Luquillo LTER

Paired elevational (strata) transects (300 m to 1000 m) in the Sonadora River watershed were sampled at 50 m intervals to decouple underlying environmental gradients (abiotic variation and vegetational variation) that are associated with changes in elevation and that are hypothesized to structure animal communities. One transect (mixed forest) reflected changes in abiotic and biotic conditions, including forest type (i.e., tabonuco, palo colorado, and elfin forests), whereas the second transect reflected changes in environmental conditions but not forest type, as its constituent plots were all located within palm dominated forest. Each stratum from the mixed forest transect was paired with a stratum in palm forest. In the mixed forest transect, plots were arranged in a 0.1 ha rectangle (50 m by 20 m) comprising 10 circular plots (3 m radius) configured in 2 parallel rows, each containing 5 evenly-spaced plots with 10 m spacing between plot centers. Rectangular strata were aligned with their long axes perpendicular to the Sonadora River. This transect is the same as used in Barone et al. (2008). Because the elevational distribution and size of palm forest patches were variable, palm forest strata were located within 1 km of the corresponding stratum in mixed forest, and always within the Sonadora watershed. By necessity, the configuration of plots within strata of palm forest was more variable than that in mixed forest. Nonetheless, plots were always contiguous, with a maximum distance between centers of the most distant plots being less than 58.3 m (mean, 43.6 m; standard deviation, 6.07m), comparable to the situation along the mixed forest transect, where 41.2 m was the maximum distance between plot centers. Walking sticks at each plot were sampled 4 times during the summer of 2007 (wet season) from the mixed forest transect and 3 times during the summer (wet season) of 2008 from the mixed forest and palm forest transects. Sampling was conducted at night (2000-0400 h). Each time a plot was surveyed, 2 people searched all available surfaces (e.g., soil, litter, rock cover, vegetation, debris) up to a height of approximately 3 m for 15 minutes or until all substrates had been completely searched, whichever was longer. Sampling was not repeated at any one elevation until after the entire gradient was sampled.