Identifying demographic and environmental drivers of recruitment and population growth in a cavity nesting sea duck population

Lawson, Abigail; Sedinger, J. S.; Taylor, Eric

4 July 2017

Traits with the greatest proportional effects on fitness are typically conserved (Stearns 1992), and traits with larger temporal variation frequently play a dominant role in population dynamics (Cooch et al. 2001). We examined recruitment patterns and population growth in Common Goldeneyes (Bucephala clangula; hereafter goldeneye), using Pradel mark-recapture models from a long-term nest box study (1997-2010). Our objectives were to estimate recruitment (f) and population growth (λ) relative to recruitment origin group (in-situ or unknown), investigate environmental and density dependent effects on these parameters, and evaluate potential immigration patterns. We detected group-specific differences for f (in-situ: 0.47± 0.13 SE, unknown: 0.31 ± 0.04), and the proportion of boxes occupied by goldeneyes the year prior to recruitment had a significant negative effect on recruitment for the in-situ group (β = -1.04; 85% CI -1.29, -0.78), and a positive effect for the unknown group (β = 0.45; 85% CI 0.30, 0.61). The negative box occupancy effect in the year prior to recruitment, when in-situ yearling goldeneyes prospect for potential nest sites, suggests that local nesting densities may limit recruitment of locally hatched females. We identified two competitive models for λ, which averaged 1.04 ± 0.03 and included interactions between recruitment origin group and a linear temporal trend, and the proportion of ducklings marked two years prior. By evaluating all levels of marking effort on λ, we determined that even if all hatched ducklings were marked in a given year, the resulting in-situ λ was consistently lower than all observed population-level λs during the study, indicating that individuals produced outside of study area nest boxes contributed to λ. Though female goldeneyes are considered highly philopatric, our results suggest that female natal and breeding dispersal may be more prevalent than previously thought, and the spatial scale at which these processes occur requires further investigation.