Adjacent forest fragments were all less than 10 ha, and bordered by roads cultivated fields, and/or rivers. Transects were placed in crop(cultivated) habitat, 11) adjacent forest fragments, and 1110 adjacent weedy "edge" habitat, with pitfalls placed in the ground every m. Transects in forest fragments followed contour lines of the often steep terrain; transects in crop, forest, and pasture areas were roughly parallel. For each transect, pitfall cup traps(414 ml were dug in between 3 and 5 pm and filled with approximately 25 ml of denatured 96% ethanol and left overnight. Contents of the pitfall traps were collected the following morning and placed in plastic bags, and arthropods were sorted to order and stored in 70% ethyl alcohol at the Rancho Mastatal Environmental Learning Center. Beetles were subsequently identified to family at the Instituto Nacional de Biodiversidad (INBio) in San Jose and at the University of Washington, Tacoma (Borror et al. 1992, Solis 2002, and spiders were sorted to family at the University of Washington, Tacoma (Levi& Levi 1968, Kaston 1972). Simpson's diversity index was calculated for beetles and spiders for each habitat, with data from each transect pooled together. Indices were then compared across habitat types using analysis of variance (ANOVA) (Sokal & Rolf 2000). Because data from the different samples at different times are not independent replicates, data from all three (two in the case of spiders samples were pooled together for ANOVA, with three replicates of each habitat type per sample(Magurran 2004). In order to explore the composition of diversity in each habitat type, the Serensen quantitative index was calculated for both beetles and spiders for each habitat. This measure incorporates both diversity and abundance data, and therefore provides an effective means of comparing the composition of specimens found in each habitat type (Magurran 2004).