Climate Variability and Plant-Pollinator Networks in the Cascade Range, Oregon

Worldwide, networks of plants and pollinators are faced with the threat of climate change. The extent of this threat and the degree of adaptability is not yet understood. In Oregon, climate change is predicted to bring hotter and drier summers which may have consequences for pollinators and the resources they rely on. This study examined a system of wild bees (solitary bees, Bombus spp., and feral Apis mellifera) and floral resources in montane meadows of the western Cascade Range, where tree encroachment has reduced meadow area by more than half since 1950. The analysis tested how climate variation was related to frequency of flowering, plant-pollinator interactions, plant phenology, and interaction timing over multiple weekly sampling periods in summers of nine years (2011 to 2018, 2021) in ten sub-plots in each of twelve montane meadows ranging from 0.25 to 4.4 ha at 1,308 to 1,536 m elevation in the H.J. Andrews Experimental Forest, western Cascade Range, Oregon. Study sites included mesic (n = 5), wet (n = 2), and xeric (n = 5) meadows in three meadow complexes separated by up to 10 km. Climate and weather were characterized using cumulative degree days and antecedent precipitation from meteorological stations in the Andrews Forest. Network structure was characterized using the bipartite function in R and results were interpreted to assess how climate variability is related to plant-pollinator network structure. Beta diversity (Sørensen index) was calculated using the betapart function in R and resulting values of turnover and species replacement were related to climate variables to assess how climate may influence network rearrangement via rewiring (defined as adaptations leading to the formation of new mutualistic relationships) and turnover (defined as extirpation or exclusion following the inability of species to adapt).

The most frequently occurring flowering species in the network were Eriophyllum lanatum, Gilia capitata, and Orthocarpus imbricatus. The most frequent pollinators were Apis mellifera, Bombus mixtus, and Bombus bifarius. The most frequent interactions were between Apis mellifera and Gilia capitata (20% of total), Apis mellifera and Eriophyllum lanatum, Bombus mixtus and Delphinium nuttallianum, and Bombus bifarius and Orthocarpus imbricatus. Species composition and interactions for both bees and flowers varied greatly within and between years in all twelve meadows. Floral abundance and the length of the flowering period were inversely related to heat (cumulative degree days) in mesic meadows but less so in wet meadows and not at all in xeric meadows. Bee abundance was not related to heat (cumulative degree days), and neither flower abundance nor bee abundance was related to moisture.

The network was characterized by high levels of redundancy and significant rewiring (Sørensen’s pairwise dissimilarity > 0.5 between all years) in plant-pollinator interactions over time. Most bee species were generalists (meaning they visited multiple flower species), but the network included specialist bee species including Dufourea calochorti and Dufourea trochantera. Turnover (species replacement) of flowers was positively related to turnover of plant-pollinator interactions. Turnover of flowers was positively related to differences in air temperature between years, but turnover of plant-pollinator interactions was not related to this
measure of climate variability. Although neither bee abundance nor turnover of plant-pollinator interactions was directly related to climate variability, shortened flowering periods during hotter and drier summers associated with climate change may reduce bee abundance and enhance turnover in floral composition, potentially speeding turnover in bees, particularly in specialists, which may not be able to adapt via rewiring.