Control of seedling recruitment by land crabs in rain forest on a remote oceanic island

Ecology, Dec, 1997 by Peter T. Green, Dennis J. O'Dowd, P.S. Lake

The exclusion experiment

Exclusion of red crabs. - The exclusion experiment consisted of 10 pairs of 5 x 5 m plots; plots in each pair were separated by a 3-m walkway. In each pair, red crabs had access to one plot (control plots), but were denied access to the other by a "semipermeable" fence (exclusion plots). The pairs were irregularly spaced over a distance of -600 m across the slope at MH.

Each exclusion plot was made from a 25 cm wide fence of sheet metal, nailed 10 cm above the ground onto supporting wooden posts. Wire mesh (12.5 x 12.5 mm aperture) was attached from the sheet metal to the ground and then bent outwards to form an apron 30 cm wide around the exclosure. These fences prevented crabs from gaining access to the exclusion plots, but allowed the movement of smaller invertebrates and vertebrates (the geckos Cyrtodactylus sp. nov. and Lepidodactylus listeri, and the skinks Emoia nativitatus and Cryptoblepharus egeriae) in and out of the exclosures. Other than red crabs, the only species excluded were robber crabs and little nippers (Geograpsus grayi), but both were rare at the site (Green 1997). To control for any shading effects generated by the exclosures, we suspended a 25 cm wide strip of heavy black plastic from fishing line strung between supporting wooden stakes around the control plots. This left a 10 cm space above the ground that allowed crabs unimpeded access to the control plots.

We constructed seven paired plots beneath closed canopy ("understory"), and three in light gaps created by breaks in the canopy ("gap"). The number of suitably sized gaps at MH limited the number of pairs used. The three pairs were sited in gaps of 81, 88, and 133 [m.sup.2], larger than the median gap area at MH (40-60 [m.sup.2]; Green 1996); this was necessary to accommodate the rectangular study blocks. The pairs were sufficiently spaced so that the same large tree(s) did not contribute to the overhead canopy of more than one pain Pairs beneath closed canopy received less light (0.26-0.66% full irradiance) than those in gaps (0.86-1.27% full irradiance, see Green 1993).

The exclusion fences were closed on 4 January 1990, denying red crabs further access to the exclusion plots, but at the same time trapping any crabs living in burrows on the plots. We rapidly reduced the density of red crabs on the exclusion plots by repeatedly visiting them over the first few weeks, and evicting any surface-active crabs (burrows were never excavated to capture crabs). Thereafter, crabs were evicted from the plots several times each month over the duration of the experiment. This reduced the mean density of active crabs on the exclusion plots to between 0 and 2 crabs/25 [m.sup.2] throughout the experiment. The density of red crabs on the control plots varied between 0 and 15-20 active crabs/25 [m.sup.2], with peaks during wet weather. Overall, we maintained crab density on the exclusion plots at [approximately]10% of control levels.

Monitoring of seedlings. - The plots were trampled to varying degrees during construction of the exclosures, killing a small number of seedlings [less than]5 cm tall. To standardize the initial density of these seedlings in all plots, all of the remaining seedlings ([less than]10 seedlings per plot) were removed at the start of the experiment. Thereafter, we monitored seedling emergence and survivorship each month for 23 mo (January 1990 to November 1991), and again in May 1992. At each census, all seedlings occurring on both the exclusion and control plots were counted and identified to species. New seedlings were marked with color-coded bamboo skewers to identify monthly cohorts.