Butterfly numbers and weather: predicting historical trends in abundance and the future effects of climate change

TitleButterfly numbers and weather: predicting historical trends in abundance and the future effects of climate change
Publication TypeJournal Article
Year of Publication2001
AuthorsRoy, DB, Rothery, P, Moss, D, Pollard, E, Thomas, JA
JournalJournal of Animal Ecology
Date Published2001
ISBN Number1365-2656
KeywordsButterfly Monitoring Scheme, rainfall, regression models, temperature
  1. The effect of weather on the size of British butterfly populations was studied using national weather records and the Butterfly Monitoring Scheme (BMS), a national database that has measured butterfly abundance since 1976.
  2. Strong associations between weather and population fluctuations and trends were found in 28 of 31 species studied. The main positive associations were with warm summer (especially June) temperature during the current and previous year, low rainfall in the current year and high rainfall in the previous year. Most bivoltine species benefited from warm June weather in the current year, three spring species and two that overwinter as adults benefited from warm weather in the previous summer, and most species with moist or semi-shaded habitats increased following high rainfall and cooler weather in the previous year.
  3. Simple models incorporating weather variables and density effects were constructed for each species using the first 15 years’ population data (1976–90). These fitted the observed data for that period well (median R2 = 70%). Models were less good at predicting changes in abundance over the next 7 years (1991–97), although significant predictive success was obtained.
  4. Parameter values of models were then adjusted to incorporate the full 22-year data-run. For the eight species whose models had best predicted population changes or fitted the data well (R2 > 85%), models were run from 1767 to 1997, using historical weather records, to ‘predict’ trends in abundance over the past two centuries. For three species it was possible to compare predicted past trends with contemporary accounts of abundance since 1800. In each case, the match between predictions and these qualitative assessments was good.
  5. Models were also used to predict future changes in abundance, using three published scenarios for climate change. Most, but not all, species are predicted to increase in the UK under warmer climates, a few species stayed stable, and only one species – the agricultural pest Pieris brassicae (Cabbage White) – is predicted to decline.
Short TitleButterfly numbers and weather