasts but with an abiotic dominant substrate off the north and west coasts. The patterns in the reef community structure could not be linked to any of the recorded environmental parameters (sea surface temperature, salinity, pH and dissolved oxygen). It is postulated that these parameters were not the appropriate indicators for monitoring the observed changes in the coral reefs.