Weather, topoclimate, and phenology: Long-term population dynamics of the threatened Bay checkerspot butterfly (Euphydryas editha bayensis)

Background/Question/Methods

Elucidating climatic drivers of population dynamics requires rigorous long-term estimates of distribution and abundance combined with knowledge of key demographic mechanisms.  Long-term population dynamics of E. editha bayensis in a large population complex in Santa Clara County, California, demonstrate how topoclimatic variability can buffer effects of weather.  Populations across 2000 ha of serpentine grasslands have been monitored by stratified sampling of postdiapause larval densities across local insolation gradients, a method that tracks overall numbers and microdistribution across slopes.  The longest record (1985-2013 at Kirby Canyon, 95 ha) is analyzed with multiple regression to identify key monthly weather variables, potential density-dependence caused by localized hostplant (Plantago erecta) defoliation, and relationships between population size changes and spatial shifts.  Shorter time series in adjacent habitats document overall synchrony and asynchrony within the population complex. Measures of phenological synchrony between the butterfly and its hostplants quantify the direct mechanism that drives population responses.  Long-term population persistence is modeled using an empirically calibrated diffusion approximation.

Results/Conclusions

The Kirby Canyon population fluctuated between 25,000 and 800,000 larvae, with three population booms and busts over 29 years.  Key weather variables were Oct-Nov precipitation (positive), Mar-Apr temperature (negative), and April precipitation (negative).  The first two factors affect the length of the grassland growing season, and heavy April precipitation can cause high adult and prediapause larval mortality. Two-year average population size had a negative impact above 400,000 larvae; these periods correspond to observed localized hostplant defoliation.  The spatial response – shifts across the insolation gradient – was correlated with population size change; years of growth exhibited shifts toward warmer slopes, and declines exhibited shifts toward cooler slopes.  Populations across the whole habitat largely tracked Kirby Canyon, but not in perfect synchrony.  The relative phenology of butterflies and hostplants strongly predicts population response, and phenological variation across slopes accounts for spatial shifts.  Mean times to extinction ranged from 300 to >4000 years under various formulations of the diffusion model.  The links between weather, topoclimate, phenology, and population dynamics provide key general insights into conservation in an era of climate change.