Climate change and butterflies: Can we use long-term data to separate direct effects on individuals from plant-mediated indirect effects?

Climate change is a contributing factor to the decline of butterflies in the Western United States. However, climate change is not one cohesive phenomenon. It can be decomposed into a network of direct and indirect pathways including rising temperature, altered precipitation patterns, and increasingly frequent extreme events. Using the longest insect monitoring dataset in North America, remote sensing data, and high-resolution climate data, we explore multiple hypotheses about the direct and indirect impacts of climatic conditions on butterflies in the same year and in the following year. Specifically, we use structural equation modeling to quantify the relationships between weather, plant phenology, and butterflies. We show that changing conditions are having a negative impact on butterflies in relatively pristine montane regions. While we find considerable heterogeneity (among species and sites) in the nature and magnitude of weather impacts, a temporally-lagged but direct effect on butterflies in the following year is common, suggesting a role for climatic conditions on overwintering stages. These results demonstrate the complexity of climate driving butterfly declines in montane regions and the importance of considering lagged effects for predicting population responses.