The relative importance of the species pool, productivity and disturbance in regulating grassland plant species richness: a field experiment

1. Ecologists have generally focused on how species interactions and available niches control species richness. However, the number of species in the regional species pool may also control richness. Moreover, the relative influence of the species pool and species interactions on plant richness may change along productivity and disturbance gradients. 2. We test the hypothesis that many species from the propagule pool will colonize into habitats of moderate productivity and moderate disturbance, and the number of species in the pool will, therefore, primarily control plant species richness in such habitats. The hypothesis also states that few species from the propagule pool will colonize into habitats of high productivity and minimal disturbance because competitive species interactions primarily control plant richness in such habitats. 3. To test this hypothesis, we experimentally varied resource availability via fertilization and irrigation, the size of the available propagule pool via sowing the seeds of 49 species and disturbance via vegetation clipping. 4. A larger propagule pool increased species richness 80% in the absence of fertilization and the presence of clipping but had no significant effect on richness in the presence of fertilization and the absence of clipping (significant fertilization x clipping x seed addition interaction). Irrigation increased species richness primarily in the absence of fertilization and the presence or clipping (significant fertilization x clipping x irrigation interaction). 5. Synthesis. These results support the assertion that the control of plant species richness shifts from the number of species in the propagule pool to the intensity of species interactions as productivity due to soil fertility increases and disturbance decreases. However, increases in productivity due to irrigation caused an increase in species richness, probably because of reduced seedling desiccation. Other studies have found similar outcomes along natural productivity gradients where productivity is confounded with other variables, but this study is the first to find such an outcome along an experimental productivity gradient. These results suggest habitats with moderate soil fertility and disturbance may be more open to colonization from the propagule pool than other habitats, a finding that has implications for plant restoration and invasive species management.