The primary objective of the study was to determine why established females tolerate new females that join their breeding group and lay eggs in their nest. Previous work on this population has shown that females suffer a cost of joint-nesting in the form of lowered hatching success. Therefore, we would expect female pukeko to attempt to disrupt the reproductive efforts of their co-nesters by ejecting their eggs from the joint nest. Two hypotheses might explain why this does not happen. The "peace incentive" hypothesis states that females would forego egg destruction to avoid retaliatory behaviour by the other female. Alternatively, females might not destroy the eggs of co-nesters because they cannot discriminate between their own and another female's eggs. To test between these, we experimentally removed the eggs of one of the females from a number of joint nests. In all S(Wen cases for which we have data on the post-removal behaviour of the females, the robbed female showed no response to the disappearance of her eggs and continued to incubate the clutch. In addition, we added eggs to eight single female nests. Again, the single females showed no sign that they could distinguish between the foreign eggs and their own. The foreign eggs were not buried, ejected, or destroyed, nor were they moved preferentially to the outer perimeter of the clutch. To perform the egg removal experiments, I needed to correctly group joint clutches of eggs into maternal sib-groups. I evaluated two methods of doing this, one relying on qualitative observer assessment and the other on statistical techniques. I determined genetic maternity using DNA fingerprinting. Qualitative assessment was more effective than statistical techniques for identifying the maternity of eggs. Such an approach may be a useful alternative to expensive and time-consuming molecular genetic techniques for measuring reproductive skew in joint-nesting birds. Predation rates on pukeko nests at our study site during the 1998/99 nesting season were significantly higher than they had been in previous years (1990-1995). In the intervening years, the local rabbit population crashed as the result of two rabbit control measures: poisoning and rabbit haemorrhagic disease (RHD). We hypothesised that the increase in predation rates was due to rabbit specialist predators seeking out alternative prey after the crash in rabbit populations. Such a scenario is of grave concern to wildlife managers in many areas of New Zealand where rabbits are abundant and threatened native bird species are already under extreme pressure from introduced predators. / Thesis / Master of Science (MS)

Habitat saturation, due to high population density, and predation pressure have both demonstrated to influence the formation of some social systems. For example, helper-at-the-nest systems are formed when offspring delay dispersal due to a lack of suitable territory availability or high cost of dispersal. In one cooperative avian species, the pukeko (Porphyrio melanotus melanotus), a previous comparison between high density (North Island) and low density (South Island) populations found that at high density, there were fewer dispersal opportunities due to habitat saturation resulting in increased relatedness amongst group members. Climatic variations between the North Island and the South Island could explain many of the results previously found. The recent decimation of a large pukeko population at a North Island site provided us with a unique opportunity to compare group structure between two populations under similar climatic conditions to determine whether or not individuals bred independently in a low density population. For my thesis I aimed to answer three questions. First, how does population density affect the dispersal of offspring from their natal territory? Second, how does population density affect relatedness of adult group members? Third, how does density affect territory size? In this thesis I was unable to determine whether or not dispersal events were occurring or if there was high or low relatedness amongst group members. Territory sizes were significantly smaller at Tawharanui over Shakespear but they were not significantly correlated with group size. Harrier presence was significantly different between paddocks. / Thesis / Master of Science (MSc) / Ecological constraints have been shown to influence the formation of social groups. The decimation of a population of cooperatively breeding pukeko on the North Island of New Zealand provided us with a unique opportunity to compare two populations under different density conditions. I explored the effect of population density on group composition, parentage, and territory size as a means to understand the costs and benefits of cooperative breeding and joint-nesting. Territory sizes were significantly larger under low density conditions. I was unable to determine whether or not population density influenced offspring dispersal or relatedness amongst group members.

Group living is a widespread social system among animals. Within these groups, decisions on interactions between individuals can be facilitated through knowledge about individual identity and kinship. Individual identity allows for the recognition of individuals from past interactions and thus, information on likelihood of reciprocity and group-membership can be gained. The benefit for cooperative interactions, specifically, increases with the level of relatedness between the helper and the recipient. Thus, knowing who is kin, is an essential ability among group-living species and remembering individual identity helps to maintain long-term relationships and inform future decisions. Kin recognition can be facilitated through temporal and spatial overlap (i.e. familiarity) or through phenotypic-templates (i.e. phenotype matching). The goal of this thesis was to explore recognition in the joint-nesting pūkeko (Porphyrio melanotus melanotus). For the first portion of this thesis (Chapter II), I tested for evidence of phenotype matching in pūkeko using a cross-fostering experiment. Comparing survival and growth between fostered and non-fostered offspring, I provide evidence that pūkeko do not use phenotype matching as their mechanism for kin recognition. In Chapter III, I show that pūkeko chick distress calls may have an individual and group signature. I found variation in the vocal parameters between individual chicks and social groups. I also tested for response of adults towards chick distress calls of their own group. Using a playback-choice experiment, I report a biased response of adult pūkeko towards the distress call of their own group's chicks rather than the call of a distressed chick from a foreign chick. / Thesis / Master of Science (MSc) / Kin recognition is an essential ability for social species. Knowing whom is kin can help inform decisions on cooperation and conflict. I explored whether the joint-nesting pūkeko use familiarity or phenotype matching to recognise cross-fostered offspring. I experimented to determine if adult pūkeko can recognise the distress vocalizations of chicks in their group. I found no evidence that pūkeko use phenotypic templates to recognise cross-fostered chicks as non-kin. However, adult pūkeko showed a bias in response towards the distress calls of their own versus unfamiliar chicks. Individual chick distress calls, while variable from day-to-day, show group-specific similarities.