Spatial variation in species interactions (interaction β-diversity) and its ecological drivers are poorly understood, despite their relevance to community assembly, conservation, and ecosystem functioning. We investigated effects of wildfire severity on patterns and four proximate ecological drivers of interaction β-diversity in plant-bee communities across three localities in the Northern Rocky Mountains (Montana, USA). Wildfires decreased interaction β-diversity but increased interaction frequency (number of visits) and richness (number of links). After controlling for interaction frequency and richness, standardized effect sizes of interaction β-diversity were highest following mixed-severity wildfires, intermediate following high-severity wildfires, and lowest in unburned landscapes, suggesting that wildfire increases spatial aggregation of plant-bee interactions. Moreover, higher effect sizes in burned landscapes were largely determined by turnover in the species composition of both trophic levels rather than by interaction rewiring (spatial turnover in local species interactions not due to species turnover). The underrepresented level of rewiring indicated spatial consistency in post-disturbance patterns of interactions among co-occurring species. Together, our findings suggest that wildfire alters the β-diversity of mutualistic species interactions via linked assembly of plant-bee communities and provide insights into how environmental change alters complex networks of species interactions.

We studied plant-bee interactions in three distinct localities—Helena (HE), Paradise Valley (PV), and Whitefish (WF)—in the Northern Rockies Ecoregion of Montana, USA. Study sites were selected to span mixed-severity fire (MX), high-severity fire (HI), and unburned areas (UN). We quantified plant-pollinator interaction networks at each site by observing floral visitors at open flowers in a 25 m diameter plot (491 m²) weekly for 20 min during times of greatest visitor activity (sunny, calm, 0900-1600) over the growing season (late-May – August, depending on the locality). Pollinators were those floral visitors contacting the reproductive parts of flowers and moving among flowers, and we captured them with hand nets. Plant and floral visitor species presence at a site was determined by their participation in at least one plant-pollinator interaction. We did not consider plant species with flowers that were never visited. Plots in Helena were observed twice in 2013, 12 times in 2014, 9 times in 2015, and 7 times in 2016. Plots in Paradise were observed twice in 2013, 9 times each in 2014 and 2015, and 5 times in 2016. Plots in Whitefish were observed twice in 2013 and 7 times each in 2014 and 2015 (Burkle et al. 2019). Floral visitors were frozen, pinned, and identified to species. Because bees were 89% of individual visitors to flowers, we focused on this group and pooled plant-bee interactions for each site across all observations for analyses.

INTERACTION_FREQUENCY refers to the number of time we observed bees of a particular species (BEE_SPECIES) visiting flowers of a particular plant species (FLOWER_SPECIES) at that site.