Litter Moisture Content as a Determinant of Litter Arthropod Distribution and Abundance During the Dry Season on Barro Colorado Island, Panama

Abstract

Animals generally are patchy in time and space. The causes of this pattern have been explored in only a few cases. We here present data on the relative amounts of migration and activity by litter arthropods during the dry season in artificially watered vs unmanipulated patches of litter on the forest floor of Barro Colorado Island, Panama. Watering significantly changed the physical environment of the litter. Litter moisture content increased by 37-70 percent (absolute increase 8-16%) and soil moisture content by 16-24 percent (absolute increase 4-8%). More and smaller arthropods were found in wetter patches. This effect was weak or variable in watered patches that had unmanipulated arthropod populations and strong in patches from which we had removed arthropods at the beginning of the experiment. Ant colonies were much more likely to migrate into wetter, empty areas, and more individuals were collected in these plots. Overall, more species of ants and more groups of arthropods were found in watered, empty plots. We thus demonstrate that at least part of the variance in plot-to-plot abundance of litter arthropods is related to litter moisture content. This may have strong effects on the community structure of litter arthropods. PATCHY SPECIES DISTRIBUTIONS are a common subject of biological investigations (review in Wiens 1976). Many species or groups are patchily distributed, often with no obvious biological explanation (Hairston and Byers 1954, Lloyd 1967). Litter and soil arthropods have been used as examples of this phenomenon (e.g., Macfadyen 1952, Lloyd 1963, Usher 1975). Soil or litter moisture content, insolation, or other abiotic factors frequently are suggested to be important in the formation of uneven distributions among sites (Usher 1970, 1975), but there have been only scattered experimental approaches to this question (Lloyd 1963, Gill 1969, Lussenhop 1976, Vandermeer et al. 1980, Whitford et al. 1981). We here present experimental data on arthropod abundance in leaf litter under a tropical semideciduous forest. Using 30 months of data, we previously suggested that the length of the rainy season and the distribution of rain showers in the dry season strongly affect litter arthropods (Levings and Windsor 1982). More litter arthropods were collected (1) when there was a short wet season with low late-wet-season rainfall and (2) when the dry season was punctuated with frequent, small rainshowers. Litter moisture content is more variable in the dry season than the wet season (Levings and Windsor 1982); wetter areas of litter are common on the forest floor during the dry season (pers. obs.). Small areas of moist litter may provide refuges from locally dry conditions for desiccation-intolerant species, permit continued reproduction or foraging activity, or be necessary for successful aesti-