Temporal variation in spatial organization of a rodent community in the southwestern Kyzylkum desert (Middle Asia)

The distribution of niches in resource space and the niche patterns of a 14‐species community of Middle Asian desert rodents were studied during two years ‐ at low and high rodent density ‐ using discriminant function analysis Nineteen quantitative environmental parameters (soil structure and vegetation), measured in 550 plots within 22 1 ‐ha grids, were considered The first three canonical axes of resource space account for 72% of the variance The first two axes represent complex environmental gradients the first axis represents a general landscape gradient from sand to clay soils, the second axis reflects a gradient of in creasing productivity The third axis reflects with in‐habitat environmental variation All community parameters, as well as parameters of individual species niches, were unstable between years At the same time, different parameters vary in different extent Position of niche centroids along macro‐habitat axes, as. well as macrohabitat niche breadth, were relatively stable between years, but these parameters for microhabitat axis and values of niche overlap were much more variable A strong correlation between changes m relative between‐habitat niche breadth and differences in average niche overlap with relative changes in species abundances indicate density dependence of these parameters Changes in niche overlap is a consequence of between‐year differences in guild patterns Guild structure was pronounced at high density when the level of niche overlap was intermediate At low density, when the level of niche overlap decreased, guild structure was incon‐spicous Different levels of diversity differed in their sensitivity to density changes α‐diversity was relatively constant as a result of between‐year stability of niche centroid positions However level of ß‐diversity varied significantly between years reflecting changes in the level of niche overlap, because a decrease in niche overlap leads to an increase in the rate of species turnover