Background/Question/Methods Restoration of terrestrial ecosystems often focuses on increasing native plant diversity and abundance, likely prompting responses from an interacting community of mutualists and consumers. Yet we still know little about the extent and nature of such responses, especially in cases where an extant native plant community persists only in tiny fragments, with the fate of historic interactors unknown. We assessed pollinator community responses to restoration (via restored fire disturbance and increased native plant propagule pressure) in the highly fragmented western Washington prairies. We followed a chronosequence of six prairie restoration sites (0-12 years of restoration) over a four-year period, collecting plant and pollinator interaction data at four seasonal timepoints. Specifically, we documented plant-pollinator interactions with 30-minute timed observations at floral patches of each forb species in the community and simultaneously documented sitewide floral resources. Floral visitors were collected and identified. The extent and frequency of prescribed fire, direct seeding of native forbs, plugging of forb seedlings, and grass-specific herbicide application at the sites were gleaned from management records. We tested effects of these variables on floral diversity, pollinator diversity, and network metrics such as nestedness and modularity. Results/Conclusions We documented 8,943 plant-pollinator interactions. Interactions were dominated by bees (48.7%) and flies (37.9%), including 16 syrphid fly genera and 24 bee genera. Wasps, beetles, and other flower visitors made up only a modest proportion of observations. Floral diversity and abundance was strongly correlated with restoration actions, and pollinator diversity increased as floral diversity increased. Nestedness of plant-pollinator interactions decreased as fire frequency and seeding rate increased. We did not detect strong effects on other network metrics that might reflect changes in pollinator foraging behavior, such as modularity and H2. The effects of restoration varied with seasonal time point, with late-season communities dominated by pollinator interactions with non-native plants. Restoration that increases floral resources is correlated with an increase in pollinator diversity and reorganized plant-pollinator community interactions, even in locations where historically associated pollinator populations are unlikely to have persisted. Future work should establish the longevity of these changes as restoration and management continue.
