Aim:The distribution of marine predators is driven by the distribution and abundance of their prey; areas preferred by multiple marine predator species should therefore indicate areas of ecological significance. The Southern Ocean supports large populations of seabirds and marine mammals and is undergoing rapid environmental change.The management and conservation of these predators and their environment relies on understanding their distribution and its link with the biophysical environment, as the latter determines the distribution and abundance of prey. We addressed this issue using tracking data from 14 species of marine predators to identify important habitat.Location: Indian Ocean sector of the Southern Ocean. Methods:We used tracking data from 538 tag deployments made over a decade at the Subantarctic Prince Edward Islands. For each real track, we simulated a set of pseudo-tracks that allowed a presence-availability habitat modelling approach that estimates an animal's habitat preference. Using model ensembles of boosted regression trees and random forests, we modelled these tracks as a response to a set of 17 environmental variables. We combined the resulting species-specific models to evaluate areas of mean importance. | METHODSThe Prince Edward Islands (46.9°S, 37.7°E) are situated in the southwest Indian Ocean sector of the Southern Ocean (Figure 1). The Results: Real tracking locations covered 39.75 million km 2 , up to 7,813 km from the Prince Edward Islands. Areas of high mean importance were located broadly from the Subtropical Zone to the Polar Frontal Zone in summer and from the Subantarctic to Antarctic Zones in winter. Areas of high mean importance were best predicted by factors including wind speed, sea surface temperature, depth and current speed. Main conclusions:The models and predictions developed here identify important habitat of marine predators around the Prince Edward Islands and can support the largescale conservation and management of Subantarctic ecosystems and the marine predators they sustain. The results also form the basis of future efforts to predict the consequences of environmental change. K E Y W O R D Sareas of ecological

International audienceAfter the breeding season, penguins mustreplenish body condition and accumulate sufficient energystores before their annual moult ashore; failure to do somay lead to starvation. Knowing where and how adultpenguins find adequate resources during this energy-intensivestage is vital to understanding their susceptibility toecosystem changes. GPS and TDR loggers were used totrack movements and record diving behaviour of macaroniEudyptes chrysolophus and eastern rockhopper E. chrysocomefilholi penguins from Marion Island (46°S, 37°E)during the pre-moult foraging trip in 2012, 2013 and 2014.Both species consistently travelled in a southerly directionto forage in cooler (~3.5 °C) Antarctic Zone waters southof the Antarctic Polar Front where they associated withmesoscale eddies and sub-mesoscale filaments. Dives werepredominantly to depths of 30 to 60 m, but macaroni penguinsdived deeper more often. Mean trip durations of both species were similar (33 ± 6 days), but maximum foragingranges of macaroni penguins (903 ± 165 km) were greaterthan eastern rockhopper penguins (696 ± 152 km). Spatialoverlap of core foraging areas between species was high,but a 2- to 3-week difference in departure dates reducedpotential interspecific competition at sea. Trip durationswere longer in 2014 compared to 2013, when decreasedproductivity may have reduced prey availability, forcingpenguins to remain longer at sea. Continued monitoring isvital to understand how crested penguins at Marion Islandadapt to the predicted southward shift of major frontalboundaries

Twenty-two South African shores were surveyed for stranded debris colonized by goose barnacles. The most frequently colonized substrata were kelp (Ecklonia maxima) and plastic, comprising 33.1 and 29.0% of total colonized items, respectively. Four goose barnacle species showed significant preferences for specific substrata: Dosima fascicularis for feathers; Lepas anserifera for rubber sandals; L. pectinata for the shells of dead cephalopods Spirula spirula; and L. testudinata for kelp E. maxima. The other species, L. anatifera and L. australis, did not show significant preference for specific substrata. All species frequently attached to plastic debris. Object size and species were significantly related (P < 0.0001), with D. fascicularis and L. pectinata colonizing relatively smaller objects than other species. Colony size differed significantly between species (P < 0.0001), with D. fascicularis and L. pectinata occurring in relatively smaller colonies and L. testudinata being the only species to regularly occur in colonies ≥1000 individuals. We conclude that the nature and abundance of available substrata greatly influences the abundance and distribution of goose barnacles, and the increasing abundance of plastic is drastically increasing their opportunities to colonize and disperse within the southern African region.

Niche partitioning plays an important role in minimising interspecific competition for resources. Using carbon and nitrogen stable isotopic analysis of feathers, we investigated how macaroni penguins Eudyptes chrysolophus and eastern rockhopper penguins E. chrysocome filholi breeding at the Prince Edward Islands partition the marine environment during the critical pre-moult period over 5 consecutive years (2011 to 2015). Both species consistently foraged immediately south of the Antarctic Polar Front, with macaroni penguins foraging farther south in years of reduced primary productivity, minimising spatiotemporal overlap between species. Macaroni penguins consistently foraged at a higher trophic level than rockhopper penguins, indicating trophic niche differentiation. Male rockhopper penguins fed at a higher trophic level than females, but macaroni penguins showed no differences between sexes. The observed dietary differences, partial allochrony and species-specific spatial responses to reduced primary productivity (i.e. prey availability) indicate that macaroni and eastern rockhopper penguins exhibit contrasting foraging strategies that limit interspecific competition during the pre-moult period.

A central theme in community ecology is understanding how similar species co-exist and how their interactions may evolve in the context of climate change. Most studies of resource partitioning among central place foragers, particularly birds, focus on the offspring-rearing period, when they are accessible, but breeding success may be determined earlier and little is known about how such species partition resources at the onset of breeding. We used a non-invasive approach to evaluate resource partitioning in co-existing females at a sub-Antarctic island during their pre-laying periods. Three hypotheses were tested using carbon, nitrogen and oxygen stable isotope ratios measured in shells and membranes of hatched eggs as ecological tracers: 1) resource partitioning by geographic location and trophic level will exist among the 12 bird species and will be enhanced within taxonomic groups; 2) given the absence of strong oxygen gradients in the Southern Ocean we will not detect spatial structuring based on oxygen isotopes, but differences will exist between resident and oceanic species as the former may use meteoric water; 3) capital and income breeder strategies can be differentiated using stable isotopes of egg remains.Two and three dimensional isotopic data showed resource partitioning among species. As predicted, segregation was evident within the four main taxonomic groups: penguins, albatrosses, burrowing petrels and giant petrels. Unexpectedly, oxygen isotopes revealed widespread use of meteoric water among a suite of sub-Antarctic birds. Stable isotopes allowed us to identify females of most species as income breeders at the onset of breeding, with the exception of the females of the two crested penguin exhibiting a mix of income and capital resources use. Multidimensional isotopic analyses revealed that resource partitioning exists at multiple stages of the annual cycle in ways likely to be important under global change, exhibiting wide potential for ecosystem analysis.

Aim: Climate change will likely lead to a significant redistribution of biodiversity in marine ecosystems. We examine the potential redistribution of a community of marine predators by comparing current and future habitat distribution projections. We examine relative changes among species, indicative of potential future communitylevel changes and consider potential consequences of these changes for conservation and management.Location: Southern Indian Ocean. Methods:We used tracking data from 14 species (10 seabirds, 3 seals and 1 cetacean, totalling 538 tracks) to model the habitat selection of predators around the Prince Edward Islands. Using random forest classifiers, we modelled habitat selection as a response to a static environmental covariate and nine dynamic environmental | 143 REISINGER Et al.

Anthropogenic climate change is resulting in spatial redistributions of many species. We assessed the potential effects of climate change on an abundant and widely

Several records describe pseudostalked barnacles Xenobalanus globicipitis attached to killer whales Orcinusorca, yet little is known of this association in Southern African waters. Here we describe Xenobalanus prevalence on killer whales in South African waters and assemble previous records. Killer whales were photographed opportunistically between July and September 2013. Sex and age class were determined, and if Xenobalanus were present, attachment site and colony size noted. A prevalence of 50% was recorded, with barnacles most commonly observed on the dorsal fin and tail flukes. The high prevalence of Xenobalanus on killer whales in South African waters and lack thereof on killer whales at Marion Island implies that these are separate populations, and thus we suggest Xenobalanus prevalence is an additional means to distinguish between these populations.