The majority of sponge species found in this study are typical mangrove sponges known to inhabit mangrove roots, seagrass beds and adjacent shallow reefs, whereas a few species are usually found in a reef environment. Most species found in this inventory were previously reported on mangrove roots in the Caribbean (e.g., Sutherland, 1980; Wulff, 2004). Species were patchily distributed within transects and several species that dominated at one site were completely absent at another site. This phenomenon is consistent with earlier reports on sponge community structures in mangroves (Rützler et al., 2000; Diaz et al., 2004; Wulff, 2004).
Results on environmental variables in this study reflect single point measurements and fail to reveal the temporal variation of these parameters, which may be large seasonally and even daily. Interpretations of apparent relations should therefore be restricted to environmental variables that show discernible gradients. Statistical data reduction suggests that the diversity patterns found in this study may relate to two variables: distance to the nearest reef and eutrophication.
The correlation between sponge diversity and proximity towards adjacent reefs validates earlier field observations in Belize, including an increased species richness with decreased distance to well developed coral reefs in off shore cays (Rützler et al., 2000), and increased species richness in a coast to barrier reef transect (Ellison and Farnsworth, 1992). A similarity between this study and the Belizean survey of Rützler et al. (2000) is that localities close to reefs harbored more species typical of nearby reefs (e.g. Desmapsamma anchorata Carter, Ircinia strobilina Lamarck and Iotrochota birotulata Higgin), while localities further from the reef are largely comprised of species typical to mangrove habitats (e.g. Tedania (Tedania) ignis, Lissodendoryx (Lissodendoryx) isodictyalis Carter, Haliclona (Soestella) caerulea Hechtel, Halichondria (Halichondria) magniconulosa Hechtel).
Sponge larvae are incapable of traveling large distances and it has been demonstrated in Florida and Belize that epifaunal communities are partly structured by proximities to source populations and larval life span (Bingham, 1992; Farnsworth and Ellison, 1996). This may suggest that the reef partly functions as a larval pool, rather than that the mangrove habitat is self-sustaining. Sponge community composition varies or clusters with respect to proximity towards the reef (Figure 2) yet the clustering is not fully consistent. In some cases, localities close to the reef are more strongly related to localities that lie furthest from the reef, which points to mangrove-derived recruitment as well. This mechanism would parallel other areas (e.g., the Mediterranean), in which larval recruitment is also partly responsible for observed patterns in sponge community structures (Uriz et al., 1998).
Although free-living phases of sponge larvae are short and dispersal abilities are limited, communities should be homogenously distributed after multiple generations. Distributions patterns are therefore controlled by a combination of factors. It has been shown by others that turbidity plays a pivotal role in structuring epifaunal communities (Bingham, 1992; Ellison and Farnsworth, 1992). In this study, statistical data reduction of combined parameters suggests a pattern of decreasing diversity with increasing eutrophication. Previous work has demonstrated the adverse effects of eutrophication on sponges under laboratory conditions (Roberts et. al., 2006) and in other habitats, including reefs and kelps (Hindell and Quinn 2000; de Voogd et al., 2006).
The importance of abiotic factors affecting sponge distributions in coastal and estuarine mangrove habitats has been emphasized in earlier studies, in which decreased species richness was attributed to large variation in temperature, salinity and tidal range (Ellison and Farnsworth, 1992; Rützler, 1995). In addition, transplantation of reef sponges to mangrove habitats was successful in off-shore cays with similar abiotic conditions in Belize (Wulff, 2005), but resulted in the death of transplants when reef sponges were transplanted to coastal mangrove habitats and off-shore mangroves in Belize and Florida (Ellison and Farnsworth, 1992; Wulff, 2004; Pawlik et al. (2007). The inability of reef sponges to survive transplantation to coastal habitats has been attributed to sub-optimal levels of abiotic variables in coastal and estuarine mangrove habitats (Pawlik et al., 2007).
The tannin concentrations of mangrove roots did not meet our expectation, as roots have significantly higher tannin contents when they are covered with sponges as compared with roots without sponge cover. In general, tannins are considered to function as anti-grazing substances, which may suggest that mangrove roots do not favour sponge coverage. However, another study provided evidence that this association is beneficial for roots, as epibiontic cover protects roots from isopod invasion, coinciding with increased growth rates of roots (Ellison and Farnsworth, 1990). In this study, roots were not examined on the extent of isopod presence, yet all sampled roots appeared rather fragile and soft to the touch, indicating that roots were damaged by isopods (Ellison and Farnsworth, 1990). Ellison et al. (1996) posed the presence of a facultative mutualistic relationship between the red mangrove and the sponge species Haliclona (Reniera) implexiformis Hechtel and Tedania (Tedania) ignis. They found a nutrient exchange between both organisms through fine rootlets, in which the sponge obtains organic carbon from the roots, and mangrove roots take up excretory nitrogen from the sponge. However, formation of such rootlets is restricted to only a few species (Ellison et al., 1996) and the release rates of excretory nitrogen are highly variable between species and depend on the size of non-photosynthetic symbiotic bacterial populations and the nature of photosynthetic symbionts (Corredor et al., 1988).
Although it is apparent that roots have increased tannin levels when covered with sponges, it remains unknown whether this increase has consequences for sponge physiology or the association, and whether tannins in roots of R. mangle are produced to act against sponge tissue. These aspects are subject of further investigation. There is evidence that the protein-binding potential of tannins is greatly reduced when the pH is greater than 7,5 (Martin and Martin, 1983, Martin et al., 1985), and the pH always exceeded 7,5 at sites harbouring sponges during this investigation. The ineffectiveness of tannins under similar conditions has also been shown by Benner et al. (1986), who provided evidence that tannin leachates of mangroves did not reduce microbial degradation. In addition, the production of tannins as chemical defence by the brown algae Fucus vesiculosus sometimes fails to affect herbivores and it is hypothesized that other metabolites may be a confounding factor in identifying chemical defences (Kubanek et al., 2004).
Alternatively, tannins may be redox active to metal ions and alter metal uptake, availability and toxicity, and the mangrove polyphenolics can have anti-oxidant properties (A.E. Hagerman, pers. comm., 2006). The greater part of the non-precipitating flavonoids in leaves of R. mangle is comprised of Quercetin, a compound highly effective in scavenging oxy-radicals (Kandil et al., 2004). Tannin leachates may also provide a carbon source for sponges. An increased tannin content may then imply favourable conditions for sponges and larvae seeking a proper substrate to settle on and may favour roots with higher concentrations of tannins in the surrounding water. Future research efforts should elucidate whether increases in tannin concentrations are induced by newly colonizing sponge species or increased tannin concentrations in the surrounding water may act as a cue for attracting sponge larvae. This may provide more insight in the role of tannins in the distribution of mangrove associated sponges.
This study aimed to quantify diversity of mangrove associated sponges in bays of Curaçao and Aruba and to correlate complexity in community structure with environmental variables. Observed patterns validate earlier observations that a combination of physical and biological factors, proximity to source populations and water quality seem important in controlling epifaunal sponge distributions on a larger scale in Caribbean mangrove ecosystems. Heterogeneity in species distributions between roots remains unresolved in many aspects, although data presented here suggest that differences in tannin content may influence the structure of Caribbean sponge communities.