|Spatio-Temporal Analysis for Bird Migration Phenology|
Ali Arab, Georgetown University
Monday, November 12, 2012
4:00pm, room 223 Weber Building
The study of changes in the patterns of bird arrival data is an important problem in phenology as the bird migratory processes can serve as potential bioindicators of climate change. To this end, several recent studies have analyzed trends of ecological processes such as bird migration patterns and provided evidence that pattern changes in these processes are correlated with climate effects. However, most studies ignore spatial and temporal variability in migration data and this often results in loss of important information across spatial and temporal scales. Critically, as the impact of climate change on natural processes may not be consistent over time and space, spatio-temporal analysis of bird migration processes allows climate scientists to better understand changes and shifts in migration patterns. We discuss a spatio-temporal modeling framework for analyzing birds spring arrival data which also allows for investigating potential links to climate indices. To demonstrate the methodology, we use two citizen science databases for Purple Martins; historic bird arrival data (1905-1940) from the North American Bird Phenology Program (BPP), and data for recent years (2001-2010) from the Purple Martin Conservation Association. Our results show significantly earlier spring arrivals for Purple Martins in most of the U.S. (South, Midwest, Mid-Atlantic, and North West) with link to Winter North Atlantic Oscillation. Finally, as most long-term phenological studies (including ours) heavily rely on data obtained from citizen science programs, we discuss potential sources of bias in these data sets and in general, sampling issues related to citizen science programs. Potential improvements will be discussed in order to obtain high quality data from these efforts.