| Summary: | For long-lived species with perennial, socially monogamous pair bonds and bi-parental brood care, decisions concerning how much to invest in reproduction and whether to remain with a partner could substantially affect lifetime fitness. I studied the fitness consequences of reproductive effort (Chapter 2) and mate change (Chapters 3-5) in black brant geese (Branta bernicla nigricans, hereafter brant), an Arctic nesting goose species with precocial young. I used a 25 year (1990–2014) dataset collected at the Tutakoke River brant colony (TRC), in southwestern Alaska, USA to investigate these questions. Brant provide an ideal study species because researchers are able monitor large numbers of individuals and obtain unbiased estimates of fitness components throughout their life cycle. In my second chapter, I used experimental manipulations of clutch and broods sizes (i.e., the number of goslings leaving the nest) to understand if the maximal clutch size laid by brant (i.e., 5 eggs) is under ultimate control as a result of tradeoffs between reproduction and residual reproductive value of females. I used the Barker robust design mark-recapture model to estimate two components of female fitness: (1) true survival and (2) breeding probability in t+1. I found no evidence that incubated clutch size affects future fitness of female brant. However, breeding probability in t+1 (0.82 ± 0.10 [95% CI]) was maximized for females tending 4–5 goslings and declined for females with smaller or larger brood sizes. Thus, the brood sizes that maximized the residual reproductive value of adult females matched the most common clutch sizes laid by brant. The unexpected result that females tending smaller broods had lower future fitness may result from their smaller family size during winter which may decrease their social status in wintering flocks and result in reduced foraging profitability which may carryover to affect reproduction. My findings support the hypothesis that the maximal clutch size in brant is under ultimate regulation because of tradeoffs with adult residual reproductive value. In Chapter 3, I investigated whether or not mate change influenced reproductive success of female brant. I hypothesized that changing mates could affect reproductive success because (1) new pairs are unfamiliar with each other and (2) females may repair with a male that is an inexperienced breeder (i.e., fewer than two previous breeding attempts at TRC). I investigated the reproductive consequences of mate change using generalized linear models to estimate relative initiation date, clutch size, and the number of goslings leaving the nest and a Cormack-Jolly-Seber mark-recapture analysis to estimate prefledging survival. I found that females breeding with an unfamiliar, but otherwise experienced male fledged at least as many goslings as females breeding with an experienced, familiar mate. However, females who had switched to an inexperienced mate initiated their nests 0.48 (± 0.26 [95% CI]) days later and incubated clutches that were 0.17 (± 0.10 [95% CI]) eggs smaller than females breeding with a familiar mate. More importantly, goslings attended by a mother who was breeding with an experienced, familiar mate had greater prefledging survival (? = 0.30 ± 0.04 [95% CI]) than those whose mother had changed mates in year t and paired with an inexperienced male (? = 0.19 ± 0.04 [95% CI]). These results support the hypothesis that there can be reproductive costs of changing mates for female brant, but male experience rather than familiarity of partners determines these costs. In Chapter 4, I estimated rates of mate retention and investigated if there were long-term fitness benefits of mate retention. This analysis included 3021 and 3039 mature female and male brant who bred at TRC from 1990–2014. From 1990–2010, I recorded 748 and 196 breeding attempts after mate change for female and male brant, respectively. I estimated mate fidelity in t+1 of brant that were breeding with a familiar or unfamiliar mate in year t using a multi-strata robust design capture-mark-recapture analysis. I investigated whether breeding with a new mate reduced true survival or breeding probability in t+1 using the Barker robust design capture-mark-recapture framework. I found that mate retention in year t+1 for brant breeding with a familiar mate in year t was high for females (0.881 ± 0.017 [95% CI]) and males (0.952 ± 0.013 [95% CI]). However, for individuals who had switched mates in year t the probability of mate fidelity was greatly reduced for females (0.277 ± 0.163 [95% CI]) and males (0.343 ± 0.246 [95% CI]). There was also long-term fitness costs associated with mate change. Such that individuals who nested with a new mate had true survival rates which were lower (S females = 0.85 ± 0.009 [95% CI]; S males = 0.80 ± 0.017 [95% CI]) than those breeding with a familiar mate in year t (S females = 0.90 ± 0.006 [95% CI]; S males = 0.89 ± 0.011 [95% CI]). Additionally, individuals nesting with a new mate had a lower probability of breeding in year t+1 (females = 0.80 ± 0.035 [95% CI]; males = 0.44 ± 0.090 [95% CI]) than those breeding with a familiar mate in year t (females = 0.96 ± 0.008 [95% CI]; males = 0.98 ± 0.006 [95% CI]). I hypothesize that the demographic costs of mate change are partially related to the relatively low rates of mate retention among newly formed pairs. As a result, individuals breeding with a new mate are more likely to be single for part of the next winter which likely reduces their social status in wintering flocks and results in additional energy expenditure while they attempt to repair. To my knowledge, these results represent the first strong evidence of a link between breeding with a familiar mate and adult demographics in a long-lived bird, with perennial, socially monogamous pair bonds. In my fifth chapter, I investigated factors influencing mate retention and the proportion of female brant that pair with an inexperienced mate after mate change. I predicted that female brant that had undergone a mate change in year t and had paired with an inexperienced breeder or who failed to produce at least one gosling would have low rates of mate retention. I also suspected that brant would pair in a positive assortative manner by body size, previous breeding experience, and age because of the potential fitness benefits of breeding with larger, older and more experienced mates. Contrary to a priori predictions I found that females who had repaired with an inexperienced male were 39% more likely to retain their mate than those who repaired with an experienced male. I found no evidence that successfully producing at least one gosling influenced future mate retention. I recorded 273 cases where experienced female brant switched mates and the new mate was previously marked. In these cases the age of each mate was more strongly correlated (r = 0.26, P < 0.001) than the previous number of breeding attempts by each partner (r = 0.16, P = 0.007). There were weak, but statistically significant correlations between tarsus (r = 0.07, P = 0.012) and culmen (r = 0.07, P = 0.016) lengths of partners (i.e., proxies for structural size). However, body mass during brood-rearing was more strongly correlated among pair members than structural measurements (r = 0.28, P < 0.001). I estimated that about 90% of females who change mates acquire a male who is likely an inexperienced TRC breeder. It is unclear why females who pair with experienced partners have lower rates of mate retention, but it could result from experienced males being in short supply, thereby, enabling them to increase their choosiness in mates. It is clear that most females who change mates will suffer short term reductions in reproductive success, because they will likely acquire partners who are inexperienced breeders. When taken together, the results of chapters 3-5 suggest that there are short and long-term fitness benefits for brant that remained paired with an experienced, familiar partner. These benefits of mate retention may at least partially explain why brant have a perennial, socially monogamous mating system.
|
|---|