Mycale (Aegogropila) magellanica (Porifera: Demospongiae) in the southwestern Atlantic Ocean: endobiotic fauna and new distributional information ; Mycale (Aegogropila) magellanica (Porifera: Demospongiae) en el Atlántico suroeste: fauna endobiótica y nuevos datos de su distribución

1 Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET). 2 Instituto Nacional de Investigación y Desarrollo Pesquero (INIDEP), Paseo Victoria Ocampo 1, 7600 Mar del Plata, Argentina. E-mail: schejter@inidep.edu.ar 3 Instituto de Investigaciones Marinas y Costeras (IIMyC, CONICET-UNMdP). 4 Departamento de Biodiversidad y Biología Experimental, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, C1428EHA, Buenos Aires, Argentina.


INTRODUCTION
The sponge Mycale (Aegogropila) magellanica (Ridley, 1881) is distributed in Antarctic, sub-Antarctic, Chilean and Argentine waters (van Soest et al. 2011).In this last region, M. (A.) magellanica is one of the most widely distributed, 39º10'S and 56º20'W being its northern distribution limit in the Atlantic (López Gappa and Landoni 2005) (Fig. 1).In spite of its wide distribution, relatively frequent finding and easy identification compared with other sponges in the shelfbreak frontal area of Argentina (Schejter et al. 2006, Bertolino et al. 2007), no studies have been performed on this species, besides the original description and subsequent faunistic records.
The relation between sponges and their associated organisms ranges from accidental or intentional commensalism to predation, mutualism or parasitism (Sará and Vacelet 1973, Pawlik 1983, Wendt et al. 1985, Wulff 2006, Winfield and Ortiz 2010, and references therein).Endobiotic taxa registered worldwide comprise representatives of up to 11 phyla (Rützler 1976).The most common endobionts are protozoans, diatoms, cnidarians, polychaetes, crustaceans, echinoderms and other invertebrates such as pycnogonids, platyhelminths, sipunculids and nemertins, but they also include fishes (see Supplementary material Appendix 1 for examples; and Wulff 2006 for review).Klitgaard (1995) suggested that the majority of the fauna associated with sponges in temperate/cold waters is composed of facultative inhabitants, while those of the warm tropical waters are apparently obligate sponge associates.Endobionts obtain refuge, direct or indirect sources of food and sites for reproduction, while they may contribute to host defence against predators or sediment removal (Wendt et al. 1985, Saffo 1992, Thiel 2000, Poore et al. 2000, Wulff 2006).
The kind of endobionts hosted by a species is related to the morphology and the chemical defence of the host sponge (e.g.Neves andOmena 2003, Skilleter et al. 2005), among other biotic interactions (e.g.Koukouras et al. 1992).Rützler (1976) found a positive relation of the sponge canal volume to total mass of endofauna, but many other studies showed no relationship between abundances of endobionts and volume of the sponge (Koukouras et al. 1985, Ota et al. 2008), presumably due to the high variability in associated endobiont fauna among sponge samples.
López Gappa and Landoni (2005) reported the presence of 196 species in the Argentine Sea; the spongeinvertebrate associations of many of these species have not been studied yet.In this paper we studied the endobiotic fauna of Mycale (A.) magellanica and we compared these results with those of other sponge species.We discuss our findings in relation to the well-known benthic richness of the study area.Additional records for the distribution of the sponge M. (A.) magellanica in the shelf-break frontal area of the Argentine Sea are also given.

MATERIALS AND METHODS
Samples for ecological purposes (more than 100 samples consisting of a 10-litre volume each) are collected as a routine procedure during Patagonian scallop Zygochlamys patagonica stock assessment cruises performed yearly by the Instituto Nacional de Investigación y Desarrollo Pesquero (INIDEP).These evaluation cruises are carried out in the shelf-break frontal area of Argentina, SW Atlantic Ocean, a region that supports the scallop fishery.General samplings during the assessment consist of 10-minute trawling or dredging (depending on the vessel) tows (1 tow per site) at a speed ranging between 3.5 and 4 knots.As a standard procedure, scallop stock assessment samples are processed on-board while ecological samples are frozen and then sorted at the INIDEP laboratory.The majority of the invertebrate species are identified to specific level during the analysis of the ecological samples in the laboratory.However, considering that the identification of groups such as ascidians, sponges and bryozoans is very difficult and time consuming, these organisms are usually grouped into single major taxa for the community assessment purposes (e.g.Schejter and Bremec, 2007b).Samples or vouchers of these organisms are often preserved for taxonomic and/or ecological studies.Sponges, particularly, could represent a variable percentage of the community that averaged 10% in biomass (Bremec and Lasta, 2002).When their contribution in biomass to the total community is considered relevant (although in relative terms), we keep vouchers or sub-samples of the morphospecies for subsequent analysis.
In relation to the objectives of the present study, while sorting ecological samples, we selected Mycale (A.) magellanica specimens from benthic community samples collected at 12 sites (localities are given in Table 1) where a conspicuous biomass of sponges was registered.Owing to the sampling methods with trawls and dredges, organisms are frequently collected broken or damaged.In consequence, the selection of these 12 samples does not mean that M. magellanica was absent at the other sampled sites, although the species remained unidentified during routine sorting.The biggest sponge pieces collected were separated and dissected for the current study: six sponge pieces between 340 and 660 g wet weight totalling 2.7 kg, from four localities between 39º24'S-41º51'S and 55º56'W-58º09'W and between 103 and 107 m depth.Each sponge piece was first weighed (wet weight) and the volume was determined by water displacement.Then, the sponge piece was cut, disaggregated and examined under a binocular microscope at the Benthos Laboratory (INIDEP).All the endobionts found inside the sponge were sorted, counted and identified to the lowest taxonomic level possible.
Identification of the sponge species was done using the classical procedure based on spicules and skeleton architecture observations, according to Rützler (1978).Scanning electron microscope (SEM) was also used for the observation of spicule morphology.
Six sponge pieces were dissected (wet weight: 369, 555, 338, 660, 417 and 355 g) and a total of 849 endobiont individuals were sorted (Table 2).Considering water displacement measures, in general 1 g of sponge corresponds to a volume of ~1 ml (in all six pieces measured).Hence, on average, the sponge M. (A.) magellanica had a mean density of ~348 individuals per litre (or kilogram).Twenty three taxa grouped into five major groups (Crustacea, Mollusca, Echinodermata, Polychaeta, and Sipunculida) were identified.Crustaceans, mainly Amphipoda and Isopoda, reached between 66% and 96% of the total number of individuals inhabiting the sponges.The most frequent and abundant species were the amphipod Aristias cf.antarcticus Walker, 1906 and the isopod Caecognathia sp.For both species, adult males, brooding females and juveniles were found.Four other species of amphipods and three of isopods were recorded in lower abundance (Table 2).The structure, morphology and internal architecture of M. (A.) magellanica were very heterogeneous in each single piece (see Fig. 2B, C).Usually the basal part was compact and intricate, and no endobionts were detected (except for a few bivalves identified as Hiatella meridionalis (d'Orbigny, 1846) and some ophiuroids).From the base to the surface, the sponge tissues displayed less dense and softer structures that hosted the majority of the amphipods, isopods and other faunal components (Fig. 2D).Living H. meridionalis were found in holes, depressions or crevices of the sponge.However, a few empty shells were also recorded completely embedded inside sponge tissues.

DISCUSSION
This study is the first attempt to elucidate the composition of endofauna (excluding bacterial or fungal symbionts) in sponges from the deep waters of the Argentine Sea.In this regard, it is important to mention that the species Mycale (Aegogropila) magellanica has been previously recorded in the Argentine Sea, mainly in South Patagonian waters, around the Malvinas Islands with scattered records from the Beagle channel and coastal waters of the provinces of Tierra del Fuego and Chubut.Until now, there was only one record from waters off Buenos Aires province (see López Gappa and Landoni 2005).Our results extend the distribution of this species northwards in the SW Atlantic Ocean (Fig. 1).Mycale (A.) magellanica seems to be important in providing habitat for at least 23 taxa of small invertebrates.Crustaceans were the most important group in abundance and diversity in this study.It is noticeable from the general literature that, among amphipods, members of the Aristiidae, Colomastigidae, Leucothoidae and Sebidae are known to be frequent endobionts of sponges and other sessile invertebrates (LeCroy 2009, Thomas and Klebba 2007, Winfield et al. 2008, White and Thomas 2009, Winfield and Ortiz 2010, Kilgallen 2010, and references therein).However, the species collected in the present study had not been previously recorded inside sponges in the Argentine Sea (see López Gappa et al. 2006, De Broyer et al. 2007).
Until this study, Colomastix bastidai Alonso de Pina, 1993 was only known from the type locality, a position adjacent to the new records mentioned herein (see López Gappa et al. 2006, Table 3).Our finding of C. bastidai represents the second record of the species in Argentine waters.Leucothoe spinicarpa (Abildgaard, 1789) has been reported from polar, temperate, and tropical waters (Thiel 2000).In the Argentine Sea, several records of L. spinicarpa have been reported, many of them in the Magellanic Biogeographic Region and only one in the Argentine Biogeographic Region (see López Gappa et al. 2006).However, according to De Broyer et al. (2007), the Southern Ocean records of L. spinicarpa probably belong to one or more southern species.Because of the need for a revision of this species and in order to avoid more taxonomic confusion, we prefer to keep the denomination of Leucothoe cf.spinicarpa.Similarly, Aristias antarcticus Walker, 1906 has been cited from the Malvinas Islands in the Argentine Sea, and it is also widely distributed in Antarctic and sub-Antarctic waters (López Gappa et al. 2006, De Broyer et al. 2007).The systematics of the southern Aristias spp. is in disarray and all previous records of A. antarcticus require confirmation (Kilgal- len 2010).Therefore, we also keep the denomination of Aristias cf.antarcticus.Because of the wide geographic distribution of Seba saundersii Stebbing, 1875 (in the southern Argentine Sea, sub-Antarctic and Antarctic waters) and the poorly preserved condition of the specimens examined, we identified them as Seba cf.saundersii.For both genera, Aristias and Seba, the new records presented herein are the northernmost known in the Argentine Sea (see López Gappa et al. 2006).Among the isopods, the second most important group of endobiont crustaceans found, the species Caecognathia sp.(Gnathiidae), was recorded in high abundances.Caecognathia antarctica (Studer, 1884) is the only species of gnathiid isopod reported from the Argentine Sea.The specimens found in our samples most probably belong to this species; however, since Studer (1884) briefly described C. antarctica based on a single juvenile, it is not possible to identify this species with certainty.The life cycle of gnathiid isopods involves a parasitic larval phase and a non-feeding adult phase (Monod 1926).The resting larvae and the adult stages are usually found in sponges, tubes of serpulid worms, coral rubble or sediment cavities (Monod 1926, Upton 1987, Wägele 1988, Klitgaard 1997, Smit and Davies 2004).Almost all the specimens of Caecognathia sp.obtained in our samples were males.The lower abundance of females in sponges was also reported by Smit et al. (2003) and Barnard (1914); the latter author found the females inhabiting the tubes of serpulid worms.Three other species of isopods were found in the studied sponge, Fissarcturus patagonicus (Ohlin, 1901), Acanthoserolis schythei (Lütken, 1858) and Iathrippa sp., but each species was represented by one or two individuals.Klitgaard (1995) pointed out that owing to sampling procedures, contamination of the sponge with foreign fauna could be expected.Since these three isopod species were recorded in very low densities and none of them were reported as sponge associates, the finding of these species on M. (A.) magellanica can be considered accidental.
The finding of juvenile stages and brooding females of Aristias cf.antarcticus and Caecognathia sp., although in low abundances, shows that at least a couple of species have a close relationship with M. (A.) magellanica, and part of their life cycles probably takes place inside the sponge, as is already known for gnathiids.
In the study area, where the main Patagonian scallop (Zygochlamys patagonica) fishing grounds are located (the shelf-break frontal area), general knowledge on the biodiversity of the small-sized taxa is very scarce.The species composition of this benthic fraction (bigger than 1 mm) was recently assessed by means of samples taken with Picard dredge (Sánchez et al. 2011).In coincidence with our results, these authors also found that crustaceans were the most diversified group.The biological material studied by these authors and our present samples come from nearby locations.Only the genus Iathrippa (Isopoda, Asellota, Janiridae) is common to both habitats, although its finding in M.(A.) magellanica could be incidental.Nonetheless, many other crustaceans (especially Lysianassidae spp.) and polychaetes were identified to family level by Sánchez et al. (2011), which could indicate a higher species overlapping between the two studies.
Other infaunal groups that also appeared in low abundance in the studied sponges were polychaetes and sipunculids (only six taxa, see Table 2).Polychaetes are usually one of the dominant groups in sponges (see Supplementary material Appendix 1).A clear relationship between faunal densities and sponge morphology was established by comparing many studies and, for example, in the case of a syllid species, it was reported that lobate sponge species are able to grow faster than massive ones, supporting higher densities of worms (Neves and Omena 2003).Considering the endofauna hosted by other Mycale species (Duarte andNalesso 1996, Ribeiro et al. 2003, Table 1), in which polychaetes were the dominant group in one of the species but crustaceans were dominant in the other one, it remains unclear whether the morphology or internal architecture of M. (A.) magellanica could be responsible for hosting only low densities of polychaetes but high densities of crustaceans.
In general, the highest endobiont richness was found in sponges at low depths and low latitudes (warmest waters) (Supplementary material Appendix 1).The highest endobiont richness could be found in Aplysina aerophoba Nardo, 1883, Sarcotragus fasciculatus (Pallas, 1766) and Agelas oroides (Schmidt, 1864) from the Aegean Sea, with more than 100 associated species each.Considering only the Mycale genus, our species showed the lowest endobiont richness, in accordance with the highest latitude and depth.
The shelf-break frontal area in the Argentine Sea is one of the most productive ecosystems in the SW Atlantic Ocean (Acha et al. 2004, Bogazzi et al. 2005) as a consequence of high levels of nutrients (Rivas 2006, Romero et al. 2006).This region is dominated by soft-bottoms (sand and mud), like more than 70% of the Argentinean continental shelf (Parker et al. 1997).In these habitats, epibiotic relationships are known to increase the specific diversity by providing substrate for the attachment of sessile species, given the lack of rocks or hard bottoms (Schejter andBremec 2007a, Schejter et al. 2011).Our study thus shows that the sponges themselves also enhance benthic biodiversity, as they are able to shelter a variety of invertebrate species.The endobiont richness is a valuable contribution to local biodiversity, though the values found in M. (A.) magellanica are smaller than those shown by other more tropical sponge species.
In soft and flat substrates, erect and sessile epifauna usually play the role of ecosystem engineers, as they structure the architecture of the sea bottom by increasing its complexity.Mass removal of this fauna could have devastating effects on local biodiversity (Coleman andWilliams 2002, Abdo 2007).Trawling and dredging activities are the main causes of loss of erect and sessile epifauna (National Research Council 2002, Bremec et al. 2000, 2008).Therefore, fishing activities on the Patagonian scallops beds could affect the settlement of M. (A.) magellanica, thus leading to the loss of associated endobionts.

Fig. 1 .
Fig. 1. -Distribution of Mycale (A.) magellanica (Ridley, 1881) in the Argentine Sea.Light shadow areas represent previous records, dark shadow areas represent sampling sites of the present study.