Isolation and self-regulation processes in simulated postfire microsites promote plant species diversity

Abstract

Mechanisms responsible for the high species richness of disturbance-prone floras remain speculative. After fire, speciose shrublands in Australia possess mosaics of microsites that vary widely in seedling density and species richness and provide an ideal context in which to test a) the self-regulation hypothesis that species survive and grow better as they become rarer in the mix, and b) the synergistic hypothesis that species perform relatively better at increasing levels of competition under poorer growing conditions. Both processes should serve to promote species co-occurrence at the local scale. We planted germinants of 10 native shrub species in nutrient-poor sands in 30 cm (high density) and 40 cm (low density) square, buried but open-bottomed, boxes. Each box simulated a field-type microsite, containing one individual (solitary), or 49 individuals with each species contributing three (5% of the total, rare), five (10%, equal), 25 (51%, dominant) or 49 (100%, monoculture) seedlings. Best performance per plant (gauged as % survival × shoot mass per survivor) occurred among all ten species when solitary in the box. When grown in the presence of other plants, mixtures of species performed better on a 1) per plant, and 2) whole microsite, basis than monocultures at both densities, and 3) when rare rather than dominant in the mix, conforming with the self-regulation hypothesis, all independent of species identity. Expected niche differentiation of soil-based resources among species mixes would explain increased fitness per plant with decreasing abundance per species. Overall performance (performance per plant × absolute number of survivors per microsite) was maximized when all species were moderately rare (10% of initial numbers among 10 species) in the mix. Under good (low density) and poor (high density) growing conditions, increasing competition (indexed as total shoot mass) from the dominant species in the mix, led to a gradual fall and merging of shoot mass of the remaining nine species at these two densities, but species richness did not change. This outcome provides only limited support for the synergistic model of species coexistence (species richness should have declined at a decreasing rate) but is consistent with the concept of self-regulation. We conclude that mosaics of microsite types maintain biodiversity of speciose, disturbance-prone ecosystems through both isolating mechanisms (in seed-poor microsites with negligible competition) and compensatory self-regulation mechanisms (in species-rich microsites with intense competition).