A test of the niche dimension hypothesis in an arid annual grassland.

The niche dimension hypothesis predicts that greater numbers of limiting factors can allow greater numbers of species to coexist through species' tradeoffs for different limiting factors. A prediction that follows is that addition of multiple limiting resources to plant communities will increase productivity and simultaneously decrease diversity. Species loss due to limiting resource enrichment might occur through reducing the number of resources that species compete for or by changing the identity of limiting factors. We tested these predictions of the niche dimension hypothesis in an arid annual grassland by adding combinations of nutrients: nitrogen (N), phosphorus (P), and potassium with other elements (O). We found that species number decreased while biomass increased with greater numbers of added resources. In particular, N in combinations with P or O resulted in the greatest species loss, while biomass increased super-additively with N and P together. The addition of greater numbers of added nutrients decreased the availability of light and soil moisture, consistent with a potential shift in the identity of limiting resources. Species also differed in their responses to different combinations of N, P, and O, supporting predictions of resource-ratio tradeoffs. These results are particularly notable because this experiment was conducted during a drought year in an arid grassland (226 mm annual rainfall), which might have been expected to be water-rather than nutrient-limited. Our results support the hypothesis that plant diversity may be maintained by high-dimensional tradeoffs among species in their abilities to compete for multiple limiting factors.