| Natural populations are typically fragmented, but current and anticipated levels of habitat modification will result in smaller, more isolated populations for many species. As a means of studying effects of habitat fragmentation in an accelerated way, I have examined patterns of diversity at both community and population scales in an experimentally fragmented California annual grassland for four years. An equal area (64 m{dollar}sp2{dollar}) was distributed among 2, 8 and 32 exclosed plots isolated by grazing, resulting in three levels of habitat subdivision. Species richness and diversity fluctuated annually in all treatments, but were consistently greater in the most fragmented treatment (32 small plots), as reported in earlier work. Species gains, losses and relative turnover were equivalent among the three subdivision treatments for four years, averaging approximately 3 species replacements per treatment per year. It was concluded from these studies that negative interactions are important in structuring the isolated communities, and that subdivision enhanced community diversity in part by reducing the effects of competition on any given species. In an experiment that tested the invasibility of the experimental habitat fragments, a new species was introduced to all experimental plots. Its success was positively related to local diversity patterns, although generally independent of plot size. It was further determined that factors influencing local species diversity, notably dominance, disturbance and plant biomass, also influenced invadability, and in similar ways. In an examination of effects of fragmentation of genetic diversity, three species with different mating systems were studied in the experimental grassland to examine relationships between habitat structure, plant community structure and population structure (as determined by genetic markers and quantitative traits). For a fully inbreeding grass, the single genetically-determined trait examined, late flowering, was associated with population decline, regardless of plot size. For a predominantly inbreeding grass, diversity was equivalent or higher with greater population subdivision and greater species diversity. For an outcrossing legume, variability was greater with fewer, larger populations, indicating a possible disruptive effect of habitat fragmentation. |
