Insights from citizen science into the spatiotemporal dynamics of Batesian mimicry in the context of climate change
OA Version
Citation
Abstract
Climate has a demonstrable impact on species distributions, with changes in climatic oscillations often producing complex downstream effects due to species-specific tolerances to changes in temperature and precipitation. In general, barring physiological or environmental barriers, species are predicted to migrate polewards, and possibly also upwards in elevation, in response to general equatorial warming (Walther et al. 2002). A major challenge to substantiating this predicted relationship between temperature and distribution, especially for highly mobile or nomadic species, has been the lack of suitable techniques for tracking species distributions with sufficiently high resolution to assess for evidence of range shifts over decadal timescales. However, the recent rise in the volume and availability of citizen science data has demonstrated its strong potential as a tool for elucidating responses to environmental change on large temporal and spatial scales (Champion et al. 2018; Hurlbert & Liang 2012; Soroye et al. 2018). Here we assessed the strengths and weaknesses of three different citizen science datasets in addressing questions relating to historical range shifts in two butterfly species endemic to the eastern United States: the pipevine swallowtail (Battus philenor) and its Batesian mimic, the red-spotted purple (Limenitis arthemis astyanax). Motivated by the prediction that there would be a quantifiable northward latitudinal range shift in both B. philenor and L. a. astyanax in response to historical regional warming, we ultimately analyzed twenty-four years of observational survey data from the North American Butterfly Association and found evidence to suggest that, surprisingly, from 1998 to 2021, there is no indication of a directional range shift in either the model or its mimic. However, consistent with expectations from classic mimicry theory we found evidence of a tight historical correlation between the ranges of B. philenor and L. a. astyanax over the years surveyed. Furthermore, we found that the annual variance in the ratio of models to mimics is significantly lower at the model’s northern range limit than in other parts of its range. This suggests that phenological coupling between these two species is subject to more intense selection at the range edges and that climate-induced changes in phenology likely have stronger fitness consequences in areas where the model species, Battus philenor, is relatively rare. Taken together, our results support the potential of citizen science data as a powerful resource for tracking historical spatiotemporal changes in highly vagile insect populations, particularly in tracking patterns linked to the long-term effects of climate change.