Ecological communities are increasingly being recognized as the products of contemporary drivers and historical legacies that are both biotic and abiotic. In an attempt to unravel multiple layers of ecological contingency, we manipulated (i) competition with exotic annual grasses, (ii) the timing of this competition (temporal priority in arrival/seeding times) and (iii) watering (simulated rainfall) in a restoration-style planting of native perennial grasses. In addition, we replicated this experiment simultaneously at three sites in north-central California. Native perennial grasses had 73-99 % less cover when planted with exotic annuals than when planted alone, but this reduction was greatly ameliorated by planting the natives 2 weeks prior to the exotics. In a drought year, irrigation significantly reduced benefits of early planting so that these benefits resembled those observed in a non-drought year. There were significant differences across the three sites (site effects and interactions) in (i) overall native cover, (ii) the response of natives to competition, (iii) the strength of the temporal priority effect and (iv) the degree to which supplemental watering reduced priority effects. These results reveal the strong multi-layered contingency that underlies even relatively simple communities.

Plant communities in abiotically stressful, or 'harsh', habitats have been reported to be less invaded by non-native species than those in more moderate habitats. Here, we synthesize descriptive and experimental evidence for low levels of invasion in habitats characterized by a variety of environmental stressors: low nitrogen; low phosphorus; saline, sodic or alkaline soils; serpentine soils; low soil moisture; shallow/rocky soils; temporary inundation; high shade; high elevation; and high latitude. We then discuss major categories of hypotheses to explain this pattern: the propagule limitation mechanism suggests invasion of harsh sites is limited by relatively low arrival rates of propagules compared with more moderate habitats, while invasion resistance mechanisms suggest that harsh habitats are inherently less invasible due to stressful abiotic conditions and/or increased effects of biotic resistance from resident organisms. Both propagule limitation and invasion resistance may simultaneously contribute to low invadedness of harsh sites, but the management implications of these mechanisms differ. If propagule limitation is more important, managers should focus on reducing the likelihood of propagule introductions. If invasion resistance mechanisms are in play, managers should focus on restoring or maintaining harsh conditions at a site to reduce invasibility.