Temporal occupancy distributions reveal multifaceted and heterogeneous effects of climatic variation on montane butterflies

Abstract

Climate change has substantially altered the phenology of many organisms, with profound implications for biodiversity and species interactions. However, phenological studies of insects have often focused on shifts over time, without explicitly examining how climate drives these changes. In addition, the metrics used to measure changes in phenology often capture only limited information about the flight period. In this study, we analyzed multidecadal observational and climate data using a hierarchical Bayesian framework to model the annual probability of occurrence distributions for 135 butterfly species across five montane sites along an elevational gradient. Our analysis used polynomial models that account for shifts in abundance and the timing and length of flight periods to investigate the effect of climate on butterfly phenology. We found that spring maximum and minimum temperatures, as well as winter precipitation, are important predictors of butterfly phenology. High winter precipitation delayed phenology at high-elevation sites where substantial snowfall occurs, while increased spring maximum temperatures generally advanced phenology across all elevations. Even modest increases in spring minimum temperatures caused substantial shift in phenology. We documented variability in the effect of climate on phenology across sites, among species within a site and among populations of the same species across different sites with variability among species within a site being especially pronounced. We also found that climate influences different aspects of the flight period differently (e.g., timing versus duration), underscoring the need to move beyond single metrics such as day of first flight. These findings highlight the importance of examining the entire flight period and considering the interplay of species-specific traits to improve predictions of how climate change impacts phenology. Such approaches might be essential for designing more targeted and effective conservation strategies in response to ongoing climate change.