Search

While there is now strong evidence that many factors can shape dispersal, the mechanisms influencing connectivity patterns are species‐specific and remain largely unknown for many species with a high dispersal potential. The rock lobsters Jasus tristani and Jasus paulensis have a long pelagic larval duration (up to 20 months) and inhabit seamounts and islands in the southern Atlantic and Indian Oceans, respectively. We used a multidisciplinary approach to assess the genetic relationships between J. tristani and J. paulensis, investigate historic and contemporary gene flow, and inform fisheries management. Using 17,256 neutral single nucleotide polymorphisms we found low but significant genetic differentiation. We show that patterns of connectivity changed over time in accordance with climatic fluctuations. Historic migration estimates showed stronger connectivity from the Indian to the Atlantic Ocean (influenced by the Agulhas Leakage). In contrast, the individual‐based model coupled with contemporary migration estimates inferred from genetic data showed stronger inter‐ocean connectivity in the opposite direction from the Atlantic to the Indian Ocean driven by the Subtropical Front. We suggest that the J. tristani and J. paulensis historical distribution might have extended further north (when water temperatures were lower) resulting in larval dispersal between the ocean basis being more influenced by the Agulhas Leakage than the Subtropical Front. As water temperatures in the region increase in accordance with anthropogenic climate change, a southern shift in the distribution range of J. tristani and J. paulensis could further reduce larval transport from the Indian to the Atlantic Ocean, adding complexity to fisheries management., Using a multidisciplinary approach combining genetic data (SNPs) with individual‐based modelling, we show that patterns of connectivity between rock lobsters Jasus tristani and Jasus paulensis changed over time in accordance with climatic fluctuations. Historical connectivity between the ocean basis was more influenced by the Agulhas Leakage (from the Indian to the Atlantic Ocean) while contemporary connectivity is stronger in the opposite direction from the Atlantic to the Indian Ocean, driven by the Subtropical Front.

Investigating historical gene flow in species complexes can indicate how environmental and reproductive barriers shape genome divergence during speciation. The processes influencing species diversification under environmental change remain one of the central focal points of evolutionary biology, particularly for marine organisms with high dispersal potential. We investigated genome‐wide divergence, introgression patterns and inferred demographic history between species pairs of all six extant rock lobster species (Jasus spp.), which have a long larval duration of up to two years and have populated continental shelf and seamount habitats around the globe at approximately 40oS. Genetic differentiation patterns reflected geographic isolation and the environment (i.e. habitat structure). Eastern Pacific species (J. caveorum and J. frontalis) were geographically more distant and genetically more differentiated from the remaining four species. Species associated with continental shelf habitats shared a common ancestry, but are geographically distant from one another. Similarly, species associated with island/seamount habitats in the Atlantic and Indian Oceans shared a common ancestry, but are also geographically distant. Benthic temperature was the environmental variable that explained most of the genetic differentiation (FST), while controlling for the effects of geographic distance. Eastern Pacific species retained a signal of strict isolation following ancient migration, whereas species pairs from Australia and Africa, and seamounts in the Indian and Atlantic oceans, included events of introgression after secondary contact. Our results reveal important effects of habitat and demographic processes on the recent divergence of species within the genus Jasus, providing one of the first empirical studies of genome‐wide drivers of diversification that incorporates all extant species in a marine genus with long pelagic larval duration.

Investigating historical gene flow in species complexes can indicate how environmental and reproductive barriers shape genome divergence before speciation. The processes influencing species diversification under environmental change remain one of the central focal points of evolutionary biology, particularly for marine organisms with high dispersal potential. We investigated genome-wide divergence, introgression patterns and inferred demographic history between species pairs of all extant rock lobster species (Jasus spp.), a complex with long larval duration, that has populated continental shelf and seamount habitats around the globe at approximately 40°S. Genetic differentiation patterns revealed the effects of the environment and geographic isolation. Species associated with the same habitat structure (either continental shelf or seamount/island) shared a common ancestry, even though the habitats were not adjacent. Differences in benthic temperature explained a significant proportion (41.3%) of the genetic differentiation. The Eastern Pacific species pair of J. caveorum and J. frontalis retained a signal of strict isolation following ancient migration, whereas species pairs from Australia and Africa and seamounts in the Indian and Atlantic oceans included events of introgression after secondary contact. Parameters estimated for time in isolation and gene flow were congruent with genetic differentiation metrics suggesting that the observed differentiation patterns are the product of migration and genetic drift. Our results reveal important effects of habitat and demographic processes on the divergence of species within the genus Jasus providing the first empirical study of genome-wide drivers of diversification that incorporates all extant species in a marine genus with long pelagic larval duration.

Aim: Quantifying biological assemblages and their environment is a fundamental, yet statistically challenging task in conservation ecology. Here, we use a recently developed approach called Regions of Common Profile (RCP) to quantify and map the distribution of demersal fish assemblages in an ecologically significant region of the Southern Ocean to (1) gain ecological and management insights and (2) evaluate the utility of the new method for ecoregionalization. Location: Northern Kerguelen Plateau, Subantarctic Islands, Southern Ocean. Methods: The RCP approach is a multispecies, model‐based approach that can overcome many limitations of traditional distance‐based approaches. It simultaneously groups sites with a similar composition of species and describes the patterns of variation in assemblages using environmental data, allowing the prediction of assemblages across the study region. We apply RCP to a unique dataset of demersal fish occurrences across the northern Kerguelen Plateau to model and map the distribution of assemblages and examine the representativeness of the Heard Island and McDonald Island marine reserve. Results: We demonstrate that the RCP approach allows a direct and quantitative interpretation of the composition of assemblages as well as their environment. Further, the model reasonably predicts the occurrence of individual species across the plateau as well as the species composition of sites. We distinguish and map seven assemblages defined by depth, surface temperature and chlorophyll a. Shallow‐water assemblages contain a high proportion of endemic species, while deep‐water assemblages contain more cosmopolitan species. With the exception of one deep‐water assemblage, assemblages were well represented within the current Heard and McDonald Islands marine reserve. Main conclusions: The RCP is a valuable tool for classifying biological regions with a range of ecological and conservation management applications. Our results extend current ecological and biogeographic knowledge for the northern Kerguelen Plateau, and maps of the distribution of assemblages will be useful for ongoing spatial management.

Three species of skate, Bathyraja eatonii , B. irrasa and B. murrayi, are commonly taken as incidental by-catch in Patagonian toothfish ( Dissostichus eleginoides ) longline and trawl fisheries, and the mackerel icefish ( Champsocephalus gunnari ) trawl fishery on the Kerguelen Plateau (KP) in the southern Indian Ocean. Data from fishery observations for 1997–2014 shows that the three skates were widely distributed across the Kerguelen Plateau, showing different spatial distributions, linked mainly with depth. Off Heard Island and McDonald Islands (HIMI), in the southern part of the KP, B. eatonii and B. irrasa were most abundant to the north and northwest of Heard Island, out to the edge of the Australian Exclusive Economic Zone (EEZ), and were caught down to depths of 1790 m and 2059 m respectively. The smallest species, B. murrayi , occurred mainly in the shallower waters down to 550 m, and was most abundant to the north and northeast, close to Heard Island. Around Kerguelen Islands, in the northern part of the KP, skates were most abundant between the 500 m and 1000 m contours circling and extending from the islands. Catch rates were modelled using zero-inflated GAMs and GLMs. The catch rates of skates from the trawl fisheries in the Australian EEZ surrounding Heard Island and McDonald Islands have shown little evidence of depletion on the main trawl fishing grounds, although there is evidence of a decrease in the average total length of B. eatonii . The marine reserves and the conservation measures employed by the Commission for the Conservation of Antarctic Marine Living Resources in the HIMI fisheries, appear to provide effective protection for the skates, at least in the shallower waters where the trawl fisheries operate. B. irrasa taken in the deeper waters where longline fishing occurs have shown a slight decline in catch rate over the years of the HIMI fishery. Although all skates are returned to the water from this fishery, survival rates are unknown and careful monitoring should continue to assess the status of these stocks. There appears to be little change in the abundance of the skate species at Kerguelen in the time period. This study provides the first review of skate by-catch across both the HIMI and Kerguelen fisheries. Ongoing monitoring of species specific by-catch levels and further research to determine the important life history parameters of these species are required, particularly for B. irrasa which is taken in both trawl and longline fisheries.

A new cryptic species of spotted barracudina, Arctozenus australis sp. nov., is described from the Kerguelen Islands, in the Indian Ocean sector of the Southern Ocean. It differs from the only congener, Arctozenus risso (Bonaparte, 1840), in the reduction of pigments on body, a more slender body, and longer head, snout and jaws. A neotype is designated for Paralepis risso Bonaparte, 1840 and Paralepis borealis Krøyer in Gaimard (1847). Note on synonymy of Paralepis risso is provided. Observation of adults of Notolepis annulata, Magnosudis prionosa and Paralepis coregonoides found loss of teeth on jaws and gill arches, suggesting more species in the subfamily Paralepidinae may commonly possess this character in adults.

Marine mammal depredation on fisheries (animals removing fish caught on fishing gear) is a worldwide issue involving socio-economic and ecological consequences. Longline fisheries are the most impacted by odontocete (toothed whales) depredation. While technological means have provided limited efficacy in reducing depredation, this study examined the fishing practices influencing both the proportion of depredated longline sets and the amount of fish removed by whales. We used an 8-year dataset from the Patagonian toothfish (Dissostichus eleginoides) longline fisheries operating in Crozet and Kerguelen Economic Exclusive Zones (EEZs) (South Indian Ocean) and GLMMs to investigate sperm whale (Physeter macrocephalus) depredation. Sperm whale depredation occurred on 61% of 5260 sets in Crozet and 41% of 16,902 sets in Kerguelen, and resulted in minimum estimated toothfish losses of 702 tons and 2649 tons, respectively, in the two areas. The probability of depredation decreased in winter months, increased with depth fished and decreased when vessels travelled over distances of >60 km from fishing grounds with encountering depredation. These findings suggest the natural spatio-temporal distribution of sperm whales and their ability to follow vessels over limited ranges influence the number of captured fish removals. The amount of depredated toothfish decreased with the speed at which longline sets were hauled and increased with the soaking time of sets suggesting that whales may depredate sets during both hauling and soaking operations. Together, these observations indicate that rates of depredation may be influenced by the conditions of fishing operations and could therefore be employed to implement strategies of avoidance in all fisheries facing similar depredation impacts.

The emergence of longline fishing around the world has been concomitant with an increase in depredation-interactions by odontocete whales (removal of fish caught on hooks), resulting in substantial socio-economic and ecological impacts. The extent, trends and underlying mechanisms driving these interactions remain poorly known. Using long-term (2003–2017) datasets from seven major Patagonian toothfish (Dissostichus eleginoides) longline fisheries, this study assessed the levels and inter-annual trends of sperm whale (Physeter macrocephalus) and/or killer whale (Orcinus orca) interactions as proportions of fishing time (days) and fishing area (spatial cells). The role of fishing patterns in explaining between-fisheries variations of probabilities of odontocete interactions was investigated. While interaction levels remained globally stable since the early 2000s, they varied greatly between fisheries from 0 to >50% of the fishing days and area. Interaction probabilities were influenced by the seasonal concentration of fishing effort, size of fishing areas, density of vessels, their mobility and the depth at which they operated. The results suggest that between-fisheries variations of interaction probabilities are largely explained by the extent to which vessels provide whales with opportunities for interactions. Determining the natural distribution of whales will, therefore, allow fishers to implement better strategies of spatio-temporal avoidance of depredation.

International audience; Although a high number of extant beaked whale species (Cetacea, Odontoceti, Ziphiidae) live in the Southern Ocean and neighbouring areas, only little is known about the past occupation of the region by these highly specialized, deep diving and echolocating cetaceans. Recently, longline fishing activities along the seafloor at depths of 500-2000 m off the sub-antarctic Crozet and Kerguelen islands, Indian sector of Southern Ocean, resulted in the accessory “capture” of tens of ziphiid fossil cranial remains. Our description and comparison of the best-preserved and most diagnostic crania from this sample lead to the identification of more than eight species in at least seven genera: the hyperoodontines Africanacetus ceratopsis, Khoikhoicetus kergueleni n. sp., Hyperoodontinae indet. aff. Africanacetus, and Mesoplodon sp. aff. Mesoplodon layardii, the ziphiines Izikoziphius rossi and Ziphius sp., and the ziphiids indet. Nenga sp. aff. Nenga meganasalis and Xhosacetus hendeysi. Unsurprisingly, with at least four species in common (A. ceratopsis, Izikoziphius rossi, X. hendeysi, and Ziphius sp.), the assemblage displays high similarities with assemblages described from deep-sea deposits off South Africa, providing thus new data on the palaeogeographic distribution of several extinct species and indicating a roughly similar geochronological age for at least a part of the as semblages. The limited amount of data available points to a pre-Pliocene age for a large part of the Crozet-Kerguelen assemblage, suggesting a relatively early, Miocene colonization of the Southern Ocean by crown ziphiids. Contrastingly, 14C radiometric dating of two specimens of Mesoplodon sp. aff. Mesoplodon layardii yielded latest Pleistocene-earliest Holocene ages. These results reveal the presence either of an extinct species of Mesoplodon in the Southern Ocean only a few thousands years ago, or of an up-to-now unidentified extant species closely related to the strap-toothed whale M. layardii.; Bien qu’un grand nombre d’espèces de baleines à bec (Cetacea, Odontoceti, Ziphiidae) habitent de nos jours l’océan Austral et les zones avoisinantes, l’occupation passée de cette région par ces cétacés à sonar hautement spécialisés, se nourrissant à grandes profondeurs, reste très mal connue. Récemment, des activités de pêche à la palangre menées sur le fond de l’océan à des profondeurs de 500 à 2000 mètres au large des îles sub-antarctiques de Crozet et Kerguelen, dans le secteur indien de l’océan Austral, ont permis la « capture » accessoire de dizaines de restes crâniens fossiles de ziphiidés. La description et la comparaison des crânes les mieux préservés et les plus diagnostiques de cette collection a permis l’identification de plus de huit espèces dans au moins sept genres : les hyperoodontinés Africanacetus ceratopsis, Khoikhoicetus kergueleni n. sp., Hyperoodontinae indet. aff. Africanacetus, et Mesoplodon sp. aff. Mesoplodon layardii, les ziphiinés Izikoziphius rossi et Ziphius sp., et les ziphiidés indet. Nenga sp. aff. Nenga meganasalis et Xhosacetus hendeysi.Sans surprise, avec au moins quatre espèces en commun (A. ceratopsis, Izikoziphius rossi, X. hendeysi, et Ziphius sp.), cet assemblage montre les plus grandes similitudes avec les assemblages décrits des dépôts océaniques profonds au large de l’Afrique du Sud. Ces similitudes fournissent de nouvelles données sur la distribution paléogéographique de plusieurs espèces éteintes et indiquent des âges géochronologiques similaires pour au moins une partie de ces assemblages. La quantité limitée de données disponibles indique un âge pré-Pliocène pour une bonne partie de l’assemblage de Crozet-Kerguelen, ce qui suggère une colonisation relativement précoce de l’océan Austral par des ziphiidés du groupe-couronne. D’un autre côté, la datation radiométrique au 14C de deux spécimens de Mesoplodon sp. aff. Mesoplodon layardii a fourni des âges allant du Pléistocène terminal au tout début de l’Holocène. Ces résultats indiquent soit la présence d’une espèce éteinte de Mesoplodon dans l’océan Austral il y a seulement quelques milliers d’années, soit la présence d’une espèce moderne, non encore identifiée et proche parente de la baleine à bec de Layard Mesoplodon layardii.

International audience; Age and sex dependent spatial segregation has resulted in limited knowledge of the ecology and demography of sperm whale adult males feeding seasonally in high latitudes. This study focused on adult males interacting with the Patagonian tooth-fish (Dissostichus eleginoides) fishery operating off the Kerguelen and Crozet Archipela-gos. Demographic parameters were estimated using a 10-yr-long photo-identification data set paired with multistate closed robust design capture-mark-recapture models. The examination of a set of 29,078 photographs taken from fish-ing vessels during sperm whale depredation events resulted in identification of 295 individuals with nine visiting both study areas. Dispersal between both study regions was estimated to be 1% per year. The mean annual number of interacting sperm whales was estimated to n = 82 (95% CI 58–141) in Crozet and n = 106 (95% CI 76–174) in Kerguelen. Transient proportions were 13% in Crozet and 26% in Kerguelen. Corrected for transience, apparent survival estimates were 0.953 (95% CI 0.890–0.993) in Crozet, and 0.911 (95% CI 0.804–0.986) in Kerguelen. These survival and population size estimates are the first for depredating adult males in high latitudes, and can be used in evaluating the current conservation status of this historically harvested stock and to investigate depredation trends in 35 both Crozet and Kerguelen Islands.

Over the past five decades, marine mammal interactions with fisheries have become a major human-wildlife conflict globally. The emergence of longline fishing is concomitant with the development of depredation-type interactions i.e., marine mammals feeding on fish caught on hooks. The killer whale (Orcinus orca) is one of the species most involved in depredation on longline fisheries. The issue was first reported in high latitudes but, with increasing expansion of this fishing method, other fisheries have begun to experience interactions. The present study investigated killer whale interactions with two geographically isolated blue-eye trevalla (Hyperoglyphe antarctica) fisheries operating in temperate waters off Amsterdam/St. Paul Islands (Indian Ocean) and south-eastern Australia. These two fisheries differ in the fishing technique used (vertical vs. demersal longlines), effort, catch, fleet size and fishing area size. Using 7-year (2010–16) long fishing and observation datasets, this study estimated the levels of killer whale interactions and examined the influence of spatio-temporal and operational variables on the probability of vessels to experience interactions. Killer whales interactions occurred during 58.4% and 21.2% of all fishing days, and over 94% and 47.4% of the fishing area for both fisheries, respectively. In south-eastern Australia, the probability of occurrence of killer whale interactions during fishing days varied seasonally with a decrease in spring, increased with the daily fishing effort and decreased with the distance travelled by the vessel between fishing days. In Amsterdam/St. Paul, this probability was only influenced by latitude, with an increase in the southern part of the area. Together, these findings document two previously unreported cases of high killer whale depredation, and provide insights on ways to avoid the issue. The study also emphasizes the need to further examine the local characteristics of fisheries and the ecology of local depredating killer whale populations in as important drivers of depredation.

A combination of morphological and molecular techniques was used to confirm the existence of a second species of the monotypic centrolophid genus Tubbia. Adults of the seamount rudderfish, T. stewarti sp. nov., which reaches about 56 cm SL, is mesopelagic at depths of 525–1438 m in the temperate waters of the Southern Hemisphere. It has a confirmed dis-tribution off Australia and New Zealand where it occurs sympatrically with the wider ranging T. tasmanica Whitley. Like most other members of the group, juveniles live in the epipelagic zone where they have been taken at 30–50 m depth. The new species has a more robust head, more slender body, more flattened interorbit, longer jaws, denser head pores, rela-tively larger eyes and nostrils, narrower caudal peduncle and more vertebral centra than T. tasmanica, and also differs sub-tly in some morphometric ratios. A rediagnosis of T. tasmanica is also provided.

International audience; Individual heterogeneity in foraging behavior has been widely documented within predator populations. In highly social apex predators such as killer whales (Orcinus orca), specialization may occur at the matriline level. A small population of killer whales has been documented to occur around the Crozet Islands. These whales feed on a wide range of prey items including seals, penguins and large whales, as well as depredate the local Patagonian toothfish (Dissostichus eleginoides) longline fishery. The level of interactions with fisheries varies greatly between matrilines. Here, we present the results on the effects of such behavioral heterogeneity on the demographic trends of this killer whale population. We used photo-identification data from 1977 to 2011 in a mark–recapture framework to test the effect of varying levels of fisheries interactions on adult survival. We documented significant differences in survival between depredating and non-depredating whales, resulting in divergent intra-population demographic trends. These differences showed low survival, and thus a negative effect, for depredating whales when illegal fishing occurred (poachers used lethal methods to deter killer whales from depredating longlines). After illegal fishing stopped (2003–2011), the survival rates of depredating individuals exceeded the survival rates of non-depredating individuals, suggesting a positive influence of “artificial food provisioning”. This effect was further supported by a higher population growth rate for depredating whales. This study highlights the potential demographic costs and benefits that cetaceans face from depredating fisheries and addresses the demographic consequences of both intra-population feeding specialization and the influence of anthropogenic changes in resource availability.

International audience; Sympatric forms of ecologically distinctive killer whales (Orcinus orca) have beendocumented worldwide. This study focused on a new case of such sympatric occurrenceof the “Crozet” type and the recently described “type D” killer whales off theCrozet Islands. The two ecotypes are morphologically and genetically distinct, butthey both depredate the same local longline fishery. We used observational,photo-identification, and fishing data, collected between 2003 and 2015, to examinedifferences in their patterns of depredation. Of the 828 sets where ecotype could beconfirmed, type D killer whales interacted with 82 (11%) of the sets, including 9(1%) sets that were simultaneously depredated by both ecotypes. Associationsbetween the two types were never observed. Type D killer whales typically occurredin larger groups and both ecotypes preferentially depredated Patagonian toothfish(Dissostichus eleginoides). GLMM modeling revealed that the probability of type Ddepredation significantly increased throughout the study period, especially in deepwaters, and photo-identification data suggested that a subset of all individuals werehabituating to depredation. This study documents the partitioning of resourcesbetween two distinct ecotypes of killer whales and provides preliminary insight intothe feeding ecology of the rare type D killer whale.

The RSV Aurora Australis survey allowed the first comprehensive study of the demersal ichthyofaunal environment and of the diversity of the Dumont d’Urville Sea. We observed a high dominance of the Notothenioidei in both the number of species and in integrated abundances. The Nototheniidae was the most abundant family with 44.7% of the total integrated abundance, followed by Bathydraconidae (18.8%). Trematomus eulepidotus was the dominant species with 19.9% of the total individuals catch. Nevertheless, 43 of the 53 species caught could be considered as very rare. The Bathydraconidae was the most diversified family with 11 species caught. The highest integrated abundances of fish were found from 400 to 800 m. Well-structured species communities were observed, with high species richness from 570 to 681 m. The richest zones were located along the basins and along their upper-sides. Statistical analyses indicated large-scale spatial patterns in species composition, with clear differences in fish communities from the continental slopes, the basins and on the shelf. At a finer spatial scale, the current in the George V Basin and iceberg scouring on the banks and their sides tended to create locally heterogeneous small-scale habitats. We suggest that the glacial history and the structured habitats allowed successive colonisations of the seabed by demersal fish.

A primary objective of the ICEFISH 2004 cruise was to collect and study notothenioid fishes from remote localities in the Atlantic sector of the Southern Ocean. Nearly 1 month was devoted to bottom trawling for fishes on the shelf and upper slope (to 1,000 m) areas around Shag Rocks, South Georgia, South Sandwich Islands and Bouvetoya. The focus was on the latter two locations, because their faunas are more poorly known. Eight species were collected at Shag Rocks with Patagonotothen guntheri most abundant; 17 at South Georgia with Lepidonotothen nudifrons, L. larseni and Gobionotothen gibberifrons most abundant; 13 at the South Sandwich Islands with L. larseni, L. nudifrons and G. gibberifrons most abundant; and 11 at Bouvetoya with L. larseni, Macrourus holotrachys and L. squamifrons most abundant. Ten new locality records were established: Shag Rocks (1), South Georgia (1), South Sandwich Islands (5), South Sandwich Trench at 5,350 m (1) and Bouvetoya (2). Total known demersal fish diversity on the shelf and upper slope at Shag Rocks/South Georgia, South Sandwich Islands and Bouvetoya is 42, 31 and 17 species, respectively. To examine population structure in the four most abundant notothenioids at Bouvetoya (L. larseni, L. squamifrons, Notothenia coriiceps and Chaenocephalus aceratus), we examined the ND2 portion of mitochondrial DNA. Chaenocephalus aceratus, N. coriiceps and L. larseni exhibited no significant genetic differentiation in comparison with samples from localities in the Scotia Sea and the Antarctic Peninsula. However, L. squamifrons showed significant genetic differentiation between the South Shetlands and Bouvetoya populations (F ST = 0.189, P = 0.015). Thus, these data combined with previous studies of two other notothenioids suggest that five of the six notothenioid species at Bouvetoya are not genetically differentiated from other localities in the Atlantic sector of the Southern Ocean. The location of Bouvetoya within the Antarctic Circumpolar Current and the long (1–2 years) pelagic stages of the notothenioids at Bouvetoya may be at least partly responsible for this genetic homogeneity.

In this work scientific echosounder were used hull mounted onboard R/V Marion Dufresne II to scrutinize the vertical distribution of micronekton according to water masses characteristics along transect which were discriminated with spatial remote sensing data. Acoustic data were collected continuously at 38 kHz frequency during eighteen transits carried out in the South-West Indian Ocean (20–60°S, 50–80°E) between 2010 and 2014. The data set used includes scientific and opportunistic fishing surveys from Integrated Marine Observing System database. A structure in three main depth layers i.e., "surface", "intermediate" and "deep", has been found continuously all along the survey. Changes of micronektonic vertical structure were investigated along North South latitude: the surface layer acoustic density and thickness decrease by going southward. The intermediate layer is generally almost empty except between 30°S and 40°S. The deep layer acoustic density increases from North to South but its thickness does not change significantly. To assess the importance of these vertical changes, a spatially constrained clustering was applied on acoustic data and a positive correlation was established between vertical acoustic organization and oceanographic fronts' position. We conclude that spatial organization of micronekton is structured depending on water masses and confirm the interest of collecting acoustic data from fishery vessels to complete scientific surveys that are often restricted in time and space.

De ´partement des Milieux et Peuplements Aquatiques, Museum National d’Histoire Naturelle(MNHN), CP 26, 43 rue Cuvier, 75005 Paris, France*Corresponding author: tel: +33 05 49 09 96 04; fax: +33 05 49 09 65 26; e-mail: tixier@cebc.cnrs.frTixier,P.,Gasco,N.,Duhamel,G.,andGuinet,C.Habituationtoanacousticharassmentdevice(AHD)bykillerwhalesdepredatingdemersal longlines. – ICES Journal of Marine Science, doi: 10.1093/icesjms/fsu166.Received2 April 2014; revised 27 August 2014; accepted 3 September 2014.Acousticharassmentdevices(AHDs)havebeenincreasinglyimplementedinvariousfisheriesthatsuffersignificantlossescausedbyodontocetedepredation.However,theefficacyofAHDstodeterodontocetesfromfishinggearremainspoorlyinvestigated.TodeterminetheeffectivenessofAHDsindeterringdepredation,weexperimentallytestedahighamplitudedevice(195 dBre1 mPa6.5 kHz1 mfromthesource)fromaPatagoniantoothfishDissostichus elegenoideslonglineroperatingofftheCrozetIslands, whileitwassubjected toheavydepredation bykillerwhales Orcinusorca.Thisspeciesusuallydepredateslonglineswithina10-to300-mrangefromthevessel,astheyonlyhaveaccesstofishinggearduringhauling.WeexpectedthisdistancetoincreaseinresponsetotheacousticdisturbancecreatedbytheAHD.Thedistancesof29killerwhalesfromthevessel(n ¼ 1812 records) were collected during phases of AHD activation and phases during which the AHD wasturned off. Two multiexposed killerwhale social units fled over 700 m away from the vessel when first exposed to the AHD. However, they remained within a 10- to 300-m rangeanddepredatedlonglinesagainpastthethirdandseventhexposures,respectively,showinganinsignificantbehaviouralresponsetofurtheractiva-tionsoftheAHD.Whentestedthroughgeneralizedlinearmixedmodels,theeffectofAHDactivationwasonlysignificantwhenkillerwhaleswerefirstexposedtothedevice.However,theeffectdisappearedaftersuccessiveexposuressuggestingthatkillerwhalesbecamehabituatedtotheAHDandmaysustainpotentiallyharmfulhearingdisturbancetoaccesstheresourcemadeavailablebylongliners.Inadditiontoraisingsignificantcon-servation concerns, this rapid return of initial depredation behaviour strongly suggests that AHDs are ineffective at deterring depredating killerwhales, and thatfisheries should favourthe use of other mitigationtechniques when facing repeated depredation by this species.Keywords: acoustic harassmentdevice AHD, depredation, killer whales, longline fisheries.

This paper provides a synthesis of results obtained as part of a long-term collaborative study involving biologists, fishers, and resource managers—centring on the occurrence of killer whales in the Crozet Archipelago before and after the implementation of a demersal longline fishery for Patagonian toothfish. Depredation behaviour was reported as soon as the fishery was initiated, with dramatic effects on both the demographic trajectories of the killer whales and on the amount of fish lost by the fishers. Killer whales interacting with the fishery exhibited very high mortality rates when illegal fishing took place, while killer whales not interacting were unaffected. However, after illegal fishing ended, killer whales interacting with the fishery exhibited both higher fecundity and survival rates compared with killer whales not interacting. Since whales typically removed fish entirely from the hooks, an adapted methodology that did not rely on determining the number of damaged fish was developed to estimate depredation rates. In the Crozet EEZ over a 10-year period, 33.9% of the total amount of Patagonian toothfish caught, representing a total of 28 million €, was estimated to be lost due to the combined effects of killer whale and sperm whale depredation. In an effort to reduce depredation losses, modifications to fishing methods, such as changing the fishing season, changing fishing areas when exposed to depredation and changing longline length and hauling speed were successfully tested. Acoustic deterrent devices were ineffective in deterring killer whales from depredating longlines. Alternative fishing gears, such as fish pots, were also tested. However, while providing encouraging results regarding the suppression of depredation and seabird bycatch, fish pots were not efficient enough to sustain an economically viable fishery. In conclusion, we discuss how the findings of this comprehensive study can be used elsewhere in fisheries confronted with depredation.

Odontocete depredation on longlines involves socioeconomic and conservation issues with significant losses for fisheries and potential impacts on wild populations of depredating species. As technical solutions to this conflict are limited and difficult to implement, this study aimed to identify fishing practices that could reduce odontocete depredation, with a focus on killer whales (Orcinus orca) interacting with Patagonian toothfish (Dissostichus eleginoides) longliners off the Crozet islands. Data collected by fishery observers from 6013 longline sets between 2003 and 2013 allowed us to statistically detect the significant influence of five operational variables using GLMMs. The probability of interactions between vessels and killer whales was decreased by (i) the number of vessels operating simultaneously in the area: the limited number of depredating killer whales may induce a dilution effect with increased fleet size, and (ii) depth of longline sets: vessels operating in shallow waters may be more accessible to whales that are initially distributed on peri-insular shelves. The cpue was negatively influenced by (iii) length of longlines: longer sets may provide killer whales access to a greater proportion of hooked fish per set, and positively influenced by (iv) hauling speed: increased speed may shorten the time during which toothfish are accessible to whales during hauling. The time it takes for killer whales to reach vessels was positively correlated to (v) the distance travelled between longline sets with an estimated threshold of 100 km beyond which whales seem to temporarily lose track of vessels. These findings provide insightful guidelines about what fishing strategy to adopt given these variables to reduce killer whale depredation here and in similar situations elsewhere. To a greater extent, this study is illustrative of how collaborative work with fishermen in a fully controlled fishery framework may lead to the definition of cost-limited and easy-to-implement mitigation solutions when facing such human-wildlife conflict.

Geometric morphometrics is a relatively new tool to fisheries research showing promise as a means of enabling researchers to cheaply and quickly categorise fish to individual stocks based on variations in otolith form, most commonly size and shape. In this study we introduce the method of elliptical Fourier analysis using two widely separated populations of striped trumpeter (Latris lineata) as a case study and compare the interpretation of results based on both unconstrained and constrained ordination techniques. There were no significant differences in otolith morphometrics between sex or age classes within each region. All form descriptors were standardised for fish length, thereby minimising confounding effects on any potential inter-regional otolith form differences. Non-metric multidimensional scaling was not sufficient to elucidate differences in otolith form between populations. However, using constrained canonical analysis of principal coordinates and canonical discriminant analysis, regional differences became evident with allocation success of 75 and 87%, respectively. Based on this study differences in otolith form reflect that the two tested striped trumpeter populations have reasonable phenotypic anonymity. This study further supports the usefulness of shape analysis and constrained non-parametric statistical tests as tools for stock discrimination and introduces elliptical Fourier analysis to the study of otolith morphometrics.

Otolith chemistry has been successfully used to reconstruct the environmental history experienced by estuarine-dependent teleost fish, including movement between estuaries and coastal areas. However, application has been more limited in species exposed exclusively to oceanic waters, where gradients in physical and chemical properties are less extreme. To test whether otolith elemental signatures record spatial information in an oceanic species, we sampled otoliths from Patagonian toothfish (Dissostichus eleginoides) and used an inductively coupled plasma mass spectrometer (ICP-MS) coupled to a laser ablation system to target the outer otolith edges corresponding to the period immediately before capture. Using multivariate analysis of variance and multivariate discriminant analysis, we found that edge signatures discriminated toothfish by geographic region with near complete success: only 5% of fish caught off South America and in the Antarctic were misclassified to sampling areas in the other region. Moreover, edge signatures showed strong differences between sampling areas within each region: fish captured off South America classified to sampling areas therein with 79%–84% success, and Antarctic fish classified to sampling areas therein with 50%–67% success. These results compare favourably with rates of classification for estuarine-dependent fish, demonstrating that otolith elemental signatures can discriminate the geographic provenance of oceanic and estuarine-dependent fish.

Soon after longlining on Patagonian toothfish (Dissostichus eleginoides) started in the Southern Ocean in the second half of the 1980s, interactions of cetaceans with these fisheries became apparent. The two species primarily involved were orcas (killer whales) (Orcinus orca) and male sperm whales (Physeter macrocephalus). Both species took substantial number of fish from the line primarily during day light hours. Catch rates of longliners declined to less than 50% when orcas occurred close to longline vessels while the loss to sperm whales was much less obvious. They were seen diving close to the line down to 400 m where they apparently took fish. Their impact on catch rates was much less notable. Sperm whales became frequently entangled in the line and part of the line was lost in a number of cases. Other cetaceans were rarely seen in the vicinity of longline vessels. They became entangled in the line only occasionally and one whale (presumably a minke whale) died.

14 pages; International audience; Using top predators as biological samplers, we collected information on the poorly known—but ecologically important—cephalopod fauna of the Southern Ocean. A total of 4527 cephalopod beaks were identified from stomach contents of Patagonian toothfish caught in slope waters at Crozet and Kerguelen Islands (Indian Ocean). Main prey were the squid Gonatus antarcticus and Kondakovia longimana at both localities, Taonius sp. B (Voss) and Slosarczykovia circumantarctica at Crozet, and Chiroteuthis veranyi and Mastigoteuthis psychrophila at Kerguelen. Fish diet together with the feeding habits of sharks and seabirds show that at least 36 and 38 different cephalopod species inhabit Crozet and Kerguelen waters, respectively. Oegopsid squid dominate the assemblages (29 and 32 taxa at Crozet and Kerguelen, respectively) over octopods (7 and 5 taxa), 1 species of sepiolid occurring at Kerguelen. These rich communities include pelagic squid, benthopelagic cirrate octopods and a few endemic benthic octopodids. The results emphasize the importance of onychoteuthids and gonatids in the nutrition of top consumers in the Southern Ocean and they shed new light on the role of chiroteuthids, mastigoteuthids and cirrate octopods in the trophic web of the marine ecosystems.

Only five species of sharks have been recorded in the Southern Ocean, where their biology is essentially unknown. We investigated the feeding habits of the three commonest species from stomach content analysis of specimens taken as bycatches of the fishery targeting the Patagonian toothfish (Dissostichus eleginoides) in upper slope waters of the Kerguelen Archipelago. The three species prey upon a diversity of fishes and cephalopods. They segregate by feeding on different species of squids of different sizes. The small lanternsharks (Etmopterus cf. granulosus; 0.3 m on average) feed on small-sized Mastigoteuthis psychrophila, while the large porbeagles (Lamna nasus; 1.9 m) feed on small-sized histioteuthids (Histioteuthis atlantica and H. eltaninae) and on medium-sized juvenile ommastrephids of the genus Todarodes. Finally, the huge sleeper sharks (Somniosus cf. microcephalus; 3.9 m) prey upon large-sized cephalopods (Kondakovia longimana and Taningia danae) and giant squids (Mesonychoteuthis hamiltoni and Architeuthis dux). Thus sleeper shark is a fish with sperm whale-like feeding habits and, hence, the second top predator known to science to rely significantly on giant squids. Prey species and biology indicate that porbeagles are pelagic predators in the entire water column, while sleeper sharks are mainly benthic top predators and scavengers. The present study also underlines the diversity and biomass of the poorly known cephalopod fauna, including giant squids, occurring in outer shelf and upper slope waters surrounding subantarctic islands.

The factors affecting the number and the mortality rates of seabirds attending long-liners and trawlers fishing in the Kerguelen area were studied during four successive seasons (1994–1997), based on observations carried out onboard by dedicated observers. Twenty-four species of seabirds were observed attending fishing vessels, representing an average of 591 birds/census. The total numbers attending varied mainly according to the year, the cloud cover and the presence of offal from long-liners. The dumping of offal increased the numbers of birds attending the vessel, especially when the offal could be easily handled by birds. The activity of the vessels also affected the numbers attending, birds being more abundant during line setting and trawl hauling. White-chinned petrels were the most abundant ship-following seabirds, followed by black-browed albatrosses, giant petrels and cape petrels. The number of white-chinned petrels, black-browed and grey-headed albatrosses attending fishing vessels increased in the time between spring and autumn, whereas it was the reverse situation for giant petrels and cape petrels. Four species of seabirds were caught by fishing gear, mainly by long-lines: white-chinned petrels, and black-browed, grey-headed and wandering albatrosses. Taking into account the number of birds from each species attending long-liners and known to be potential by-catch, some species appear to be more susceptible to being caught than others. White-chinned and grey-headed albatrosses are caught in much higher proportions than the numbers present, whereas black-browed albatrosses are caught in lower numbers. Giant petrels are abundant around long-liners but were never caught. In long-liners, most birds were killed when the lines were set during the day or when the deployment of the scaring device was not successful, with an overall figure of 0.47 birds/1000 hooks. Only one albatross was caught when the lines were set during the night. White-chinned petrels represented 92.2% of all birds killed by long-liners. The number of birds caught varied significantly among months and among years. The type of bait used also affected the catch rate. The catch rate was related to the number of birds attending the long-liner only for black-browed albatrosses. Most birds killed by trawlers were entangled by the netsonde cable. The efficiency of mitigation measures in order to reduce seabird mortality is discussed and it is stressed that night setting is the most efficient way to reduce mortality and should be enforced everywhere when possible. However, further methods should be developed to reduce the mortality of species active at night, especially white-chinned petrels whose populations in the Indian Ocean may by threatened by long-line fisheries.

Mesopelagic fish assemblages were investigated in the Polar Frontal Zone off the Kerguelen Islands during summer 1995, in parallel with a king penguin tracking study. During the day, the upper offshore water layers (0–200 m) have low potential prey diversity and abundance with only three fish species: a lanternfish, Krefftichthys anderssoni, a member of the Muraenolepididae, Muraenolepis marmoratus, and the early stages of the nototheniid, Lepidonotothen squamifrons. The mesopelagic fish community, including the typical myctophids, first appears in the deeper layer (300 m). At night, the surface layer (50 m) is invaded by the mesopelagic Myctophidae Electrona antarctica, Gymnoscopelus braueri, G. piabilis, G. fraseri, G. nicholsi, Protomyctophum bolini and P. tenisoni. Deeper (>100 m), a cline of species assemblages from the coast to offshore is observed.

A longline fishery for Dissostichus eleginoides has recently developed in the vicinity of South Georgia and Kerguelen islands, two internationally important breeding areas for procellariiform birds. The number of hooked birds and a method to reduce mortality were investigated during 13 days of fishing activity in Kerguelen waters in February 1994. Between 100 and 600 seabirds were always observed behind the longline vessel during daytime. The main ship-following species were the white-chinned petrel Procellaria aequinoctialis (67% of counts), giant petrels Macronectes spp. (8%) and the wandering albatross Diomedea exulans (11%), black-browed albatross D. melanophris (6%) and grey-headed albatross D. chrysostoma (2%). Only diving species were caught on the lines, i.e. the white-chinned petrel (n = 36) and the grey-headed albatross (n = 2). Marked differences in the mortality rate were observed between day and night (1·00 versus 0·38 birds per 1000 hooks), and at night when the decklights were on or off (0·59 versus 0·15 birds per 1000 hooks). Dumping of homogenized offal during line settings greatly reduced incidental capture of seabirds, mainly because birds were more attracted by offal than by hooked baits. We therefore propose that the dumping of offal during line settings should be included in the regulations of the longline fishery for Dissostichus in order to minimize seabird mortality.

Enzymatic polymorphism was used to detect variability within Notothenia rossii from two sites on the Kerguelen Plateau and Champsocephalus gunnari from the same sites and the South Orkney Islands. No polymorphism was found in the second species and it was low but not statistically significant in the first. This apparent homogeneity does not substantiate suggestions from other results that the populations can be separated, especially in the case of C. gunnari. Other approaches will be necessary to solve definitively the question of population separation.

In the winter 2007–2008, the CAML-CEAMARC cruises prospected in the Eastern part of the Antarctic continental shelf (Dumont d'Urville Sea, off Terre Adelie). The Australian R/V “Aurora Australis” and the Japanese R/V “Umitaka Maru” sampled in locations and at depths previously uninvestigated in this region. In total, 538 teleost specimens collected during these cruises were sequenced for the mitochondrial cytochrome oxidase I gene (COI), with the goal of barcoding a representative sampling from the campaign. The efficiency of barcoding for identification has been questioned for some taxonomic groups, thus we compared the COI results for a few of the families and genera included here (genus Trematomus , Artedidraconidae, Liparidae) to results for other markers for the same specimens. To better explore intra- and interspecific variability, sequences from previous campaigns and public databases were added to the analysis for these groups. The congruence among the results for different genes (COI, cytochrome b, D-loop and the nuclear rhodopsin retrogene) and morphological identification was used to assess the efficiency of the COI dataset at recovering species delimited using other data. Where discrepancies were present among the different data sources, a morphological re-identification was performed. The partial COI sequence yields reliable identification in most Antarctic teleost families when using their position in the clusters on a NJ tree. However, for several groups of species neither COI nor the other molecular markers investigated nor morphology recover unambiguously the currently accepted species. The taxonomy of these groups needs to be reconsidered. Identification through sequence similarity using the Barcode of Life Data System (BOLD) works for some groups, but is hampered by the incompleteness of the taxonomic coverage for antarctic teleosts. For four families (Artedidraconidae, Zoarcidae, Liparidae and Channichthyidae), several interspecific divergences were very small, and of the same magnitude as intraspecific divergences for other antarctic species. Despite these small divergences, almost all the species investigated in artedidraconids have molecular synapomorphies in the COI sequences, and a barcoding gap from the closest species. In the genus Trematomus , almost all species are well separated except for two pairs of closely related species that could not be distinguished by the other molecular markers either. For the typically hard to identify zoarcids and liparids, the results of barcoding are in agreement with in-depth morphological study. Once a reasonably complete reference dataset is available, barcoding will be invaluable to discriminate species from one another in these families. A careful comparison of the morphological and molecular results for our specimens allowed us to add numerous well-identified specimens (including some rare species) and sequences to BOLD. It helped to pinpoint the specimens that needed to be re-identified morphologically, and highlighted groups where barcoding is most helpful for specimen identification ( Chionodraco species). This large-scale project underlines the need for further taxonomic work in antarctic actinopterygians.

Mesopelagic fishes play a major ecological role in the Southern Ocean by partitioning energy throughout the water column during diel vertical migration. In the region, lanternfishes (Myctophidae) represent the most abundant mesopelagic fish family by species, number and biomass. There has been some intensive, localized sampling of myctophids in the Kerguelen Plateau region, but this has been geared mainly to the elucidation of larval fish biology, and to trophic studies of bird and mammal predators, and was focused in the upper 200-300 m depth stratum and mainly to the east of the Plateau. In this paper, a number of published and unpublished datasets are examined in order to clarify “regional boundaries”, and to establish the distributional characteristics and nuances of some representative convergence, subantarctic and antarctic species in the light of their proposed phylogeny. To improve this knowledge base going forward, the high costs of ship-board sampling could well be off-set by the use of modeling techniques.

International audience; Fatty acid (FA) and fatty alcohol (FAlc) compositions of both total lipid and neutral lipid fractions were studied for five myctophid species sampled in Kerguelen waters. Both qualitative and quantitative FA signature analyses were then performed to investigate their diet over longer time scales than the conventional stomach content analysis. Regarding their lipid class, FA and FAlc compositions, the five species could be discriminated into two groups: wax-ester-rich species (Electrona antarctica, Krefftichthys anderssoni) characterised by large amounts of monounsaturated FAs ([73% of total FAs) and triacylglycerol- rich species (Electrona carlsbergi, Gymnoscopelus nicholsi, Protomyctophum bolini) with major amounts of saturated and monounsaturated FAs ([29 and[46% of total FAs, respectively). Qualitative and quantitative FA analyses showed that K. anderssoni mainly preyed upon copepods, E. antarctica upon copepods and more euphausiids and P. bolini and E. carlsbergi mainly upon euphausiids with some copepods, while G. nicholsi had a more diverse diet. This study shows the usefulness of quantitative statistical analysis to determine the diet of Antarctic and sub-Antarctic predators and stresses the need of increasing the lipid and FA analyses of more zooplanktonic and micro-nektonic marine species.

International audience; Incidental seabird mortality associated with longline commercial fishing is a worldwide conservation concern. To develop conservation strategies, it is essential to estimate the likelihood of seabird bycatch and the degree of overlap between birds' foraging grounds and commercial fishing areas. We tracked 21 adult white-chinned petrels Procellaria aequinoctialis breeding on Kerguelen Island, southern Indian Ocean, during their breeding period in 2006 and in 2008. At-sea foraging distribution of white-chinned petrels was mainly confined to Antarctic waters. Commercial longline fisheries targeting toothfish were operating in both the French Exclusive Economic Zone and in other Commission for the Conservation of Antarctic Marine Living Resources areas during the study. We analysed concurrent data on the positions of both birds and vessels to estimate overlap. Static analysis using indices (home-range and utilization distribution overlap) revealed that, at a large scale, spatial and temporal overlap occurred, but varied among areas and with breeding stage. Dynamic analysis (detection of operating vessels respective to bird locations within a time/space window) revealed little overlap at a small scale. Our study revealed a mismatch between large- and small-scale overlap estimates, suggesting that birds and vessels occupy the same overall zone with infrequent co-occurrence (19% of birds in the vicinity of vessels). This result was confirmed by the relatively low occurrence of fishery-related items (4 to 22%) in chick food samples. However, given the large size of seabird populations, overall, large numbers of birds overlap with vessels, and management authorities should maintain and promote the implementation of strict mitigation measures to further reduce bycatch.

International audience; Ecoregions are defined in terms of community structure as a function of abiotic or even anthropogenic forcing. They are meso-scale structures defined as the potential habitat of a species or the predicted communities geographic extent. We assume that they can be more easily defined for long-lived species, such as benthos or neritic fish, in the marine environment. Uncertainties exist for the pelagic realm because of its higher variability, plus little is known about the meso-and bathypelagic zones. A changing environment and modification of habitats will probably drive new communities from plankton to fish or top predators. We need baseline studies, such as those of the Census of Antarctic Marine Life, and databases like SCAR-MarBIN as tools for integrating all of these observations. Our objective is to understand the biodiversity patterns in the Southern Ocean and how these might change through time. (C) 2010 Elsevier B.V. and NIPR. All rights reserved.

Since the article published by Loeb et al. [Loeb, V.J., Kellermann, A., Koubbi, P., North, A.W., White, M., 1993. Antarctic larval fish assemblages: a review. Bull. Mar. Sci. 53(2), 416–449.] about Antarctic ichthyoplankton, many surveys were carried out in different sectors of the Southern Ocean focusing on different aspects of the ecology of fish larvae. Some of these researches were conducted in the Subantarctic Kerguelen Islands and others on the continental shelf off Terre Adelie and Georges V land. Oceanographic and geographic features influence fish larvae ecology such as island mass effects, gyres, canyons. Antarctic fishes show also temporal segregation of spawning which induces temporal succession of early stage larvae. This avoids competition and probably the predation on early stages for species having few recruits. In that case, we have to understand how these larvae can deal with the match–mismatch with their preys and how they find sufficient food to survive. But our knowledge on Antarctic fish larvae is still insufficient as we do not know larvae for quite a lot of species and because of the difficulty to sample during winter.

Large-scale transport of seawater in ocean currents may generate spatially complex population structure through the advection of life stages of marine fish species. To test this, we compared the chemistry of otolith nuclei from Patagonian toothfish (Dissostichus eleginoides), presently managed as spatially discrete populations corresponding to fishing management areas along the Antarctic Circumpolar Current (ACC), which transports water eastward around the Southern Ocean. The chemistry of otolith nuclei, laid down during early life, differed significantly between fishing areas off South America and the Antarctic and between some Antarctic areas. However, we found significant discrepan- cies from expectation for a population structure corresponding to fishing areas. We also found evidence of four groups of fish with different early life chemistry: one associated with South America and three Antarctic groups showing mix- ing consistent with advective transport along the ACC. These results suggest that toothfish populations are structured by their physical environment; population abundance and persistence may rely on a restricted number of breeding members with access to spawning grounds, whereas fisheries may rely substantially on nonbreeding vagrants trans- ported from fishing areas upstream. Resume : Le transport a grande echelle de l'eau de mer par les courants oceaniques peut produire des structures de population complexes par advection des differents stades vitaux des especes de poissons marins. Pour tester cet enonce, nous avons compare la chimie des noyaux des otolithes chez la legine australe (Dissostichus eleginoides), une espece qui est actuellement geree comme etant formee de populations discontinues dans l'espace et correspondant aux unites de gestion de la peche le long du courant circumpolaire antarctique (ACC) qui transporte l'eau vers l'est dans l'Ocean austral. La chimie des noyaux des otolithes, determinee tot dans le cycle biologique, differe significativement dans les zones de peche au large de l'Amerique du Sud et dans celles de l'Antarctique, ainsi que dans certaines zones de l'Antarctique. Nous trouvons cependant d'importantes discordances par rapport a la structure de population a laquelle on pourrait s'attendre si celle-ci etait reliee aux zones de peche; il y a des indications de l'existence de quatre groupes de poissons avec une chimie differente au cours des premiers stades, un associe a l'Amerique du Sud et trois groupes antarctiques qui presentent des melanges compatibles avec le transport advectif le long de ACC. Ces resultats indiquent que les populations de legines sont structurees par leur environnement physique; l'abondance et la persistance des po- pulations doit peut-etre alors dependre d'un petit nombre de reproducteurs qui ont acces aux lieux de reproduction, alors que la peche depend de maniere importante de poissons errants non reproducteurs qui sont transportes dans le courant a partir des zones de peche.

International audience; The Kerguelen Plateau constitutes a natural obstacle for the eastward progress of the Antarctic Circumpolar Current, especially around the Kerguelen Islands. However, there is little quantitative knowledge of the current field around the islands due to lack of long-term current measurements. We performed a systematic analysis of a total of 28917 points of fishing gear drifts from setting and recovery positions of demersal longlines deployed between 2002 and 2007 for Patagonian Toothfish (Dissostichus eleginoides) fisheries in Kerguelen waters. This enabled the construction of a realistic field of depth-averaged time-mean slope currents along the 1000 m isobath all around the Kerguelen Islands. The strongest depth-mean velocities of 25 cm s−1 are associated with the Polar Front, which rounds the islands from the south and flows northward along the inner continental slope immediately east of the islands, strongly supporting previous hydrographic evidence. These results demonstrate the potential for hitherto unexploited historic longline drift data from demersal fishing grounds to provide valuable quantitative information on the regional circulation

Electrona antarctica is one of the most abundant mesopelagic fishes in the oceanic zone surrounding the Kerguelen Archipelago in the Indian sector of the Southern Ocean. Generalized additive models (GAM) combined with geographical information systems (GIS) were used to predict and map the abundance of this species according to three environmental variables: sea surface temperature, bathymetry and surface chlorophyll a. The model was applied on the Antarctic Polar Front in the eastern part of Kerguelen Archipelago. E. antarctica seems to be linked to areas presenting low chlorophyll a concentrations, depths greater than 500 m and temperatures lower than 5°C. The model was then applied to the Kerguelen’s plateau for three different years: 1998, 1999 and 2000. The position of Antarctic Polar Front and the intensity of an upwelling play an important role in the abundance variability of E. antarctica. Furthermore, the model allows the understanding of the habitat of E. antarctica and its trophic place in the pelagic ecosystem.

18 pages; International audience; The large temporal and spatial variability in marine productivity encountered by marine predators may negatively influence breeding success. The Antarctic fur seal Arctocephalus gazella (AFS), a marine predator in the Southern Ocean (SO) ecosystem with a circumpolar distribution, exhibits a short, 4 mo lactation coinciding with increased summer marine productivity. The diet of AFS, and the distance to significant and productive oceanographic features, such as the Antarctic Polar Frontal Zone (PFZ), varies considerably between populations. We studied the foraging activity, foraging efficiency and the pup provisioning strategies of lactating AFS at a key breeding site in the southern Indian Ocean, the Kerguelen Archipelago. Foraging parameters were examined in relation to interannual variability in oceanographic conditions and prey availability in the PFZ over 3 consecutive breeding seasons (1998 to 2000). The location of foraging zones, diving activity, diet and foraging efficiency varied significantly between years, concurrently with annual changes in seasurface temperature (SST) and prey availability. The strongest recorded El Niño Southern Oscillation event in 1997–1998 coincided with anomalously warm waters in the vicinity of the Archipelago. Deeper diving by females, reduced maternal and pup body condition, and minimal pup growth rates and low catch per unit effort of the primary prey species, lanternfishes (Myctophidae) were all recorded in this period. Maternal size was positively related to the growth performance of pups only in this period, indicating the importance of age/size and/or experience in mediating environmental fluctuations. Foraging efficiency over a foraging cycle and variability in mean provisioning rates (trip duration), were identified as proxies of prey availability within the foraging range of seals, emphasising the effectiveness of the use of AFS foraging behaviour as an indicator of both food and oceanographic variability and climatic anomalies. The increasing frequency of anomalously warm SST events in sectors of the SO, however, may elicit specific behavioural responses from ‘central place foragers' (i.e. species that return to breeding sites to feed their young) to avoid sustained poor body condition of females and their weaned offspring.

RésuméDes observations nouvelles et quelques échouages complètent le premier bilan faunistique établi en 1989 pour les îles Kerguelen. Aux neuf espèces identifiées précédemment ( Eubalaena glacialis , Balaenoptera acutorostrata , Megaptera novaeangliae , Physeter macrocephalus , Mesoplodon layardii , Cephalorhynchus commersonii , Lagenorhynchus cruciger , Orcinus orca , et Globicephala melas ) s'ajoutent, grâce à des échouages, deux nouvelles espèces: Ziphius cavirostris et Hyperoodon planifrons . En outre, la présence occasionnelle de Phocoena dioptrica est confirmée. L'événement le plus marquant est cependant le grand nombre d'observations (n=1649) de cachalots ( Physeter macrocephalus ) faites depuis 1989 au delà du plateau des îles, dans une zone peu fréquentée autrefois par les bateaux de pêche. Pour le dauphin de Commerson ( Cephalorhynchus commersonii ), espèce la plus commune, des données biologiques et écologiques s'ajoutent aux données faunistiques. Nous faisons également état, pour trois espèces ( Physeter macrocephalus , Orcinus orca , et Globicephala melas ), d'interactions avec la pêcherie à la palangre de la légine australe ( Dissostichus eleginoides ).

The diet of the onychoteuthid squid Moroteuthis ingens was investigated through stomach content analyses of 72 individuals collected aboard a trawler targeting Patagonian toothfish Dissostichus eleginoides in the upper slope waters of the Kerguelen Archipelago. M. ingens is pri- marily piscivorous (67% by number and 87% by reconstituted mass), although the diet also includes squids (12 and 12%, respectively) and crustaceans (21 and