Development of negative soil feedback by an invasive plant near the northern limit of its invaded range

Invasions at high latitudes are recognized as an emerging threat to native biodiversity, but non-native plants still are scarce in northern Canada. One factor potentially inhibiting further invasions may be below-ground interactions; in particular, it is unclear whether interactions with soil biota are likely to help or hinder the spread of new species into often challenging northern soils. In the Canadian subarctic, the non-native plant Linaria vulgaris has invaded human-disturbed soils in the town of Churchill, Manitoba (58.8 degrees N) but for decades has failed to spread into natural communities. One explanation for this stasis might be greater resistance by soil communities in uninvaded areas relative to areas where this plant is established; however, no local evidence for plant-soil feedbacks exists. In one of the first papers to test the potential role of plant-soil feedbacks in an invasion at high latitudes, we planted L. vulgaris in soil serially inoculated with live and sterilized field-collected soil sampled from invaded or uninvaded sites, and measured plant performance (biomass) over three iterations. We also conducted soil chemical analyses to determine whether pH, and carbon, nitrogen, and phosphorous contents differ between invaded and uninvaded areas. There was no initial difference in biomass between inoculation treatments in the first two iterations. However, by iteration 3, we found that sterilization significantly increased L. vulgaris biomass in invaded soils, indicating feedback becomes negative in invaded soils compared to uninvaded soils. Soil chemistry did not differ significantly between invaded and uninvaded soils, though there was a tendency for invaded soils to contain more carbon and nitrogen. These results do not support the hypothesis that L. vulgaris is absent from uncolonized sites because soil communities resist invasion. Instead, they provide evidence that L. vulgaris is inhibited by plant-soil feedbacks in invaded soils, while feedbacks in native-dominated soils are not a barrier to further local spread. Thus, explanations for the restriction of this species must lie elsewhere.