Life history traits, social mating system and genetic variation in the buff-banded rail Gallirallus philippensis on Heron island

Abstract

This thesis describes the social behaviour of the buff-banded rail Gallirallus philippensis during the breeding season. This is the first detailed study on a colour-banded population of the buff-banded rail, and one of the few studies on a socially monogamous rallid. The population on which this study is based is on Heron Island, a small coral cay, approximately 70km off the east coast of mainland Australia. Two thirds of the island contains areas of human habitation, and vegetation is made up of small gardens of native and exotic plants. The remainder of the island is a National Park, and is composed of native vegetation dominated by Argusia argentea, and Pisonia grandis. I documented 59 breeding attempts by 44 pairs of buff-banded rails over three breeding seasons. The broad aim of this study was to contribute to our understanding of the occurrence of social mating systems in the Rallidae, and to investigate the influence of ecology on the mating system of a species of rallid presumed to be socially monogamous. More specifically, this study aimed to describe the social behaviour of the buff-banded rail, including the social mating system, alternative reproductive strategies and behavioural paternity assurance mechanisms. I also investigated factors that may have favoured the occurrence of social monogamy in this species. The buff-banded rail was found to be primarily socially monogamous; all breeding groups were made up of pairs, and there was no evidence of social polygamy. Pair bonds continued outside the breeding season, and lasted for the lifetime of the individuals involved. No divorce was observed, and courtship feeding and/or allopreening may act as bond maintenance mechanisms in this species. The majority (70%) of females were single brooded, but 26% of females laid more than one clutch, and clutch size was highly variable (range 1-8), though second clutches were significantly larger than first clutches. The life history characteristics of the buff-banded rail are typical of other Rallidae, in that annual fecundity was high, and incubation (15-18 days), nestling (up to 24 hours) and chick dependence (4-6 weeks) periods were short. There was bi-parental care of the precocial young, as male and female putative parents incubated, brooded and fed young. Chick mortality was very high (72%-85%) in the first two weeks after hatching, and it is likely that bi-parental care is essential for the successful production of independent young. There is evidence to suggest that bi-parental care, along with limited food resources, may have favoured the evolution of social monogamy in the buff-banded rail. Although the buff-banded rail on Heron Island is primarily socially monogamous, extra pair copulations (EPCs) were observed, and 10% of copulations were extra pair. This is the first species within the Rallidae in which this behaviour has been reported. Both males and females initiated EPCs, though generally these copulations occurred on the male's territory. The role of EPC behaviour in this species is unknown, as most of the EPCs observed occurred when the participating female was not fertile. However, females may gain some type of direct benefit from participating in EPCs, such as food or nesting material, or the acquisition of mates in subsequent breeding seasons. I also found evidence to suggest that frequent copulation and/or mate-guarding may be used as paternity assurance tactics, in this population. Levels of genetic variation were investigated in the Heron Island population of buff-banded rails. Genetic variation, measured as mean number of alleles per locus was 2.87 in this population. This is very low, relative to levels of genetic variation found in other bird species. However, there was enough genetic variation to successfully conduct parentage assignment tests, as the parental exclusion probability for one, and two parents were 0.943 and 0.995, respectively. A comparative analysis was used to examine variation in life history traits in the Rallidae. Life history traits, body size and mating systems were found to be highly variable. I suggest that in this family, life history traits vary with body size, and mating systems are associated with variation in life history traits and diet, independent of the effects of body size.