Despite contrasting views on the impacts of loss of live coral cover and habitat complexity on reef fishes, experimental studies designed to assess the relative impacts of these two factors have not been conducted (Pratchett et al., 2008). Lirman (1999) used a mensurative experiment to analyze the impacts of coral mortality and structural degradation, via comparisons of reef fish assemblages on live and dead A. palmata colonies. His results suggest that loss of structure provided by A. palmata plays a more important role than loss of living tissue in determining fish assemblage composition, although the distribution of a number of specialists was related to the presence of live coral tissue. However, mensurative experiments are prone to confounding factors and can only correlate measurements. The lack of manipulative experiments has inhibited researchers from properly assessing relative importance of live coral cover and structural complexity. Thus, the objective of this study was to determine how coral biodiversity loss affects coral reef ecosystem processes, using a series of manipulative and mensurative experiments designed to answer three specific questions: 1) How does coral biodiversity loss affect habitat availability? Since the densities of multiple coral species have simultaneously declined (Green et al., 2008), and the impacts of such cumulative losses on habitat complexity are uncertain, the first experiment quantitatively assesses the impacts of coral species richness and identity on reef rugosity using both mensurative and manipulative methods. I also investigated the impacts of coral biodiversity loss on several fish community metrics, including species richness and overall abundance, in order to assess the appropriate measure of habitat quality for coral reef fishes. 2) How does loss of live coral and structure influence fish assemblages? I used mensurative survey techniques and experimental manipulations to determine the individual effects of both live coral loss and three-dimensional structure on reef fish assemblages. Experimental plots were created using artificial A. palmata mimics to reconstruct exact reef shapes, as opposed to cinder blocks, which are proven to be ecologically irrelevant approximations of rugosity (Kroutil, 2005). 3) What are the responses, and the causal mechanisms, of trophic interactions to habitat loss? To determine how ecosystem processes, i.e. grazing and predation, are affected by habitat loss, and the spatial extent to which trophic interactions have been affected, processes were measured in Complex' A. palmata stands, simple sites adjacent to (Near') and removed from (Far') structurally complex sites , and within experimentally recreatedCurrent' habitats using established tethering techniques (Valentine et al., 2007, 2008; Eklof et al., 2009). The `Current' sites were constructed to mimic the current state of isolated, non-living A. palmata skeletons that currently characterize Caribbean reefs.
