Testing the intermediate disturbance hypothesis on species diversity in herbaceous plant communities along a human trampling gradient using a 4-year experiment in an old-field

The intermediate disturbance hypothes's predicts unimodal relationships between species diversity and disturbance frequency/intensity. To test this hypothesis, species diversity in herbaceous plant communities along a human trampling gradient was investigated by conducting a 4-year experiment in an old-field. In general, species richness (S), the Shannon-Weiner index (H') from plant clover data and species evenness (J') showed negative linear relationships with trampling frequency, in contrast to the prediction of the intermediate disturbance hypothesis. However, the significant relationships between trampling frequency and species diversity were not observed in the fourth year without J', which showed a unimodal relationship. In all experimental years, the number of new species that colonized the plots after 1 year was small under frequent trampling, and the number of species lost from the plots was large under infrequent trampling. The relative number and the relative cover of perennial species increased as trampling frequency increased in the first and second years, but this pattern was not observed in the following years because the dominance of perennials further increased at decreasing frequencies of trampling. The similarity in the species composition and the yearly changes in species dominance indicated that trampling at higher frequencies eliminated more trampling-intolerant species only in the early years of the experiment. These results suggest that trampling mediated early changes in species diversity patterns, but competitive interactions were more important in the later experimental years. The time lag in the effects of trampling and competition appears to be attributable to the infrequent occurrence of unimodal patterns of species diversity.