Foraging forms the cornerstone of an animal’s life-history. An individual's foraging success shapes the demography and health of a population. Understanding key facets of maternal foraging behaviour are crucial to get a holistic picture of both regional and local environmental factors that drive foraging behaviour. This study aimed to measure the maternal foraging behaviour of a marine top predator, the Subantarctic fur seal (Arctocephalus tropicalis), from Marion Island (MI) over a range of spatial and temporal scales. Arctocephalus tropicalis females from MI have one of the longest duration foraging trips for the species. They are most similar to conspecifics at temperate Amsterdam Island, but differ considerably from those at subantarctic Îles Crozet and Macquarie Island. Hitherto, no diving data existed for MI females. I illustrate how their diving behaviour is more similar to individuals from Îles Crozet despite their differences in foraging trip parameters. Together with Îles Crozet, MI females have one of the deepest mean diving depths (34.5 ± 2.2 m , 45.2 ± 4.8 m summer and winter respectively) and longest dive durations (70.2 ± 3 s , 104.3 ± 7.8 s summer and winter respectively) for the species. In summer, females follow the diel vertical migration of their myctophid prey. Counter intuitively, during the winter, females performed short and shallow crepuscular dives, possibly foraging on different prey. Considering that these individuals dive in deep waters, this is most likely related to myctophids occupying lower depths in the water column during winter. At dusk and dawn they are inaccessible to diving fur seals. At-sea data from multiple foraging trips per female illustrated that females have both a colony- and individual preferred foraging direction which varied seasonally. Individuals travelled consistently in the same direction regionally, but locally appear to track prey in a heterogeneous environment. The few trips in the winter to the west of MI suggest that this is a short-term response to varying prey availability rather than a long-term foraging tactic. Six years of observer-based attendance cycle data were used to augment telemetry data. Multi-state mark-recapture models were used to determine the probability of a female being missed when she was present (detection probability). Attendance data were corrected accordingly. Neither El Niño (EN) nor anomalous seasurface temperature (SSTa) influenced any of the attendance cycle parameters, as foraging trip duration is a poor predictor of weak environmental change. Only season and pup sex had a significant impact on female provisioning rates. Foraging trip duration was longer during winter than during summer. Females spent a higher percentage of time on land when they had female pups rather than male pups. Although observational attendance data remain useful it ideally requires concomitant data on pup growth, production and female body condition to elucidate changes in female provisioning rates. Temporally, season had the most influence on female foraging behaviour. Spatially, it appears that a lack of prominent local bathymetrical features overshadows MI's favourable position in the productive Polar Frontal Zone. Arctocephalus tropicalis females from MI work harder at foraging than at any other island population of conspecifics. / Dissertation (MSc)--University of Pretoria, 2013. / gm2014 / Zoology and Entomology / Unrestricted