| Summary: | Alaska sea duck populations have experienced substantial declines in the last half-century (Hodges et al. 1996). Though environmental conditions and anthropogenic activities have been implicated in some declines (Dickson and Gilchrist 2002), other potential causes remain unknown (Henny et al. 1995). In recent years, population trajectories among sea duck species (tribe: Mergini) have differed markedly (U.S. Fish and Wildlife Service [USFWS] 1999, 2012). In particular, eider (Somateria Polystitcta spp.) and scoter (Melanitta spp.) populations have declined to historically low levels, whereas populations of cavity nesting species (e.g., Bucephala, Mergus spp.) have generally increased or stabilized. Variation in population trajectories likely reflects a combination of proximate causes, such as differences in habitat quality (e.g., wetlands and tundra vs. forest), as well as variation in life history strategies, which influence how a population may respond to potential disturbance (Stearns 1992). My thesis research examined life history patterns and population dynamics of Common Goldeneyes (Bucephala clangula; hereafter goldeneye) in the northern boreal forest of interior Alaska. Goldeneyes are a cavity nesting sea duck with a holarctic distribution (Eadie et al. 1995). Between 1977 and 1994, goldeneye breeding populations in Alaska declined by 45% (Hodges et al. 1996), though current trends indicate a stable or increasing population in both Alaska (USFWS 1999) and continent-wide (USFWS 2012). I used capture-mark-recapture data from a long term (1997-2010) nest box study within the Chena River State Recreation Area, approximately 50 km. east of Fairbanks, Alaska. Goldeneyes are an ideal species for demographic study because they readily use artificial nest boxes and exhibit high rates of natal and breeding philopatry (Nilsson 1971, ii ii Dow and Fredga 1983, Savard and Eadie 1989, Pöysä et al. 1997, Ludwichowski et al. 2002), which allows for ease in capture and the ability to encounter the same individual for many years, often with a known breeding history. In chapter 1, my objective was to describe life history patterns of goldeneyes in relation to other sea duck species, which are typically described as long-lived, with delayed maturity, low and variable annual breeding probability, and invariant adult female survival. I used multistate capture-mark recapture models (Arnason 1972, 1973) in Program MARK (White and Burnham 1999) to estimate multiple demographic parameters, and used several explanatory variables to investigate the effects of climate on breeding and wintering areas, nesting density (as indexed by nest box occupancy), and individual variation (e.g., body condition) on my demographic parameters of interest. I detected substantial annual variation in adult survival (φA = 0.66 ± 0.10 SE to 0.82 ± 0.09). In contrast, breeding probability remained relatively high and invariant (ψ BB = 0.98 ± 0.13) and was positively related to individual nest success the year prior. Nonbreeding individuals in year t were more likely to remain a nonbreeder, than attempt to breed in year t+1. Probability of recruitment into the breeding population conditioned on survival to two-years of age was age invariant and followed a negative linear trend over time (ψ PB = 0.96 ± 0.01). In chapter 2, I used Pradel capture-mark-recapture models (Pradel 1996) to evaluate variation in per-capita recruitment (f) and population growth (λ), and determine the extent to which emigration and immigration may influence study area population dynamics. I detected significant differences in demographic patterns among two groups within my study population: (1) in-situ (IS) individuals that were marked as ducklings on the study iii iii area and later encountered as breeding adults; and (2) unknown recruitment origin (UN) individuals that were initially encountered as adults. In-situ per-capita recruitment was negatively related to the proportion of boxes occupied by goldeneyes in the year prior to recruitment ( = 0.52 ± 0.12), whereas recruitment in the unknown group was positively related to the proportion of boxes occupied ( = 0.30 ± 0.03). In the year prior to recruitment, yearlings typically prospect for potential nest sites (Eadie and Gauthier 1985), therefore, these results suggest that conditions during the prospecting year may facilitate density-dependent dispersal. Population-level λ varied substantially over time, and averaged 1.04 ± 0.03, the top two competitive models contained interactions between recruitment group and a linear temporal trend, and an effect of the proportion of ducklings marked two years prior. I investigated varying levels of marking effort on λ, and determined that even under maximum effort (100% hatched ducklings marked), λ s for the in-situ and unknown groups were significantly <1 (λIS = 0.69, λUN = 0.45), suggesting the population would decline in the absence of immigration. Furthermore, the non-zero value of λUN suggests that individuals produced outside nest boxes continue to contribute to λ, even when all potential in-situ recruits are marked. The demographic patterns I detected in goldeneyes are most consistent with a bethedging life history strategy (Sæther et al. 1996), rather than a survivor species strategy, as observed in other sea duck species (Goudie et al. 1994). Bet-hedging species are thought to persist in high-quality breeding habitat that enables annual breeding attempts, contrasted with survivor species that breed in lower quality habitat that constrains annual breeding opportunities (Sæther et al. 1996). As such, the importance of suitable nest sites iv iv to goldeneye demography is a recurring theme within my thesis research. High probability of breeding at the earliest age possible likely reflects both the necessity of obtaining a nest site, and the need to accumulate as many breeding attempts as possible over a lifetime, as expected of bet-hedging species— females that nest in the same nest site multiple years generally have earlier nest initiation dates and higher nest success than females that change nest sites (Dow and Fredga 1983). Furthermore, the negative relationship between in-situ per-capita recruitment and conspecific nesting density I detected, suggests that under certain conditions, a paucity of available nest sites may inhibit local recruitment patterns and mediate natal dispersal. The contrast in life history patterns in goldeneyes and other sea ducks may be a contributing factor to observed differences in current population trajectories, therefore, understanding variation in life history strategies within the sea duck group should be of fundamental interest to managers. Furthermore, I detected substantial evidence for both emigration and immigration in our study area, which suggests that nest box populations may be porous, despite assertions of perceived benefits to breeding of natal philopatry (Dow and Fredga 1983). Finally, a lack of demographic parameter estimates for age and sex classes throughout the goldeneye’s range has precluded the construction of a population model (Sea Duck Joint Venture 2008). The estimates presented here are among the first for goldeneyes within Alaska (Schmidt et al. 2006) and will contribute to a better understanding of goldeneyes throughout their North American range.
|
|---|