Assessing long‐term dynamics of non‐breeding Brown Pelican (Pelecanus occidentalis) populations using Christmas Bird Count data

Abstract

Brown Pelican Pelecanus occidentalis populations are recovering in many parts of their geographical range in coastal areas of North and Central America following near‐decimation in the mid‐20th century. However, populations have exhibited varying recovery rates that could be rectified with improved understanding of influencing factors, including population stressors, particularly during the non‐breeding period. Brown Pelicans are central‐place foragers and sensitive to changes in surface fish distributions that may result from stochastic events such as the El Niño Southern Oscillation. We used 34 years of Audubon Christmas Bird Count (CBC) data (1984–2017) to evaluate long‐term density‐dependent and density‐independent population dynamics of Brown Pelican populations in coastal areas of Alabama, California, Louisiana and Mississippi. We used exploratory dynamic factor analysis across four population time series to model a linear combination of common trends from CBC data. We then developed a suite of state space models to assess effects of density‐dependent and density‐independent factors (e.g. Southern Oscillation Index) on Pelican populations dynamics over time. Generally, Brown Pelican populations increased substantially following recovery efforts in the mid‐1980s, but have stabilized and fluctuated over time, represented by three latent trends. Populations generally were subject to density‐dependent regulatory mechanisms, with only the Louisiana population exhibiting a significant negative association with the fall/winter Southern Oscillation Index from 2 years earlier. Improved understanding of underlying regulatory mechanisms driving or hindering Brown Pelican population recovery may help in targeting regional or sub‐regional conservation measures, particularly during lesser‐known periods of their annual cycle.