Abstract
New Zealand fur seals (Arctocephalus forsteri) are now increasing around the coasts of New Zealand after being extirpated from most of their former range through human exploitation. Fur seals breed on land and so their distribution will be limited by both terrestrial and marine factors. In this study, I examine the relationship between geographic and temporal variation in reproductive performance of the New Zealand fur seal and assess how this relates to variation in the marine and terrestrial environments. The main aim is to determine which factors control the rate and extent of colonisation by New Zealand fur seals. During three years (1996-1998), I captured, weighed and measured 6856 New Zealand fur seal pups from 20 breeding colonies around South Island, New Zealand, and estimated the annual pup production at each colony. I measured terrain characteristics at each site to assess the role of terrestrial habitat in site selection. I obtained data on prey distribution and abundance, profiles of the continental shelf, and coastal substrate. Pup condition was estimated from a morphological index combining mass and length and showed variation and spatial clustering that appeared to follow geographical variation in ocean productivity. Pup condition was also related to pup density in El Nino/Southern Oscillation (ENSO) years only. Terrestrial habitat type showed a relationship to pup density, suggesting that sites with the highest pup densities have terrain features that facilitate thermoregulation onshore. I estimated the probability of pup survival using an intensive mark-recapture technique at colonies on Otago Peninsula, South Island. Pup survival (after correcting for tag loss) was ≥0.918 and did not differ between the sexes. Other sources of mortality, such as predation by sea lions, do not appear to occur often enough to limit the population of fur seals. At a finer spatial scale (Otago Peninsula), I found that there was clustering of new colonies in relation to the distribution of established breeding colonies, and that the population was expanding at an average annual rate (r) of 0.30. This suggests that pup density, philopatry and site fidelity may also play important roles in the colonisation process. Finally, using artificial neural network (ANN) models I determined the relationship between environmental parameters (i.e., prey abundance, bathymetry, coastal substrate) and pup condition to predict the potential areas suitable for colonisation by New Zealand fur seals. Inference rules failed to describe consistent relationships between pup condition and oceanographic variables among years. Nonetheless, the model determined potential availability of coastline sites suitable for colonisation by fur seals. Ultimately, the rate and extent of colonisation by New Zealand fur seals appears to be controlled by the spatial and temporal configuration of the marine and terrestrial environment, in addition to the effects of population demography and individual behaviour operating at different spatial scales.