Grizzled Skippers stuck in the south: Population‐level responses of an early‐successional specialist butterfly to climate across its UK range over 40 years

Abstract

AbstractAimClimate change has been predicted to facilitate poleward expansion of many early‐successional specialist invertebrates. The Grizzled Skipper, Pyrgus malvae, is a threatened butterfly in long‐term decline that has not met expectations of northern expansion in Britain, possibly indicating that climate change has not improved northern habitat suitability or that another driver (e.g. land use change) is masking its effects. Here, we explore the effect of climate on population size trends over four decades, and whether any regions show an improving population trend that may be a precursor to northern expansion. Examining detailed spatio‐temporal abundance data can reveal unexpected limitations to population growth that would not be detectable in widely used climate envelope models.LocationCentral and southern England.MethodsMixed models were used to investigate P. malvae population size in relation to time and monthly climate measures across its UK range since 1976, based on repeated transect walks.ResultsWe found that P. malvae population size declined more over time in the north and west of its UK range than in the south and east, and was negatively related to high December temperature and summer rainfall. However, the effect sizes of temperature and rainfall were minimal.Main ConclusionsThe last 40 years of climate change have not ameliorated climate suitability for P. malvae at its range edge, contrary to expectations from spatial‐only climate envelope models. The clear long‐term downward trends in population size are independent of climate change and we propose probably due to habitat deterioration. Our findings highlight potential hazards in predicting species range expansions from spatial models alone. Although some climate variables may be associated with a species’ distribution, other factors may be more dominant drivers of trends and therefore more useful predictors of range changes.