The effect of seed predators on plant distributions: is there a general pattern in mangroves?

Selective crab predation has been hypothesized to maintain mangrove forest zonation by preventing the establishment of certain species' seedlings at p articular tidal elevations. Support for this hypothesis consists of observa tions that the propagules of some mangrove species are consumed at a rate t hat is inversely related to the species' relative abundance in the adult ca nopy. In addition, one previous study found that rates of crab predation we re lower in large light gaps than in small ones. We conducted a mensurative field experiment to assay rates of crab predation on dispersing propagules of the dominant mangrove species (Avicennia germinans Laguncularia racemos a, and Rhizophora mangle) in our study forests at Punta Galeta on the Carib bean coast of Panama. The experiment specifically examined whether rates of crab predation varied 1) with species of propagule, 2) with changes in sta nd composition along the tidal gradient, and 3) inside versus outside of li ght gaps. Lightning-created canopy gaps are the primary sites of tree regen eration in these and other tall-stature mangrove forests. Propagules of the three mangrove species were tethered inside and in the un derstory surrounding nine light gaps, with representatives in each of the m ajor stand types that occur along the intertidal gradient. Fates of propagu les were monitored over a four-wk period. Rates of propagule predation vari ed spatially, with highest rates of removal occurring in lower intertidal R . mangle/L. racemosa stands, where the large herbivorous crabs, Ucides cord atus and Goniopsis cruentata are common. In mid- and upper intertidal stand s, dominated by A. germinans or L. racemosa, where detritivorous Uca spp. a re abundant but larger herbivorous crabs are rare, few propagules were cons umed. The propagules of A. germinans were preyed upon more heavily than tho se of the other two species; rates of predation on R. mangle and L. racemos a propagules did not differ. Light environment (i.e. inside versus outside a gap) had no significant effect on the rate of propagule predation. The en hanced survival of juvenile mangroves in light gaps that we have observed i s probably due to other factors such as higher light and nutrient availabil ity; and in some situations, lower insect herbivory. Only the propagules of A. germinans experienced a pattern of predation cons istent with the dominance-predation model: they were consumed at higher rat es in the low intertidal where adults of the species are rare, and at lower rates in high intertidal areas where the species dominates the canopy. Thi s suggests that predation on mangrove propagules by herbivorous crabs could influence the species composition of low intertidal forests by differentia lly reducing the recruitment of A. germinans On the other hand, rates of cr ab predation on A. germinans propagules seem insufficient to prevent A. ger , from establishing in low intertidal areas. In opposition to the dominance -predation hypothesis, crab predation cannot account for the distribution p atterns of R. mangle or L. racemosa whose propagules were preyed on most he avily in lower intertidal areas where these species are very abundant. Our results, together with those of two earlier studies, indicate that predatio n by crabs on dispersing propagules is not a general explanation for the zo nation of tree species along tidal gradients in mangrove forests.