Early snowmelt advances flowering phenology and disrupts the drivers of pollinator visitation in an alpine ecosystem

Rose-Person, A, M. J. Spasojevic, C. Forrester, W .D. Bowman, K. N. Suding, M. F. Oldfather, and N. E. Rafferty. 2024. Early snowmelt advances flowering phenology and disrupts the drivers of pollinator visitation in an alpine ecosystem. Alpine Botany. Available at https://doi.org/10.1007/s00035-024-00315-x.

Black sand early and control treatments at Niwot Ridge. Image credit: Cliff Bueno de Mesquita

Abstract

Climate change is altering interactions among plants and pollinators. In alpine ecosystems, where snowmelt timing is a key driver of phenology, earlier snowmelt may generate shifts in plant and pollinator phenology that vary across the landscape, potentially disrupting interactions. Here we ask how experimental advancement of snowmelt timing in a topographically heterogeneous alpine-subalpine landscape impacts flowering, insect pollinator visitation, and pathways connecting key predictors of plant-pollinator interaction. Snowmelt was advanced by an average of 13.5 days in three sites via the application of black sand over snow in manipulated plots, which were paired with control plots. For each forb species, we documented flowering onset and counted flowers throughout the season. We also performed pollinator observations to measure visitation rates. The majority (79.3%) of flower visits were made by dipteran insects. We found that plants flowered earlier in advanced snowmelt plots, with the largest advances in later-flowering species, but flowering duration and visitation rate did not differ between advanced snowmelt and control plots. Using piecewise structural equation models, we assessed the interactive effects of topography on snowmelt timing, flowering phenology, floral abundance, and pollinator visitation. We found that these factors interacted to predict visitation rate in control plots. However, in plots with experimentally advanced snowmelt, none of these predictors explained a significant amount of variation in visitation rate, indicating that different predictors are needed to understand the processes that directly influence pollinator visitation to flowers under future climate conditions. Our findings demonstrate that climate change-induced early snowmelt may fundamentally disrupt the predictive relationships among abiotic and biotic drivers of plant-pollinator interactions in subalpine-alpine environments.