Alien plants can be associated with a decrease in local and regional native richness even when at low abundance

The impacts of alien plants on native richness are usually assessed at small spatial scales and in locations where the alien is at high abundance. But this raises two questions: to what extent do impacts occur where alien species are at low abundance, and do local impacts translate to effects at the landscape scale? In an analysis of 47 widespread alien plant species occurring across a 1,000 km(2) landscape, we examined the relationship between their local abundance and native plant species richness in 594 grassland plots. We first defined the critical abundance at which these focal alien species were associated with a decline in native alpha-richness (plot-scale species numbers), and then assessed how this local decline was translated into declines in native species gamma-richness (landscape-scale species numbers). After controlling for sampling biases and environmental gradients that might lead to spurious relationships, we found that eight out of 47 focal alien species were associated with a significant decline in native alpha-richness as their local abundance increased. Most of these significant declines started at low to intermediate classes of abundance. For these eight species, declines in native gamma-richness were, on average, an order of magnitude (32.0 vs. 2.2 species) greater than those found for native alpha-richness, mostly due to spatial homogenization of native communities. The magnitude of the decrease at the landscape scale was best explained by the number of plots where an alien species was found above its critical abundance. Synthesis. Even at low abundance, alien plants may impact native plant richness at both local and landscape scales. Local impacts may result in much greater declines in native richness at larger spatial scales. Quantifying impact at the landscape scale requires consideration of not only the prevalence of an alien plant, but also its critical abundance and its effect on native community homogenization. This suggests that management approaches targeting only those locations dominated by alien plants might not mitigate impacts effectively. Our integrated approach will improve the ranking of alien species risks at a spatial scale appropriate for prioritizing management and designing conservation policies.