Chemical warfare on coral reefs: Sponge metabolites differentially affect coral symbiosis in situ

Coral reef ecosystems are characterized by high species diversity and intense levels of biotic interaction, particularly competition for space among sessile benthic invertebrates. Using in situ pulse amplitude modulated fluorometry, we demonstrate that secondary metabolites present in the tissues of some Caribbean sponge species have rapid allelopathic effects on the symbiotic algae (zooxanthellae) that live in coral tissues and provide the energy for coral growth and reef formation. When incorporated into stable gels at natural concentrations and placed in contact with brain coral heads on shallow reefs for ,18 h, secondary metabolites of some sponge species caused a decrease in the photosynthetic potential of the symbiotic algae and bleaching of the coral tissue, whereas those of others interfered with algal photosynthesis without causing bleaching. Some sponges produce metabolites with different modes of action for competing with reef corals. The global decline in the health of coral reefs has been ascribed to a variety of possible reasons (Hughes 1994; Pandolfi et al. 2005). Among the hypotheses include those of ‘‘indirect effects’’ due to the loss of coral reef predators or herbivores. For example, the mass mortality of the urchin Diadema antillarum on Caribbean reefs has been blamed for greater growth of macroalgae, which in turn compete for space with reef-building coral species (Hughes 1994). More recently, it has been demonstrated that sponge-eating fishes control the abundance of some sponge species on reefs (Pawlik 1998), so that decreases in fish abundance may permit the growth of otherwise cryptic sponge species, some of which may more effectively compete for space with corals (Plucer-Rosario 1987; Hill 1998). One sponge species, Chondrilla nucula, was recently observed to overtake reefs in Belize (Aronson et al. 2002). Similar trophic cascades have been implicated in the restructuring of other marine ecosystems (Franks et al. 2005). Mechanisms for competition between coral species have been well described (shading, sweeper tentacles, gastric filaments; Lang 1973; Richardson et al. 1979), but those that permit the overgrowth of corals by sponges have not. Although it has long been hypothesized that chemical warfare is important in competitive interactions among benthic invertebrates on coral reefs (Bakus 1981), ecologically relevant tests of allelopathy are scarce. Past laboratory-based respirometry measurements of the effects of sponge metabolites on corals were neither properly controlled nor ecologically realistic (Sullivan et al. 1983). Because of the diluting effects of turbulent flow, allelopathic interactions on coral reefs are most likely to occur on contact or at very short distances (Kubanek et al. 2002). Although the negative effects of sponges in close proximity to corals have been clear (e.g., Plucer-Rosario 1987; Hill 1998), experimental demonstration of allelopathy, as opposed to a physical effect of smothering, has been lacking. We began developing a field assay for assessing sponge allelopathy against corals over a decade ago, adapting techniques developed for performing field antifouling assays (Henrikson and Pawlik 1995) in which crude organic extracts and purified secondary metabolites were incorporated into stable gels that slowly released metabolites (1– 2% per day; Engel and Pawlik 2000) and could be deployed in the field. Although our early attempts showed promise in that we could see obvious effects on corals that had been exposed to gels treated with sponge extracts in the field, there was no simple or nondestructive technique available to quantify these effects until the advent of the diving pulse amplitude modulated (PAM) fluorometer. PAM fluorometry is a noninvasive technique that has been used to assess the photosynthetic properties of the tissues of photosymbiont-containing invertebrates (Beer et al. 1998; Jones et al. 1999). The diving-PAM fluorometer (Heinz Walz) can be used to measure the maximal quantum yield of photosystem II (Y) of zooxanthellae present in dark-adapted coral tissue in the field. After application of a saturating light flash, fluorescence rises from the groundstate value (Fo) to its maximum value, Fm. The potential quantum yield (Y) is then calculated as Y 5 (Fm 2 Fo) Fm (also termed Fv Fm). For our purposes, PAM fluorometry of adjacent, dark-adapted coral tissue that has been exposed to control gels versus metabolite-treated gels yields two useful parameters: Fo, the ground-state fluorescence, a proxy for the number of zooxanthellae in the coral tissue, or relative bleaching as zooxanthellae are lost (see below), and Y, a proxy for the photosynthetic efficiency, or Notes 907