American kestrel population trends and vital rates at the continental scale

Abstract The American kestrel ( Falco sparverius , hereafter referred to as kestrel) has declined across much of its North American range since at least the mid‐1960s. Kestrel population dynamics have been explored through a multitude of local studies and two broad reviews of available data. Across large geographic extents, however, the demographic cause(s) of kestrel population declines remain(s) largely unknown. As part of a collaborative effort to elucidate the drivers of kestrel population declines, we developed a continental‐scale integrated population model using band‐recovery data, productivity data, and Breeding Bird Survey indices from 1986 to 2019 to estimate indices of annual population sizes, survival, and productivity rates across the continental United States. We detected a decline in population size of ~1%–2% per year. Overall estimates of population growth from 1986 to 2019 suggest a 29% decline in population size (95% CI = −34% to −23%). There was little evidence of a trend in brood size. However, survival of juvenile birds (mean = −0.015, SD = 0.008 and mean = −0.024, SD = 0.010 for females and males, respectively) and adult males (mean = −0.016, SD = 0.010) in the summer declined, suggesting that these vital rates could be contributing to declines in populations over time. Winter adult survival rates (mean = −0.004, SD = 0.009 and mean = −0.009, SD = 0.010 for females and males, respectively) also declined but to a lesser extent than summer survival. For juvenile birds, winter survival increased (mean = 0.006, SD = 0.008 and mean = 0.002, SD = 0.009 for females and males, respectively); however, this was not enough to offset declines in summer survival and annual survival rates declined over the time series. Annual adult survival was also low relative to previous research on kestrel survival rates. Given the importance of survival to population trends, our findings provide support for several previously proposed broad classes of factors potentially contributing to observed population declines: declines in arthropod prey, second‐generation rodenticides, neonicotinoid insecticides, and predation.