Introduction

The diversity and analysis of macrobenthic molluscs assemblages are likely to contribute to a better understanding of the structure and interactions inside the more complex assemblages in which they live and interact with other bottom organisms, because they have a particularly wide ecological and ethological spectrum.

Although frequent studies have also been made in recent years on the benthos assemblages living in the Southern Ocean (e.g., Sáiz-Salinas et al. 1997; Arnaud et al. 1998; Linse et al. 2002; Bowden 2005; Rehm et al. 2006; San Vicente et al. 2007), few studies have dealt with molluscan assemblages and ecology (Arnaud and Hain 1992; Cattaneo-Vietti et al. 2000; Arnaud et al. 2001; Schiaparelli et al. 2006), and only one work on Bellingshausen Sea (Troncoso et al. 2007) by means of a quantitative approach of soft-bottom molluscs from box-cores samples.

The BENTART research programme were carried out on board the RV Hespérides from 24 January to 3 March 2003 (BENTART 03) and from 2 January to 17 February 2006 (BENTART 06) in West Antarctica from Low Island in the South Shetland Islands to Thurston Island in the Bellingshausen Sea, including Peter I Island. As an integrated study of the benthic ecosystem (Sáiz-Salinas et al. 2008), the BENTART programme is a good opportunity for analysing the molluscan assemblages.

All molluscan classes were present in this region, but we will only include the shelled Gastropoda, Bivalvia and Scaphopoda in this study, because Aplacophora, Polyplacophora, and some Opisthobranchia have taxonomic gaps in their identification. On the basis of the data obtained during this survey, we try to discover how many molluscan assemblages are living in the study area; what are their species diversity; which species characterize those assemblages and which species they have in common; which are the most abundant, more distributed and rare species; and which are the possible ecological and physical causes that characterize those assemblages.

Materials and methods

The research programme of the BENTART cruises aboard RV Hespérides was carried out during the austral summers of 2003 and 2006. Forty-five stations were sampled at depths of 53–3,304 m (Table 1), in a study area including Low Island (South Shetlands Islands), West Antarctic Peninsula, Bellingshausen Sea and Peter I Island (Fig. 1). Benthic samples were collected using several sampling gears, including a quantitative box-corer (with an effective surface of sampling of 0.06 m2), an Agassiz trawl (2.01 m wide and 1.12 m high, having a mesh size of 10 mm; duration of tow 8 min), an epibenthic sledge (with an opening-closing system, which is activated by contact with the sea floor and with superimposed nets of 0.5-mm mesh size; rectangular opening, 40 × 80 cm; duration of tow 8 min) and a rock dredge (0.8 m width and 0.3 m height, with a mesh size of 10.0 mm; duration of tow 8 min). Since the total haul of an Agassiz trawl, epibenthic sledge and a rock dredge does not provide an accurate measure of density, the abundance per station was derived from a subsample of 50 l taken at random in each haul (Arnaud et al. 1998). The content of each sample was sieved using three mesh sizes: 5, 1 and 0.5 mm (in the box-cores), and 10, 5 and 1 mm (in the Agassiz, epibenthic sledges and rock dredges). The molluscs retained in the two largest sieves were counted to estimate their relative abundance in the total macrobenthic fauna, then were fixed in borax-buffered 4% formaldehyde in seawater and after 1 day transferred and preserved in 70% ethanol for further taxonomic study. Environmental parameters in stations were measured [depth, redox (Eh) and organic matter] and the bottom type was characterized (Table 1).

Fig. 1
figure 1

Position of the benthic stations sampled on the BENTART 03 and BENTART 06 cruises. Inset shows the location of the study area in Antarctica

Full size image
Table 1 Location, depth and environmental parameters on the surface of sediments from the survey stations: redox potential (mV), organic matter (OM, %) and bottom type
Full size table

Species were identified, quantified and data were organized into station by species matrices. For each sampling station, the total abundance (N) and species richness (S) were calculated, the Shannon–Wiener index (H′), and Pielou’s evenness index (J′) were used to assess species diversity and evenness, also the density (individuals × m2) was calculated in box-corer samples. For population studies, data were standardized (fourth root), so that the absolute number of individuals were comparable between stations. To identify the main assemblages, non-parametric multivariate techniques as described by Field et al. (1982) were performed using the PRIMER (Plymouth Routines in Multivariate Ecological Research) version 6.0 programme (Clarke and Gorley 2005). The Bray–Curtis similarity coefficient (Clarke and Warwick 1994) was applied to the transformed grouped data obtaining a similarity matrix. Hierarchical clustering with group-averaged linking and non-metric multidimensional scaling (MDS) was then performed using the similarity matrix. Possible differences in faunistic composition between groups of stations were tested using a one-way ANOSIM test. The SIMPER programme was next used to identify species that greatly contributed to the similarity into each station group. Based on the groups of stations defined using the Bray–Curtis coefficient, the species were classified according to the criteria of constancy C and fidelity F (Dajoz 1971), with formulae and scales according to Arnaud et al. (2001).

Results

A total of 2,889 living specimens of molluscs belonging to three classes, 49 families and 139 species were recorded (Fig. 2). The class Gastropoda was the best-represented in terms of species richness (25 families and 87 species), followed by Bivalvia (20 families and 47 species) and Scaphopoda (4 families and 5 species). The most diverse family was the gastropod Buccinidae (17 species), followed by Trochidae (12 species), Turridae (10 species), Muricidae (6 species), and the bivalves Yoldiidae and Galeommatidae (6 species each). In terms of abundance, the class Bivalvia accounted most of individuals (1,884 specimens) followed by Gastropoda (643 specimens) and Scaphopoda (362 specimens, Fig. 2b). The tiny bivalves Cyamiocardium denticulatum (460 individuals), Thyasira debilis (430 individuals), Adacnarca nitens (228 individuals), Yoldiella profundorum (161 individuals), the scaphopod Dentalium majorinum (159 individuals) and gastropod Neobuccinum eatoni (89 individuals) were the most abundant species.

Fig. 2
figure 2

Number of species and specimens of each molluscan class, indicating the best-represented families and most abundant species

Full size image

Faunistic parameters

Species richness (S), abundance (per sample gear), density, diversity (H′) and evenness (J′) varied widely (Table 2).

Table 2 Faunistic parameters at each station: species richness (S), abundance (N), density (D) (ind. m−2 in box-cores), Shannon–Wiener diversity index (H’) and Pielou’s evenness index (J’)
Full size table

Highest species richness was recorded in the Antarctic Peninsula (station PA39, 37 species), followed by two station of Peter I Island (PI 8 and PI 5, with 23 and 22 species, respectively), while the lowest values corresponded to stations whit one species, two of Bellingshausen Sea (MB10 and MB12) and two of Antarctic Peninsula (PA18 and PA40).

Maximal abundances were found at stations of Peter I Island preferably, being PI 5, PI 8 and PI 6 the stations with highest values (705, 344 and 211 specimens, respectively), the lowest were found at two sites of Antarctic Peninsula (PA18 and PA40) and one of the Bellingshausen Sea (MB10), each with one specimen only, agreeing with stations of lowest richness. The highest densities in box-cores samples agreed with sites of maximal abundances (PI 5, PI 8 and PI6).

Diversity (H′) varied between 2.93 (station PA39) and 0 (stations with one species: MB10, MB12, PA18 and PA40), with high values at stations close to Antarctic Peninsula and several stations in the middle of Bellingshausen Sea (e.g., stations MB38 and MB34). Evenness (J′) was usually high in most sampling sites, except in stations of Peter I Island and some of Bellingshausen Sea (e.g., MB9).

Most of the stations close to Peter I Island (PI 5, PI 6, PI 7 and PI 8) had the greatest number of molluscs (119–705 specimens) but simultaneously had the low values of diversity and evenness as a consequence of the presence of extremely abundant species, the tiny bivalves C. denticulatum, T. debilis and A. nitens; for example, among the three species added of about 600 specimens at station PI 5.

Molluscan assemblages and community structure

The cluster analysis revealed the presence of four groups of stations in two main clusters (Fig. 3) with low similarities levels of about 5%, each group is separated into two subgroups. The group A includes the stations of the northern Antarctic Peninsula and shallower stations of Peter I Island (Subgroup A1) and the majority of the stations of Bellingshausen Sea (Subgroup A2). Group B splits the stations of the southern Antarctic Peninsula, Low Island and deeper stations of Peter I Island (Subgroup B1) from some stations of Bellingshausen Sea (Subgroup B2). MDS ordination (Fig. 4) showed similar results to those of the dendrogram, with an acceptable stress value of 0.16. Two major groups (A and B) are segregated from top to bottom MDS graphic, and a gradient depth from the deeper stations of Bellingshausen Sea to Peter I Island, Antarctic Peninsula and Low Island can be identified from left to right. The ANOSIM test showed the highest significant differences on the faunistic composition between subgroups A1 and B2, and between A2 and A1 (Table 3).

Fig. 3
figure 3

Mollusc assemblages in the study area determined by cluster analysis based on the Bray–Curtis coefficient. Two main clusters are indicated and subgroups are delineated

Full size image
Fig. 4
figure 4

MDS plot of assemblages illustrating a deep-shallow water gradient across the study area. The subgroups delimited are derivate of cluster analysis

Full size image
Table 3 ANOSIM, pairwise test comparing groups (global R = 0.619); R values and significance levels are presented
Full size table

The results of the SIMPER analysis for the similarity into the several subgroups and the species that more contributed to similarity are shown in Table 4. The subgroup A2 is the most similar with an average similarity of 30.11. On the other hand, B1 is the less similar subgroup, with only a value of 10.91 of average similarity. Dissimilarity between the subgroups was high (above 0.98), being contributed mainly by T. debilis, Cyclocardia astartoides and C. denticulatum (for the pairs of subgroups A1 vs. B2 and A1 vs. B1), D. majorinum, Adamussium colbecki and Bathyarca sinuata (for the pairs of subgroups A2 vs. B1 and A2 vs. B2), Rhabdus cf. perceptus, Y. profundorum and T. debilis (B1 vs. B2), and T. debilis, D. majorinum and C. denticulatum (A2 vs. A1).

Table 4 Results of the SIMPER analysis
Full size table

The values of the constancy index (C) and fidelity index (F) for species in each group are listed in Table 5. The subgroup A1 (69 species) is located in the northern Antarctic Peninsula and western Peter I Island, except for MB4, which is off Thurston Island in Bellingshausen Sea. In all stations the sediment is muddy or fine sandy, except in PA20, the shallowest station, where is gravely. Shallow depths from 53 to 425 m characterized this subgroup, but PA24 was the only deeper station. Of the commonest species in this subgroup, no species were constant, 1 very common (T. debilis, preferential) and 12 common (Pseudamauropsis anderssoni, and Falsilunatia delicatula, exclusive; N. eatoni, Philobrya sublaevis, C. denticulatum, and Siphonodentalium dalli f. antarcticus, elective; Margarella refulgens, A. nitens, C. astartoides, and Cuspidaria infelix, preferential; and Onoba gelida, and Chlanidota signeyana, accessory).

Table 5 Constancy index (C, %) and fidelity index (F, %) in the subgroups A1 (Stations of Northern Antarctic Peninsula and Peter I Island), A2 (Stations of Bellingshausen Sea), B1 (Station of Low Island, Southern Antarctic Peninsula and Peter I Island) and B2 (Stations of Bellingshausen Sea)
Full size table

The subgroup A2 (54 species) gathers the shallowest 14 stations of Bellingshausen Sea from 438 to 1,814 m. The predominant sediment is muddy with gravel. This subgroup has one constant species (D. majorinum, accessory) and six common (B. sinuata, and Limopsis longipilosa, elective; Anatoma euglypta, and Striopulsellum cf. minimum, preferential; A. colbecki, accessory; and C. infelix, occasional). The other species are not very common.

The subgroup B1 (75 species, the highest record of richness) consists of the stations of Low Island, the southern Antarctic Peninsula and eastern Peter I Island, except MB34, of Bellingshausen Sea. The predominant sediment is muddy with gravel or small stones, except Low Island, whose bottoms are composed by sessile epifaunal communities. The majority of the stations presented low depths (<402 m), although some stations were located at very deep waters (PI 27, PI28 and PA42). The most noteworthy species are not constant and very common, only seven common species were recorded (Marseniopsis conica, elective; O. gelida, Pseudamauropsis aureolutea, and Y. profundorum, preferential; and A. euglypta, C. signeyana, and T. debilis, accessory).

The subgroup B2 (10 species, the lowest record of richness) includes the deepest stations from the Bellingshausen Sea (>1,847 m), which muddy and gravely sediments, except the station PA19, with comparatively low depth (513 m) and bottom of stones. The commonest species only are composed by one constant (R. cf. perceptus, elective) and one very common (D. majorinum, occasional).

Three species, the bivalves Y. profundorum and A. colbecki, and the scaphopod D. majorinum, are present in all these groups (Table 5). On the other hand, two species do not belong in any group, the gastropod Solariella antarctica and the bivalve Nucula austrobenthalis. The subgroup B1 had the highest number of exclusive species (32), followed by A1 (29), A2 (21) and B2 (2).

The scaphopod D. majorinum was the most abundant and widely distributed species throughout the Bellingshausen Sea. In Peter I Island the most abundant species were the tiny bivalves C. denticulatum, T. debilis and A. nitens, with the highest abundances observed in the whole study area. In the Antarctic Peninsula, the bivalve Yoldia eightsi was the most abundant, although it was followed by Y. profundorum, which was the widely distributed species throughout this area. In Low Island, the bivalve Lissarca notorcadensis was the most abundant, followed distantly by P. aureolutea, the most widely distributed.

Discussion and conclusions

This study is a semiquantitative approach of the entire study area of BENTART 03 and 06 cruises considering all samples gears for the analysis. Therefore, these results agree with the comparison with neighbouring areas made in the specific works of quantitative analyses of soft-bottom molluscs from box-cores samples (Troncoso et al. 2007) and bathymetric zonation and diversity gradient of gastropods and bivalves (Aldea et al. 2008). However, this work extends the number of species and analyses of molluscan assemblages from 62 species (Troncoso et al. 2007) and 134 species (Aldea et al. 2008) to 139 species.

Gastropods occurred in smaller abundances than bivalves, but with higher species richness, agreeing with other studies in neighbour areas (Arnaud and Hain 1992; Arnaud et al. 2001). At families’ level, the most diverse family of gastropods was the Buccinidae, which has a high diversity in Antarctic waters (Dell 1990), in the bivalves was the Yoldiidae that coincidently is the most diverse family in the Scotia Arc Islands (Linse 2004). At species’ level, the most abundant species in Antarctic Peninsula and Peter I Island were C. denticulatum and A. nitens, which agree with those found in the South Shetland Islands (Arnaud et al. 2001), while T. debilis and N. eatoni were abundant in the Ross Sea (Cattaneo-Vietti et al. 2000).

Molluscan assemblages exhibit a bathymetric pattern, which was previously detailed and explained in this area (Aldea et al. 2008) and some trends were reported in soft-bottom molluscs (Troncoso et al. 2007). Furthermore, a depth gradient from deeper stations of the western sector of the Bellingshausen Sea to the eastern shallow water stations closer to the Antarctic Peninsula was identified in benthic assemblages (Sáiz-Salinas et al. 2008). Shallow bottoms of muddy or fine sandy sediments are dominated by T. debilis and D. majorinum, and deeper bottom of muddy and gravely sediments are dominated by R. cf. perceptus and D. majorinum.

In continental shelf zone there are two factors (i.e, bottom type and depth) that group the northern Antarctic Peninsula with Peter I Island (subgroup A1), of low depths (53–425 m) and muddy or fine sandy sediments, although the values of O.M. and redox do not indicate any trend. On the other hand, southern Antarctic Peninsula with Low Island and eastern Peter I Island (subgroup B1) were grouped. The blackish colour of the sediment revealing a medium-high values of O.M. (Troncoso, pers. obs.), and muddy sediments with gravel or small stones indicate the grouping of the stations of these distant areas with predominantly low depths (<402 m).

Deeper bottoms are composed principally of muddy and gravely sediments and are disperse in subgroups A1, A2 and B1. The subgroup B2 gathers four deepest stations of Bellingshausen Sea. In fact, the distribution of assemblages of macrobenthos are improved by a synergy of depth, redox values and O.M. content of the sediment (Sáiz-Salinas et al. 2008), the qualitative and quantitative differences between molluscs assemblages follow the same tendency.

Knox (1994) and Arntz et al. (1994) considered that the decreases in the supply of organic matter as depth increases are the main limiting factor for the Antarctic benthos. The benthic assemblages are controlled by three major factors too: food supply, temperature and sediment regime (Mühlenhardt-Siegel 1988, 1989; Brey and Clarke 1993; Sáiz-Salinas et al. 1997; Piepenburg et al. 2002; Skowronski and Corbisier 2002; Lovell and Trego 2003). Temperature and food and their seasonal oscillations influence metabolism and growth of Antarctic marine invertebrates and hence are important factors structuring benthic assemblages (Clarke 1988; Brey and Clarke 1993; Brêthes et al. 1994).

High rates of sedimentation may inhibit suspension-feeding organisms (Lovell and Trego 2003). In this sense in the molluscs’ distribution in the Bellingshausen Sea, two trophic strategies appear: filter feeders like C. denticulatum, and deposit-feeders like Y. eightsi are present in the shallow waters, and scavengers with broad bathymetric distribution, like C. signeyana, and carnivorous, like Trophon drygalskii in deep waters, this pattern was previously reported with similar assemblages of molluscs in neighbour areas by Arnaud et al. (2001).

Interactions with other organisms and biological factors such as developmental mode, fecundity, egg size or mobility could play an important role in shaping patterns of species and molluscs assemblages distribution (Aldea et al. 2008).

Finally, ice or iceberg impacts can also have a very considerable influence on the structure of benthic assemblages (Peck et al. 1999; Gutt 2001). Disturbance from ice in Antarctic benthic marine systems varies at different spatial and temporal scales (Peck et al. 1999; Gerdes et al. 2003). This heterogeneity may enhance the diversity of benthic assemblages (Gutt et al. 1999, 2001; Peck et al. 1999; Gerdes et al. 2003). Sáiz-Salinas et al. (2008) confirm this fact in the Bellingshausen Sea, by mean of the intense iceberg traffic from the mainland (such as iceberg scouring over the sea-floor, high sedimentation rates and intense fall of drop-stones onto the sea floor).