[John Anderson added the following from info provided by Justin Van Zee and edited by Michelle Buonopane.] Overview: In semiarid ecosystems the diversity of plant functional types ( grasses, shrubs, succulents, and so on) and of species may interact with the severe stresses imposed by the desert environment to influence ecosystem processes. Erosion and transport of surface sediment by wind and water is one process that may be affected by the physical structure of the plant community. The Jornada plant diversity experiment, in which the diversity and structure of the plant community have been manipulated in large (25 m x 25 m) plots, offers the opportunity to examine the relative importance of vegetation characteristics and landscape position in determining rates of sediment movement within the plots. Soil erosion is also an important indicator of relative disturbance effects of the treatment manipulations. Each of the 48 plots of the plant diversity experiment contains 5 pans or trays on the downslope side; these accumulate sediments and plant litter that are moving within the plots (carried by wind or by water). Data have been collected on the amount of sediment accumulated in the pans during rainy and during dry seasons, with material sorted and weighed as fine (< 2 mm diameter) or coarse > 2 mm) mineral sediment, plant litter, or rabbit/jackrabbit pellets. Previous statistical analyses found that the mass of material collected per plot is explained only poorly by the treatment (plant community manipulation) of the plot and by block (a rough indication of location on the slope). Objectives: We will test the relative significance of the following variables in explaining plot-level accumulations of sediment and litter: treatment, block, position on slope (the row, from 1 (top of slope) to 10 (bottom), in which the plot is located), the treatment of the plot immediately upslope from the plot, and indices of plant cover and volume (total and by functional group) from the plot-level sampling of vegetation (using fall 1997 data). In addition, we will test the significance of the following variables in explaining the accumulation of sediment and litter in individual pans within a plot: all variables listed above for the plot, plus indices of the vegetative cover and volume located immediately upslope of the pan (weighted for distance from the pan itself).