Species extinctions and the resulting loss of biodiversity are one of the largest impacts humans are making on ecosystems. Biodiversity-ecosystem function research has focused on potential changes in ecosystem functioning that this loss of species may impose. Species extinctions do not happen at random, however, and the identity of the lost species may determine changes that the ecosystem undergoes. I investigated the role that plant functional group composition plays in determining ecosystem function and whether this role is influenced by changes in environmental conditions. To investigate this question, I established a removal experiment in a natural grassland in northern Canada in 2003 with four continuous removal treatments: a control (no-removal), and removal of legumes, non-leguminous forbs, and grasses. As biodiversity loss is occurring in concert with environmental changes, I crossed removals with a fertilization treatment to determine if the role of a functional group was determined by the environment in which it is found.

Results/Conclusions

I examined a variety of ecosystem properties, including soil moisture, light interception and nutrient availability between 2003 and 2006. Removal treatments had a significant effect on all variables examined, but effects depended highly on year. Removal plots had higher soil moisture in general, but only removal of forbs consistently created a soil moisture effect. All removal treatments decreased light interception initially, with the effect all removals except forbs disappearing within 4 years. Significant effects of removal were also found for soil nutrients including total N, NO3-, P, and Fe. Interactions between removal treatments and fertilization were rare, indicating that the role of a functional group will likely remain consistent through environmental changes. This study highlights the importance of species composition in biodiversity-ecosystem functioning research, but also the importance of multi-year experiments, especially in northern ecosystems.