Incorporating Phenology to Estimate Species' Population Trends from Snapshot Citizen-Science Data

Citizen-science data are increasingly used to contribute to our understanding of biodiversity change, but analysing such data requires suitable statistical methods, often to deal with forms of bias. We develop a new approach for modelling data from a snapshot, mass-participation citizen-science scheme for UK butterflies, the Big Butterfly Count (BBC). Butterfly abundance varies throughout the year as one or more generations of each species emerge and die off, and the timing (phenology) of emergences varies annually due to weather and climate. Thus, counts from the short 3-week BBC sampling period are susceptible to bias due to this inter-annual variation in phenology. We adapt the Generalised Abundance Index, drawing upon phenology estimates from standardised monitoring scheme data, to account for phenological bias in the estimation of species' abundance trends from BBC data. The method is demonstrated via application to empirical and simulated data, revealing that not accounting for phenology leads to biased trend estimates, particularly for summer-flying single-generation species. Drawing upon phenology information, the new approach allows for the reporting of abundance trends from a snapshot citizen-science scheme, creating the potential to maximise available data sources to increase our understanding of changes in butterfly populations, particularly in urban environments.