Predicting trends in humpback whale (Megaptera novaeangliae) abundance using citizen science

Abstract

The Great Whale Count (GWC) is an annual citizen science event that monitors changes in humpback whale (Megaptera novaeangliae) sightings in Maui County during the breeding season. The study includes 15 years of observations (1995–1996 and 1999–2011) with over 11 000 whale sightings. We provide a critical examination of the utility of the citizen science data given the challenges of observer-, site- and year-specific biases in counts, as well as an immeasurable and imperfect detection process. We estimate an annual increase of 5.16% per year (±2.76%), which closely resembles earlier trend estimates for Hawai’i. We demonstrate how uncertainty estimates in citizen science data can be strongly influenced by sampling processes, especially observer effects. Although such effects are now widely recognized in ecological studies, citizen science data often predate the mainstreaming of sampling protocols which measure and adjust for imperfect detectability. Here, we propose random effect models to minimize such effects in lieu of detectability techniques, and urge citizen science programs to adapt their protocols to handle observer processes at the planning and data collection stage.