Cephalopoda as prey of juvenile Southern elephant seals at Isla 25 de Mayo / King George , South Shetland Islands

The aim of the present study was to enhance the knowledge of the feeding habits of the juvenile component of the population of Southern elephant seals [Mirounga leonina (Linnaeus, 1758)] from Isla 25 de Mayo, South Shetland Islands, age class whose diet information is scarce. A total of 60 individuals were stomach lavaged in the spring summer seasons of three consecutive years (2003, 2004 and 2005) of which 53.3 % (n = 32) presented food remnants. The Antarctic glacial squid Psychroteuthis glacialis Thiele, 1921 was the dominant prey taxon in terms of frequency of occurrence (68.7%), numerical abundance (60.1%) and biomass (51.5%), contributing 84.1% to the total relative importance index. Other squid prey species of importance were Slosarczykovia circumantartica Lipinski, 2001 in terms of occurrence (37.5%) and numerical abundance (14%) and Moroteuthis knipovitchi Filippova, 1972 in terms of biomass (16%). All identified cephalopod prey taxa are distributed south of the Antarctic Polar Front, except for the squid Martialia hyadesi Rochebrune & Mabille, 1889 which has a circumpolar distribution associated to the Polar Frontal Zone. No significant differences in the sizes of P. glacialis preyed upon by elephant seals were found between sexes and years. However, significant interannual differences were found in the taxonomical composition of their diet. This would be associated with temporal changes in food availability at the foraging areas of seals, which in turn may have been influenced by changes in oceanographic conditions as a result of the El Niño Southern Oscillation (ENSO) phenomenon that occurred during part of the study period. Furthermore, a differential response of males and females to this temporal variation was observed, with the former being also associated to a predation on octopods. This would suggest a sexual segregation in foraging habits of this species from the early stages of its life cycle.

The Southern Ocean is a highly versatile and unpredictable environment, where the variability of oceanographic factors leads to a variation in the distribution, structure and abundance of biological communities.It is one of the most productive of the oceans in the world as a result of the brief and intense phytoplankton blooms occurring during Spring (Smetacek et al., 2005).Physical processes play an important role in determining the distribution and abundance of organisms in the ocean (Pakhomov et al., 1994;hunt et al., 1999;Pollard et al., 2002).These provide the physical structure in which phytoplankton and food webs will develop (mann et al., 1996).More important is the variability and heterogeneity of the physical processes at mesoscale level (lima et al., 2002), this being applied particularly to the Indian, Atlantic and Pacific sectors of the Southern Ocean subdivided into different regions well marked by fronts (BoSt et al., 2009).
Southern elephant seals are opportunistic generalist predators with a broad feeding niche, which annual cycle is a combination of two terrestrial phases (molt and breeding) and two large aquatic phases (post-molt and post-breeding) (le Boeuf & lawS, 1994).In these feeding phases, seals make long migrations.They can travel more than 5,000 km from their breeding and molting areas (mcconnell et al., 1992; hindell & mcmahon, 2000; hindell et al., 2003) and dive to great depths (1,500 m) (hindell et al., Iheringia, Série Zoologia, Porto Alegre, 105(1):12-19, 31 de março de 2015 1992).Consequently, they can access to a wide range of topographic and oceanographic conditions in the Southern Ocean.Reserves stored during their aquatic phase are essential to carry out their reproductive efforts on land (Boyd et al., 1994;knox, 1994;hindell et al., 2003;Biuw et al., 2007).Southern elephant seal breeding colonies are located mostly around the Antarctic Convergence (ling & Bryden, 1992;lawS, 1994;le Boeuf & lawS, 1994) with four major population cores: Macquarie, Kerguelen, Peninsula Valdes and South Georgia (Slade et al., 1998;hoelzel et al., 2001).The colony analysed in this study is located at Isla 25 de Mayo, South Shetland Islands and corresponds to the South Georgia stock, in the Atlantic sector of the Southern Ocean.The population at South Georgia has remained stable over the past 45 years (mccann, 1985;hindell et al., 1987;lawS et al., 1994;Boyd et al., 1996;Burton et al., 1997;guinet et al., 1999;SCAR, 2002SCAR, , 2006;;mcmahon et al., 2003).
The juvenile survival appears to be one of the determining factors in the demographic dynamics of M. leonina (hindell et al., 1994; mcmahon et al., 2003), and this is influenced by changes in morphology, behavior and hunting experience during ontogeny (field et al., 2007).Cephalopods (squid and octopus) constitute a common prey in the diet of the southern elephant seal (clarke et al., 1982;rodhouSe et al., 1992;daneri et al., 2000), although fish have also been reported as a relatively important food item, at least seasonally (daneri & carlini, 2002;BradShaw et al., 2003;newland et al., 2009).The southern elephant seal has a key role as a top predator in the food webs of the southern marine ecosystem.In the Scotia Sea it has been reported that this phocid species accounts for 75 % of the 3.7 x 10 6 tonnes of cephalopods and 45% of the 2 x 10 6 tons of fish predated by birds and seals in the area (croxall et al., 1985).In addition, further studies have shown that the contribution of different prey species to its diet may vary with locations, seasons and between sex/ age type (rodhouSe et al., 1992;green & Burton, 1993;SliP, 1995;daneri et al., 2000;PiatkowSki et al., 2002;BradShaw et al., 2003;van den hoff et al., 2003;field et al., 2005;among others).
For the I. 25 de Mayo population of M. leonina, studies have reported that juveniles (in their first year of life), females and adult males develop their foraging activity in different areas of the Southern Ocean (Bornemann et al., 2000;toSh et al., 2009).Additionally, dietary studies conducted to date on this population, have shown that its diet differs from those of other colonies since it is mainly represented by fish and squid species distributed almost entirely south of the Antarctic Polar Front (daneri et al., 2000, 2005;daneri & carlini, 2002;PiatkowSki et al., 2002).Nevertheless, it should be noted that dietary information of the juvenile component of M. leonina in the Atlantic Sector of the Southern Ocean is relatively scarce.
It is worth highlighting the total absence of fish remains in the stomach contents of juvenile seals sampled in this study.Therefore, the aims were to (1) examine the cephalopod component of juvenile individuals of M. leonina from I. 25 de Mayo colony; (2) assess temporal variations in their cephalopod diet during the study period and (3) analyze differences in cephalopod predation patterns between sexes.

MATERIALS AND METHODS
The study was carried out from November to December of three consecutive years (2003, 2004 and 2005) at Stranger Point (62°14'S, 58°40'W), Isla 25 de Mayo/ King George (Fig. 1).The sampling area corresponds to an Antarctic Specially Protected Area (ASPA No. 132), located next to "Carlini" Scientific Station (formerly known as Jubany).
Sixty one recently hauled out juvenile Southern elephant seals were immobilized by intramuscular injection of ketamine hydrochloride (n=15 in 2003; n= 19 in 2004 and n= 27 in 2005) and subjected to stomach lavage following daneri & carlini (1996).
In the field, stomach content samples were filtered in a sieve of 1 mm mesh size.Cephalopod remains, mainly represented by mandibles (beaks) and few eye lenses, were preserved in 70% ethanol for further laboratory analysis.Beaks were identified using appropriate guides (okutani & clarke, 1985; clarke, 1986; xavier & cherel, 2009) and by comparison with reference collections deposited in the Museo Argentino de Ciencias Naturales "Bernardino Rivadavia" (MACN) and the Instituto Antártico Argentino (IAA).
The Index of Relative Importance (IRI) was calculated according to PinkaS et al. (1971), but as a modified version, by replacing the volume percentage by weight percentage according to reid (1995): IRI = (% N +% W) *% FO Furthermore, and in order to make easier the interpretation of the IRI, this index was expressed on a percent basis (%IRI) following cortéS (1997).A three factor nested ANOVA design (year of sampling, seal sex, and specimen sampled) was performed to detect seasonal differences in the sizes of the dominant squid species (Psychroteuthis glacialis) preyed upon by seals.The variable "specimen" was nested within year and sex factors.The model assumptions were corroborated.Finally, to determine whether or not there existed significant differences between years and sexes in the taxonomical composition of the diet, a log-linear analysis was applied.

RESULTS
A total of 356 beaks were extracted from stomach contents (194 lower and 163 upper).Most of these were in a fresh state.Only one of the lower beaks had to be discarded because of its degradation level.The average number of lower beaks found per stomach was 6 (range 1 -36).
In the total period of the study, 10 taxa were identified (7 teuthoids and 3 octopods) (Tab.I).All teuthoid species together accounted for over 99% of the overall IRI.The predominant prey taxon was the Antarctic glacial squid, Psychroteuthis glacialis, which occurred in 68.8% of stomach samples and represented respectively 60.1% and 51.5% in terms of numbers and mass of the total cephalopods consumed.Moreover, this squid taxon had the highest %IRI (84.1) followed by Slosarczykovia circumantartica (6.2%), both taxa thus accounting for over 90% of the percentage contribution to the overall IRI.While the predominance of P. glacialis in terms of IRI% was constant throughout the 3 years of sampling, a substantial decrease was observed in 2004.This year coincided with a stronger contribution of Moroteuthis knipovitchi to the diet of seals, reaching a %IRI value of 29, almost equaling P. glacialis (Fig. 2).
The beak size (LRL or LHL), mean estimated dorsal mantle length (ML) and mean estimated mass of the cephalopod species taken by elephant seals are shown in Table II.The log-linear analysis indicated significant differences among years in the cephalopod diet composition (p<0.05) with sexes also predating differentially according to this temporal variation (p<0.05).While octopods occurred in the diet of some male individuals, females preyed exclusively on squid taxa (Fig. 3).
Tab. II.Beak size (LRL/LHL, mm) estimated mantle length (ML, mm), and estimated mass (M, g) of cephalopods preyed on by juvenile southernelephant seals at Stranger Point, Isla 25 de Mayo/King George Island (n = the number of lower beaks found of each taxon).
Elephant seals have a quick digestion of 13 hours approximately according to krockenBerger & Bryden (1994).However, in several phocid species, the hard parts of prey can be retained in the stomach for at least 7 days (harvey & antoneliS, 1994; tollit et al., 1997).Moreover, the swimming speed of juveniles of M. leonina can reach up to 100 km per day (field et al., 2005).Hence, we assume that, at the time of sampling, food remnants that were extracted from the stomachs could well refl ect prey captured at a distance of about 700 km from Isla 25 de Mayo.
If we consider that: (a) The juvenile seals sampled in the present study belong to the southernmost breeding population of M. leonina; (b) studies of movements at sea by satellite tellemetry of juvenile seals from this colony had shown that most of them travelled to the southwest of the South Shetland islands and centred their foraging activities in ice free areas west of the Antarctic peninsula (Bornemann et al., 2000;toSh et al., 2009).Then, the almost exclusively Antarctic distribution of their cephalopod prey taxa is probably a refl ection of their foraging areas and not a bias from the sampling technique.
Notwithstanding, in 2003, two lower beaks of Martialia hyadesi were removed from the stomach contents of two seals (one male and one female).This species has a circum-Antarctic distribution associated with the Polar Frontal Zone The Antarctic Peninsula is one the most rapidly warming regions on earth, having experienced a 2°C increase in the annual mean temperature and a 6°C rise in the mean winter temperature since 1950 (ducklow et al., 2007).Moreover, delivery of heat from the Antarctic Circumpolar Current has increased signifi cantly in the last decades, suffi cient to drive to a 0.6°C warming of the upper 300 m of shelf water (Stammerjohn et al., 2008).Thus, it should be considered that the squid M. hyadesi might be (if its occurrence in the diet of M. leonina increases) expanding its spatial niche to higher latitudes.Another explanation might be associated to more northerly feeding grounds of juveniles which may reach the Antarctic Polar Front or even further north.In this regard Bornemann et al. (2000) reported one juvenile (underyearling) elephant seal from Isla 25 de Mayo which traversed the Drake Passage and almost reached the patagonian shelf when the satellite transmitter failed.
In the present study, the squid P. glacialis proved to be the most important prey in terms of frequency of occurrence, numerical abundance, biomass and relative importance index (Tab.I).These results are in line with previous information on food habits of M. leonina from Isla 25 de Mayo/King George (daneri et al., 2000, 2005).Signifi cant differences were found between years in the cephalopod diet composition of seals.Moreover, a differential response of female and male individuals to this temporal variation was observed.While satellite telemetry studies have not indicated sexual segregation in the foraging areas of juvenile males and females of this colony (Bornemann et al., 2000(Bornemann et al., ), daneri et al. (2000(Bornemann et al., , 2014) ) reported gender differences in the cephalopod diet of the adult component of the population, with octopod prey becoming more important in males.
Iheringia, Série Zoologia, Porto Alegre, 105(1):12-19, 31 de março de 2015 The interannual differences observed in the composition of the diet of juveniles of M. leonina could be linked to temporal changes in the availability of prey, mainly krill (Euphausia superba) in the feeding areas.
Euphausiids are a common food item of some of the squid prey species identified in this study, mainly P. glacialis (collinS & rodhouSe, 2006).In this regard, hewitt et al. (2003) estimated intra and interannual variations in the density and spread of krill biomass, E. superba, inferred from acoustic surveys on areas near the South Shetland between the summers of 1991/92 to 2001/02.These authors found that in the 1991/92 and 1997/98 seasons (coinciding with the Southern Oscillation event of "El Niño") the recruitment of krill from laying eggs was very low and the krill biomass density declined sharply over these two periods as in subsequent years.It should also be noted that PiatkowSki et al. (2002), analyzing the temporal variation in the cephalopod diet of adult females of M. leonina from this same colony, indicated that the contribution of the squid P. glacialis to their diet, was proportionately less during events of "El Niño".Moreover, they suggested an increase availability of this squid species to females on the Bellingshausen Sea area concomitantly with a decrease in the temperature of the sea surface associated with "La Niña" events.All these findings fit well to the results of this study since the marked decrease in the contribution of P. glacialis to the diet of juveniles of M. leonina was observed in conjunction with a warm episode of El Niño Southern oscillation which occurred in 2004 (Tab.III).Notwithstanding, it is of major importance to continue monitoring the diet of M. leonina from Isla 25 de Mayo for a longer period of time (ideally a decade) to assess more precisely the possible existence of a correlation between oceanographic and climatic changes in the Southern Ocean resulting from the "El Niño Southern Oscillaton" and the feeding pattern of the different components of the population of this colony.
Tab. III.Indicated in bold are warm episodes of the ENSO based on a threshold of +/-0.5°C for the Oceanic Niño Index (ONI) [3 month running mean of ERSST.v3bSST anomalies (Smith et al., 2008)

Fig. 3 .
Fig. 3. Frequency of occurrence of the different cephalopod taxa preyed upon by juvenile elephant seals at I. 25 de Mayo/King George, discriminated by sex.