There is widespread evidence that climate change is affecting the migration and reproductive phenology of birds. However, we do not understand within and between species responses heterogeneity and the scope work on this topic is primarily focused on insectivorous songbirds. Here, we used a comparative approach to examine how regional differences in climate change and migration distance might explain spring arrival dates in a raptorial species with a widespread distribution, the American kestrel (Falco sparverius). We used eBird data (www.ebird.org), a database of citizen science bird observations that provides long-term tracking records about abundance and spatial distribution of birds, such as where and when migratory birds arrive in the spring. Using eBird checklists from 2002-2018, we binned presence-absence data for grid cells (200 km²) across North America and calculated the daily proportion of checklists reporting kestrels for each year. We fitted a generalized additive model to each cell by year combination to estimate annual arrival dates indicated by the inflection point of the curve. Then, for each grid cell, we calculated mean daily minimum temperature anomalies for spring using Daymet climate data. We performed a generalized linear mixed model using arrival date as the response variable and flyway, latitude, and temperature anomaly as predictors.
Results/Conclusions
We found that kestrels arrived earlier in the central flyway (day of year 96.3 ± 17.2) compared to the eastern (101.1 ± 11.5) and western flyways (108.1 ± 14.5). An interactive effect between latitude and spring temperature revealed that at lower latitudes (short distance migrants) arrived earlier with warmer springs, but kestrels at higher latitudes (long distance migrant) arrived later with warmer springs. Our results suggest that kestrel arrival dates are temporally heterogeneous across North America and are consistent with the hypothesis that short distance migrants are better able to respond to shifts in phenology than long distance migrants. However, this hypothesis does not explain why long distance migrants are later in warmer springs. Long distance migrants will be more vulnerable to phenology mismatch because of rapid warming at high latitudes and later arrival times.