| Plant species diversity frequently declines from intermediate to high productivity. I review three competition-based hypotheses to explain the productivity-diversity relationship, and test them in an old-field perennial community. The total competition hypothesis predicts that competition for all resources becomes more intense and increases competitive exclusion. The light competition hypothesis predicts that competition for light becomes more intense while root competition becomes less intense, and light competition increases competitive exclusion. The density hypothesis predicts that increased competition for light kills small individuals of all species, and species are lost randomly from small plots.; Many previous tests of these hypotheses measured the effects of competition on individuals, not on the community. In a first-year old field, I used low-density plots to characterize the community with no competition. Although competition had strong effects on individuals, it had no effect on species diversity. Clearly, community-level experiments are necessary to test these hypotheses.; Two community-level tests support none of the three hypotheses. Artificial shading and fertilization both reduced light, but shading did not decrease diversity while fertilization did. Thus, increased competition for light did not decrease diversity. Fertilized plots lost more species than predicted from random mortality, so the density hypothesis was not supported. When competition was directly manipulated with a combined-monocultures design, diversity dropped with fertilizer in communities experiencing full competition or only root competition, but not in communities with no interactions or only shoot interactions. Therefore, belowground competition alone explained reduced diversity. Neither the total competition hypothesis or the light competition hypothesis predicted this pattern.; Two further experiments explore the mechanisms of root competition in this community. Species that increase in abundance with fertilizer have high root allocation and can further increase root allocation to exploit added nutrients. Size asymmetry of root competition may also contribute to competitive success. At least one species, Bromus inermis, had size asymmetric effects on neighbors in root competition.; Thus, the previous hypotheses to explain the productivity-diversity relationship were rejected, and a new hypothesis was proposed. Diversity may decline with productivity only when competition becomes more size asymmetric, either because light becomes limiting or because small-scale heterogeneity of soil resources increases. |
