Measuring floristic homogenization by non-native plants in North America

Aim To determine if non-native plant species are homogenizing species composition among widely dispersed plant communities. Location Twenty localities in North America. Methods Species lists among localities were compared to measure the influence of non-native species richness at each locality on the Jaccard Index (JI) of similarity between localities. Results After removing the effects of distance, because shared native species decreased with distance, three (nonexclusive) lines of evidence indicate that non-native species promote homogenization. First, pairs of sites with a high combined total of non-native species tend to have higher similarity than those with a low total of non-natives. Second, for a given distance, more non-native than native species tended to be shared among localities. Third, whereas most of the site comparisons with high total non-native richness have a non-native/native JI ratio greater than 1 (often much greater), only half of the comparisons with low total non-native richness have a ratio greater than one. Main conclusions These findings provide quantitative support for the widely held, but rarely tested, notion that non-native species tend to homogenize biological communities because they are more commonly shared among communities. Such testing is important as non-native species could theoretically have no impact or even reduce homogenization among communities, if non-native colonizers consist of different species pools.