Community-wide impacts of a dominant invasive plant on native plant-pollinator interactions

Background/Question/Methods

Invasive plants compete with native plants for pollination when pollinators preferentially visit the flowers of the invader. Over a broader scale, invaders may facilitate native plant pollination by attracting visitors to the invaded site or by increasing population sizes of pollinators at the site. Reviews of studies of invasive-native pairs have concluded that competition for visitation is more prevalent than facilitation, yet these studies may reflect researcher bias in selecting pairs likely to compete based on floral traits or pollinator overlap.  We assessed whether the invasive plant Lythrum salicaria competes for or facilitates visitation to native plants by comparing pollinator visitation to all plants in seven invaded and seven uninvaded communities. We also removed L. salicaria flowers in three of the sites in a subsequent year.  We measured flower and visitor densities, predicting lower visitor densities per native flower in invaded sites.  Effect size (LnRR) of L. salicaria on flower visitation was calculated for all native species that occurred in both invaded and uninvaded sites and used to test whether native species with flowers phenotypically similar to L. salicaria in morphology and color competed more strongly than species with dissimilar flowers. Plant height and relative abundance were additional explanatory variables. 

Results/Conclusions

Visitation to 126 species of native plants was lower in invaded sites than uninvaded sites. Furthermore, invasion altered the relationship between visitor density and native flower density. In uninvaded sites, visitor density was negatively related to flower density, suggesting that visitors may be limited at high flower densities.  In invaded sites, visitor density was uniformly low across the whole range of native flower densities.  Removal of L. salicaria flowers resulted in visitor densities that were as high as, but not higher than, those in uninvaded sites, refuting the hypothesis that invasion of L. salicaria augmented pollinator populations. The average effect size of L. salicaria on visitor density for 39 native species was moderately, but significantly, less than zero (LNRR=-0.81). These results support competition rather than facilitation for pollinators between invader and native plants.  Analysis of floral traits revealed stronger negative effects on visitation for species with similar flower color to L. salicaria than those with dissimilar flower color, supporting the idea that competition results from foraging preferences expressed at a relatively small scale by resident pollinators. Other flower traits, plant height, and relative abundance did not influence size of the effect on native species.