A unique 20 year study addressing best practices for ecological restoration across dozens of sites within the Canadian side of the Great Lakes bioregion includes research designed to compare how much influence connectivity has on the outcomes of efforts to restore vegetation and pollinators in meadow and forest habitats. Much of the meadow and forest habitat is relatively isolated by hard edges where each type habitat abuts the other. The remnants of meadows and forests are often small (< 2 ha). Both factors may mean that attempts to restore vegetation may not lead to a concomitant and unaided restoration of pollinators, hence the vegetation communities dependent on pollinators may fail over time. Attempts to translocate pollinators might succeed in the short term - but in small and isolated habitats, inbreeding depression or disease risk might mean that the pollinator populations fail as well. I tested the influence of size of the habitat restored, whether pollinators were translocated, and amount of structural connectivity (measured by percolation) – where it was relatively absent in many cases but restored in others. Analysis was performed by comparing generalized linear models (GLMs) with variance partitioning; this provided explanatory comparisons of all possible combinations of variables. In this manner, I tested and determined the key ecological management features needed for successful restoration of habitat and pollinators.
Comparisons of GLMs indicated that restored or existing structural connectivity plus pollinator translocation were most successful at restoring pollinators, growth of restored vegetation, and successful sexual reproduction measured by pollen transfer and seed set. The key result is that pollinator translocation and/or restoring vegetation alone did not have much long-term impact or success at restoring the pollinators or even the vegetation. The pollinator populations declined significantly over 20 years if reintroduced but still isolated. The plants that need pollinators for sexual reproduction also declined or failed to establish viable populations if they were isolated even if pollinators had been translocated. In situations where pollinators were not translocated but connectivity with habitats with appropriate pollinators was restored, the plants established sexually reproducing and viable populations over 20 years. The long-term dataset and extensive replication in this study makes it more generalizable; the important outcome is that it is not enough to be proficient at restoring vegetation or translocating pollinators. Restoration ecologists should ensure that habitat connectivity is established along with restoration of pollinators and plants.