Climate change is shifting the environmental cues that determine the phenology of interacting species. Plant-pollinator systems may be susceptible to temporal mismatch if bees and flowering plants differ in their phenological responses to warming temperatures. While the cues that trigger flowering are well-understood, little is known about what determines bee phenology. Using Generalized Additive Models, we analyzed time-series data representing 67 bee species collected over nine years in the Colorado Rocky Mountains to perform the first community-wide quantification of the drivers of bee phenology. Bee emergence was sensitive to climatic variation, advancing with earlier snowmelt timing, while later phenophases were best explained by functional traits including overwintering stage and nest location. Comparison of these findings to a long-term flower study showed that bee phenology is less sensitive than flower phenology to climatic variation, indicating potential for reduced synchrony of flowers and pollinators under climate change.

We gathered data at 18 sites around the Rocky Mountain Biological Laboratory (RMBL) in the Elk Mountains of western Colorado, USA from 2009 to 2017. Sites were located along an elevation transect (2456-3438 meters above sea-level) in montane and sub-alpine habitats dominated by a diverse mixture of perennial flowering species. We sampled bees in habitat types that were representative of dominant vegetation types: wet meadows dominated by Veratrum tenuipetalum, those dominated by Salix spp., rocky dry meadows, and Artemisia spp. steppe. We conducted biweekly bee abundance surveys at each site using pan traps (following LeBuhn et al. 2003). We set out 10 each of white, fluorescent yellow, and fluorescent blue pan traps along two approx. perpendicular 45-meter transects at intervals of 3 meters, an array that passively attracts bees by mimicking a display of flowers. We deployed pan traps between at approx. 0800 and 1700 (the period of maximum bee activity) only on warm, calm, sunny days and removed traps when these conditions changed drastically. Further details of the bee sampling are provided by Gezon et al. (2015).

The catches per day and sampling method are given, and the data processing code is included.

Several taxonomic groups and sites were excluded for the analysis. See the manuscript for details on which species and sites were used. We encourage other researchers to use these data for further studies, and also to contact Michael Stemkovski (m.stemkovski@gmail.com) or Rebecca Irwin (reirwin@ncsu.edu) for insight on the ideosynchrosies of the dataset. We wish to enable others to do work using these data. The long-term bee phenology monitoring project is an ongoing effort that is headed by Rebecca Irwin, so we encourage users to get the most up-to-date data at https://osf.io/kmxyn/. The repository on Open Science Framework includes more recent years of data since the analysis of the present dataset.