Fitness and Implications of Reproductive Decisions for Black Brant Nesting on the Yukon-Kuskokwim Delta, Alaska

Cost of breeding is influenced by numerous decisions during reproduction. In black brant (Branta bernicla nigricans; hereafter brant), an arctic goose, I studied effects of choices individual female brant which influence numerous life history traits. For birds maintaining long-term monogamous relationships, mate loss might be expected to reduce fitness, either through reduced survival or reduced future reproductive investment. In the second chapter, human harvest of male brant was used as an experiment to examine effects of mate loss on fitness of female brant. Nineteen years of recapture/resighting and nesting data were used from 2338 marked pairs of brant from the Tutakoke River colony in southwest Alaska. The Barker model in program MARK was used to examine effects of mate loss on annual survival, reporting rate, and permanent emigration. The best supported CMR model allowed survival to differ between females that lost their mates versus those that did not lose mates, reporting rate and fidelity to remain constant, and recapture probabilities at and away from the study area to be year specific. Survival rates decreased from 0.847 ± 0.004 for birds that did not lose their mates to 0.690 ± 0.072 for birds that lost mates. Ring reporting rate for females that lost their mates were two times higher than those that did not lose mates, 0.12 ± 0.086 and 0.06 ± 0.006 respectively, indicating that mate loss increased vulnerability to harvest and possibly other forms of predation. Little support was found for effects of mate loss on fidelity. Linear mixed models were used to examine variation in clutch size and nest initiation date within individuals relative to treatment (mate loss), number of years delay in breeding following treatment and number of years since resuming breeding. Females that formed new pair bonds and bred following loss of a mate enhanced their clutch size by approximately 0.83 eggs (compared to pre-loss size) when compared to the colony wide mean clutch size for that year and other pairs where the male was not removed. Relative nest initiation dates were approximately 0.45 days earlier for individuals that lost their mates and formed new pair bonds when compared to colony wide mean nest initiation dates. Results indicate substantial fitness costs to females associated with mate loss, but that females which survived and were able to form new pair bonds may have been higher quality than the average female in the population. Ideal free distribution theory predicts that individuals distribute themselves spatially to produce equal mean fitness across patches. Individual female brant at the Tutakoke River, Alaska, brant colony typically return to the same brood rearing areas across years (Lindberg and Sedinger 1998). Over 21 years, this study found that specific foraging areas consistently produced high quality goslings. Thus, some females consistently reared their broods on areas that produced offspring with low fitness. The fact that adult brant apparently did not respond to variation in fitness of their offspring among brood-rearing areas suggests three hypotheses: (1) broods were arranged such that adult brant that were socially dominant regularly occupied brood rearing areas producing goslings with greater fitness; (2) adult brant had insufficient information to accurately assess the quality of brood rearing areas; or (3) variation in offspring fitness was counter balanced against other components of adult fitness such that adult fitness was equal among brood rearing areas. In this dissertation, I evaluated all components of adult fitness to assess the hypothesis that individuals distribute themselves among seven distinct brood rearing areas in such a way that trade-offs among different life history traits result in equal fitness among areas.When examining variation in fitness across different habitat patches used during brood rearing, a consistent ranking of gosling mass corrected for age was found across brood rearing areas and years (Aikaike model weights, Σ wi = 1.00 for models including additive effects of brood rearing area and year). I examined the relationship between brood rearing area rank and future breeding propensity of adult females using these areas. Growth of goslings is known to influence their future fitness. That is, areas where goslings grew most rapidly also produced goslings with the highest mean fitness. We used a multistate robust design approach to estimate the transition probability from a breeding state to a non-breeding (unobserved) state in relation to these measures of quality of brood rearing area. The best supported model allowed the transition from a breeding state to a non-breeding state to be positively related to gosling growth rates across brood rearing areas. Adult female brant that used brood rearing areas that produced the smallest goslings experienced a 0.002 ± 0.008 probability of transition to a non-breeding state. Conversely, individuals that used brood rearing areas that produced the largest goslings experienced a probability of transition to a non-breeding state of 0.48 ± 0.11. These results are consistent with trade-offs by individual brant between fitness of their offspring and their own reproductive value. The growth period is an important determinant of fitness later in life through its effects on first-year survival and future reproduction. Choices by adult females about where to rear their broods strongly affect growth rates in geese. I explored the potential that gosling growth rates (and associated fitness consequences) are traded off against other vital rates influencing fitness of either adult females or goslings. Growth of goslings primarily influences fitness after fledging, so one hypothesis is that survival before fledging, which is influenced by predation, is traded off against growth rates and post-fledging survival. I estimated pre-fledging and post-fledging survival for goslings reared on areas used by broods.I used marking with webtags and standard leg rings, recaptures, ring recoveries, and resightings on the breeding area, and analyzed these data with the Burnham and Barker models within program MARK to derive separate estimates of pre- and post-fledging survival for 18 cohorts (1987-2004) of black brant goslings across seven brood rearing areas. Estimates of pre-fledging survival probability varied from 0*00 ± 0*00 (mean ± SE) to 0*92 ± 0*05; and estimates of post-fledging survival probability varied from 0*00 ± 0*00 to 1*00 ± 0*04. Substantial variation existed both among areas and years. Pre- and post-fledging survival were positively correlated and exhibited a quadratic relationship (ßpre-fledging survival = 1*00 (± 0*24)x -0*83 (± 0*41)x2, where x = post-fledging survival). Therefore, I did not find a trade-off between pre- and post-fledging survival in brant goslings across brood rearing areas, suggesting that factors other than foraging conditions and predation on goslings must influence selection of brood rearing areas.I used a suite of estimates of vital rates from other studies of this population to estimate brood-rearing-area-specific per capita female recruitment rates (clutch size, apparent nest survival, pre-fledging survival, post-fledging survival, juvenile survival, and breeding probability). I estimated brood-rearing-area-specific adult female annual survival rates which varied among years and areas (range 0.46 ± 0.26 to 0.99 ± 0.001). I used the sum of brood-rearing-area-specific per capita recruitment and apparent adult survival to calculate year and brood rearing area specific estimates of lambda. I used the approach of Coulson et al. (2006) to provide estimates of individual (brood rearing area by year combinations) relative contributions to lambda (-0.0005 to 0.0004), a measure of fitness. Because I found no variation in λ among brood-rearing areas and years, adult female brant appear to distribute themselves in an ideal free manner, resulting in equal fitness among females using these areas. It appears that females trade-off among different vital rates during their lifetimes such that individual fitness across habitat patches is equal