Search for a command to run...
Context As advanced satellite-based environmental data become widely accessible, emerging opportunities exist to understand avian lifecycle events at continental scales. Although this growing toolbox offers much promise, an abundance of options may appear overwhelming to ecologists and point to the need for interdisciplinary collaborations to develop and interpret complex, spatio-temporal models. Aims Here, we demonstrate that satellite-based environmental variables complement conventional variables in spatio-temporal phenology models. The objective of this case study was to assess the degree to which including more sophisticated, satellite-based greenness data in association with a customised growing degree-day metric, can improve traditional phenological models based solely on monthly temperature and precipitation. Methods Using 2001–2018 purple martin (Progne subis) first arrival dates (n = 49 481) from the Purple Martin Conservation Association, we develop a predictive model for their first arrival dates on the basis of traditional temperature and precipitation values from ground-based meteorological stations, the MODIS satellite-based greenness index, and a more sophisticated growing degree-day metric. We used a Bayesian framework to construct 10 spatio-temporal candidate models on the basis of different combinations of predictor variables and our best model included a combination of both traditional and customised MODIS-based variables. Key results Our results indicated that purple martins arrive earlier when greening occurs earlier than the mean, which is also associated with warmer spring temperatures. In addition, wetter February months also predicted earlier martin arrivals. There was no directional change in purple martin first arrival dates from 2001 to 2018 in our study region. Conclusions Our results suggest that satellite-based environmental variables complement traditional variables such as mean monthly temperature and precipitation in models of purple martin migratory phenology. Implications Including emerging and conventional variables in spatio-temporal models allows complex migratory changes to be detected and interpreted at broad spatial scales, which is critical as Citizen Science efforts expand. Our results also pointed to the importance of assembling interdisciplinary research teams to assess the utility of novel data products.