Modeling spring migration patterns of scoters and loons in the Bay of Fundy

Populations of scoter and loon species that winter in the Atlantic are understudied in North America, but coastal observatories may provide the data required to fill some of the knowledge gaps. The migration of scoters and loons has been monitored at the Point Lepreau Bird Observatory (PLBO) in the Bay of Fundy every spring since 1996, but little peer-reviewed research based on the resultant database has been published. Using data collected over 18 years at the Bay of Fundy (2000–2017), our objectives were to (1) determine the most accurate method of modeling hourly migration rates for Surf (Melanitta perspicillata), White-winged (M. deglandi), and Black (M. americana) scoters, and Red-throated (Gavia stellata) and Common (G. immer) loons, and (2) assess trends in hourly migration rates for our five focal species to determine if the numbers of migrants passing PLBO have changed over time. We calculated hourly migration rates for each of our five focal species and evaluated drivers (i.e., timing and environmental conditions) of migration and annual trends using zero-inflated generalized linear mixed models (GLMMs). We found that drivers of migration differed among species. Specifically, hourly migration rates decreased with increasing tide height for all species except Red-throated Loons. In addition, hourly migration rates increased with increasing wind vector (i.e., a tailwind) for the three scoter species, but decreased with increasing wind vector for the two loon species. Scoter migration rates peaked daily between 11:00 and 13:00 UTC, but we found no daily peak for either loon species. Peak hourly migration rates of Black and Surf scoters occurred from 12 to 26 April, but migration rates of White-winged Scoters and both loon species continued to increase throughout our migration-monitoring window. Finally, we found no changes in hourly migration rates over time for any of our focal species, suggesting no changes in abundance over the 18 years of data collection. Our study reveals the importance and utility of long-term, coastal observation stations, and we recommend their continued funding and use as valuable sources of monitoring data.