Desert plants are thought to be strongly affected by patterns of precipitation and water availability; however few studies have examined how communities respond to variations in the spatial distribution of soil water as affected by soil depth and the amount and timing of precipitation events.  We asked how community cover, diversity, and species assemblages of the sagebrush steppe respond to long-term manipulations of soil depth and a doubling of average precipitation.  To provide an indication of how communities on the plots were affected by functional differences among species, species-abundance distributions (SADs) of plots were compared with SADs generated with a simulation that assumed functional equivalence among species.  Experimental plots were 8 x 8 m and had one of four soil profile types; 1-m loam, 2-m loam, and two profiles that included a 0.5-m gravel/cobble barrier placed at different depths within the profile (0.5- and 1-m Cobble).  Precipitation treatments included ambient precipitation or 200 mm of irrigation applied in either the growing or dormant season.  Community properties and species composition were sampled from 1997-2007 using the point-frame method. Repeated measures MANOVA was used to determine differences in community properties.  Canonical Correspondence Analysis was used to determine differences in species composition.

Results/Conclusions

The deepest soil profiles (2-m loam, 0.5-m cobble, and 1-m cobble) had greater community cover and a higher value of the inverse Simpson’s diversity index (sensitive to common species) than the shallowest soil (1-m loam), however there were no differences between species richness, evenness, or the value of the Shannon-Weiner’s diversity index.  Regardless of seasonal application of the irrigation treatment, a doubling of average precipitation increased community cover and inverse Simpson’s index but not the Shannon-Weiner’s index, species richness, or species evenness.  As a whole, the species assemblages of experimental communities were only marginally affected by the supplemental precipitation treatments; with differences occurring between control and irrigated plots but not between plots irrigated in growing or dormant seasons.  Comparisons of SADs created from species compositions in 2007 were substantially steeper and more resembled a log-series distribution in all plot types than were SADs resulting from the simulation that assumed functional equivalence among species.  These results suggest that plot diversity results in part from functional differences among species.  However, whereas the amount, timing, and storage (soil depth) of precipitation should relate to the primary functional differences among species, they had only minor affects on diversity in these desert systems.