Abstract
The effect of climate change on the phenology of plants and birds of eastern North America has been well studied in recent years, but insects have received less attention. In this study we investigated whether the response to climate warming of 10 short-lived butterfly species from the Lycaenidae family in Massachusetts is similar to responses seen in other taxonomic groups. We also determined the relative value of museum and citizen science data in ecological and conservation research, and how best to analyze these data. We obtained over 5000 records of butterflies in flight using museum collections (1893–1985) and citizen science data (1986–2009). We analyzed the data using linear regression models with sighting date as the response variable and temperature, precipitation, geographic location, and year as predictors. Temperature in the months during and prior to flight explained more variation in sighting date than the other predictors, with the average advance of flight date being 3.6 days/°C. Statistical tests using the first 20% of observations of flight in a year explained much more variation than tests using all observations. The response of these butterfly species to temperature is similar to plant flowering and bee flight times and is significantly greater than bird arrival times, suggesting the possibility of trophic mismatches. Citizen science data were more abundant and useful than museum data for studying climate change effects on butterflies. Conservation biologists and ecologists will need to develop innovative statistical techniques to deal with the sampling issues associated with citizen science data.