Fine-scale environmental heterogeneity is expected to enhance species diversity within communities through niche partitioning. Observational studies from plant communities report positive heterogeneity-diversity (H-D) relationships, but, surprisingly, results from field experiments generally do not corroborate positive H-D relationships. One explanation is that past soil heterogeneity experiments may be too reliant on the use of fertilizer to create different patch types. Specifically, fertilizer may undermine plant-soil interactions in ways that obscure H-D relationships. To overcome this potential bias, we created soil heterogeneity treatments by reorganizing the natural variation in soils that occurs in the vertical soil profile. Heterogeneous plots were established by excavating three strata from the vertical soil profile and randomly reallocating these soils to 0.5x0.5 m patches within 2.4x2.4 m plots. Homogeneous plots were created by combining the three strata before redistribution into the patches comprising the plots. Thirty-four native prairie species were sown uniformly across all plots prior to the first growing season.
Results/Conclusions
As predicted, species richness was higher in heterogeneous than homogeneous plots in years 1-4 and this difference was driven by species sorting among patch types comprising the heterogeneous plots. However, in years 5-7, richness converged to similar levels in homogeneous and heterogeneous plots. Spatial analysis indicated that by year 5, homogeneous plots had equivalent intraspecific aggregation to that of heterogeneous plots despite the fact that soils and seeds were homogenized at the outset of the experiment. These results, along with a theoretical exploration of plant-soil feedbacks, suggest that plant species endogenously generate spatial structure, independently of soil heterogeneity, within an assembling grassland community potentially explaining the convergence of plant diversity between homogeneous and heterogeneous plots.