Composition and distribution of subtidal and intertidal crustacean assemblages in soft-bottoms of the Ria de Vigo (NW Spain)

The intertidal and subtidal soft-bottoms of the inner area of the Ria de Vigo (NW Spain) were sampled in November and December 1999, and spatial distribution of crustacean species was examined. Environmental variables from water and sediment were measured at each sampling site. Amphipods and myocopids were the numerically dominant orders (49.9 and 26.9% dominance), amphipods accounting for more than 54% of identified taxa. The highest crustacean densities occurred with 55-41 species and 5953.6-4346.4 ind. m–2 in external areas, where the diversity index reached the maximum values. Multivariate techniques revealed that distribution of crustaceans in the inlet was highly dependent on depth. Ordination analysis determined three major assemblages: Intertidal bottoms colonized by seagrasses and subjected to strong variations of salinity were dominated by the amphipod Melita palmata, harpacticoids and the isopod Idotea baltica (Group A). The amphipod Corophium cf. runcicorne and the cumacean Iphinoe tenella predominated in the muddy bottoms of central areas (Group B). These species were also present in the deep muddy bottoms of the mouth of the inlet, with high carbonate and gravel contents, and with the myocopids and the amphipod Microdeutopus cf. armatus displaying maximum dominances (Group C).


INTRODUCTION
The study of the composition and distribution of benthic communities is of great interest, because they are considered as good indicators of the conditions of marine sediments (Grall and Glémarec, 1997;Conradi and López-González, 2001).Crustaceans are an important component of soft-bottom benthic commu-nities in temperate latitudes.Amphipods in particular play an important role in structuring benthic assemblages (Duffy and Hay, 2000) as secondary and tertiary producers in marine communities (Guerra et al., 2002).Dauvin (1988) and Beare and Moore (1996) showed amphipods to be an important source of food for benthic fauna of commercial interest.Amphipods are also very ecologically sensitive organisms and good indicators of natural or disturbed environmental conditions (Conradi et al., 1997).Moreover, benthic databases are essential for comparisons which can be used in impact studies or monitoring programmes, in order to preserve the environments and the species of commercial importance that they support (Desroy et al., 2002).
The highly macrobenthic diversity characteristic of the Galician rias (NW Spain) is due to their great variety of habitats and sediments (Troncoso and Urgorri, 1993;Moreira et al., 2009).They are characterized by a regular incoming of nutrients due to upwelling (Nombela et al., 1995), which is translated into a high primary productivity.However, the ria seashores are also highly populated and therefore subject to many anthropogenic perturbations (e.g.culture of bivalves and construction of harbours).This is translated into organic enrichments and changes in sedimentary composition (López-Jamar and Mejuto, 1985;Moreira et al., 2009).
The Ria de Vigo is one of the largest and most complex estuarine systems on the Galician coasts, and the first in terms of economic importance and human population.During the last forty years it has been extensively studied, especially with regard to its oceanography, fisheries, mussel culture on rafts and shellfish resources (Abella et al., 1996;Pérez-Arlucea et al., 2000).Previous studies on the benthic fauna of the Ria de Vigo only referred to specific areas and/or faunistic groups (López-Jamar and Cal, 1990;Moreira et al., 2004Moreira et al., , 2009;;Cacabelos et al., 2008a, b).Except for the studies of Cacabelos et al. (2008c) about the macrofaunal assemblages found in the Ensenada de San Simón as a whole, in which some of the most relevant arthropods are mentioned, the crustacean fauna of the inlet has not been properly studied.
Since grain size is one of the most often reported factors affecting the distribution of crustaceans elsewhere, we hypothesized that in this particular inlet this factor could influence the pattern of crustacean zonation.Therefore, the objectives of the present study are (i) to describe the structure of the crustacean communities in intertidal and shallow subtidal areas of the Ensenada de San Simón, (ii) to analyse the interactions between the different assemblages and (iii) to relate any observed faunal pattern to sediment characteristics and other environmental variables.This knowledge will be essential to ensure the correct management of resources in the area, especially since it has been included in the Nature 2000 Network as a Special Conservation Zone.

Study area
The Ensenada de San Simón is located in the innermost part of the Ria de Vigo, between 42º17' and 42º21'N and between 8º37' and 8º39'W (Fig. 1).Soft-bottoms of this inlet are mainly muddy with high organic matter content (0.69-10%) (Vilas et al., 1995).Intertidal and shallow subtidal areas have meadows of the seagrasses Zostera noltii Hornem.1832 and Zostera marina L. The culture of mussels on rafts is a common practice in large areas of the mouth of the inlet, where 75 rafts are exploited.Two small harbours are located in the inlet sides.The major hydrological features of the inlet are the large freshwater inputs occurring in the innermost part of the inlet, resulting in drastic temperature and salinity fluctuations on a tidal and seasonal basis (6-35.3 psu;Saiz et al., 1961;Nombela and Vilas, 1991).Previous studies (Cacabelos et al., 2008a, b, c) have included detailed information about this study area.

Sampling and sedimentary analysis
Quantitative samples were collected from 29 sites using a van Veen grab in November and December 1999 (Fig. 1).Five replicate samples were taken at each site (0.056 m 2 ).Samples were sieved through 0.5 mm mesh and the retained material was fixed in 10% buffered formalin.Fauna was sorted from the sediment and preserved in 70% ethanol for identification.Temperature and pH were measured in situ from water and sediment.An additional sedimentary sample was taken at each site for later grain-size analysis and to determine calcium carbonate and total organic matter contents.Sedimentary types were determined according to Junoy (1996).Median grain size (Q 50 ) and sorting coefficient (S o ) were also determined for each sample.Kurtosis (Kg) and skewness (Sk) coefficients were calculated according to Folk and Ward (1957).Calcium carbonate content (%) was estimated by sample treatment with hydrochloric acid, and total organic matter content (%) was estimated from the weight loss after placing samples in a furnace for 4 hours at 450ºC.

Data analysis
Abundance data of each crustacean species were organized in matrices, and the five samples taken at each site were pooled (0.28 m 2 ).Total abundance of crustaceans (N), number of species (S), Shannon-Wiener diversity index (H', log 2 ) and Pielou's evenness index (J) were determined for each site.Correlations between these diversity measures, the abundance of dominant species and environmental variables were determined using Spearman's non-parametric correlation coefficient.Crustacean assemblages were determined through non-parametric multivariate techniques using the Plymouth Routines of the Multivariate Ecological Research software package (PRIMER; Clarke and Warwick, 1994).A similarity matrix was performed using the Bray-Curtis coefficient after applying the fourth-root transformation to species abundance.Classification and ordination of sites and species were performed by cluster analysis through the algorithm UPGMA and non-metrical multidimensional scaling (MDS), respectively.
Relationships between abundance of crustaceans and environmental variables were studied by means of the BIOENV procedure (PRIMER package) and canonical correspondence analysis (CANOCO package;ter Braak and Prentice, 1988).Environmental variables expressed in percentages were previously transformed by log (x+1) and all of them were normalized.

RESULTS
The soft bottoms of the Ensenada de San Simón were characterized by a predominance of muddy sediments with a high organic matter and low calcium carbonate contents.Sandy sediments were present in tidal channels in the inner inlet where low total organic matter content was found.The shallow sediments became increasingly muddy towards the deeper bottoms in the centre and at the mouth of the inlet.The areas around the outer part had muddy sands with a large gravel fraction composed by mussel shells.
The highest numbers of species and densities were recorded at sampling sites 27, 22 and 26, with 55-41 species and 5953.6-4346.4ind.m -2 .The lowest densities and number of species were recorded at sites 12, 29 and 24, with 0-35.7 ind.m -2 and 0-7 species.The Shannon-Wiener diversity index ranged between 4.20 (site 26) and 0.00 (sites 12 and 29).Evenness showed low values on bottoms with high dominances of myocopids and the amphipods Harpinia spp.(site 17) or the harpacticoids (site 10).
Spearman's correlation coefficient showed that depth was positively correlated with number of species and Shannon-Wiener diversity index (p<0.01).Temperature of bottom water and sorting coefficient showed positive correlations with number of species (p<0.05), while carbonate content was positively correlated (p<0.05) with abundance of individuals (Table 1).

Crustacean assemblages and species affinities
The dendrogram obtained through cluster analysis based on abundance data showed the presence of three major groups (Fig. 3; assemblages illustrated in Fig. 1): Group A, composed of intertidal sites; Group B, located in the shallow muddy bottoms of the centre of the inlet and Group C, composed of the deeper muddy bottoms of the mouth of the inlet.These same groups appeared in the MDS ordination (Fig. 4), in which site 12 was eliminated due to its absence of individuals.The abiotic and faunistic characteristics of these assemblages   are summarized in Table 2. Dominant species and their constancy and fidelity indices are also shown in this table .Cluster analysis based on abundance data of the species with dominances higher than 1% (inverse analysis) showed 4 main groups (Fig. 5).
Group A was located in the innermost part of the inlet (Fig. 1), in intertidal sediments subject to strong variations of salinity close to the mouth of the rivers Oitabén-Verdugo and Xunqueira.Sites showed high granulometric differences, with sedimentary types ranging from mud to very coarse sand.These bottoms were poor in total number of species and Shannon-Wiener diversity index (Table 2).The seagrasses Zostera marina and Z. noltii were spread across most of these intertidal and shallow bottoms, with depths from 3.6 m to intertidal levels.Species with the highest abundance were, cited in decreasing order, the amphipod Melita palmata, the harpacticoids, the isopod Idotea baltica, the amphipod Microdeutopus gryllotalpa, the podocopid Propontocypris sp., the tanaidacean Zeuxo normani and the podocopid Cytheracea sp.The harpacticoids, the amphipods M. palmata, M. gryllotalpa and Corophium spp., the podocopid Propontocypris sp. and the isopod I. baltica were the species with the greatest similarity contribution for the group.A total of 22 species were exclusive to this assemblage, while only the harpacticoids were constant.This assemblage was linked to Group 1 of inverse analysis (Fig. 5), since it included M. palmata, I. balthica and M. gryllotalpa.
Group B was present in the shallow muddy bottoms of the centre of the inlet.Sediments were mainly composed of silt and clay with a high organic matter content.The average individual density was the lowest of the inlet, and the sites showed the greatest mean evenness Table 2. -Ecological features of the crustacean assemblages determined in the Ensenada de San Simón, indicating ranges and average ± standard deviation of biotic (values per 0.28 m 2 ) and physical characteristics.S, number of species; N, number of individuals; J', Pielou's evenness; H', Shannon Wiener diversity index; Q 50 , median grain size; Bt, bottom type (M, mud; CS, coarse sand; SM, sandy mud, MS, muddy sand); OM, organic matter content; CO 3 , carbonate content.Species that add up to 75% of the total abundance in each assemblage are listed in order of dominance (%), and constancy (Ct, constant; VC, very common; C, common) and fidelity (Ex, exclusive; El, elective; Pr, preferential; Ac, accessory; Oc, occasional) are indicated.2, 3, 4, 5, 6, 7, 10, 11, 15, 20, 25 8, 9, 13, 16, 18, 19, 28 14, 17, 21, 22, 23, 26 Group C was situated in the mouth of the inlet, reaching up to 28.2 m depth.Sediments were predominantly composed of silt and clay, but showed the highest carbonate and gravel contents.The number of species, the density of individuals and the Shannon-Wiener diversity index were the highest found in the inlet (Table 2).Species with the highest abundance in these bottoms were myocopids sp. 4 and sp. 1, the amphipods M. cf.armatus, M. simplex, H. crenulata, H. dellavallei, H. antennaria and A. tenuicornis and the tanaid Tanaopsis graciloides.A total of 38 species were exclusive to this assemblage, and 17 were constant.Among the dominant species, H. crenulata, Microdeutopus versiculatus, Myocopida sp.1, L. incisa, H. dellavallei and the tanaids T. graciloides and Leptochelia savignyi were exclusive to this group.Groups 3 and 4 of figure 5 were linked to this assemblage.

Relationships between crustacean fauna and environmental variables
According to the BIOENV analysis, the crustacean assemblages correlated best with bottom temperaturedepth-kurtosis coefficient and bottom temperaturedepth-sediment temperature-kurtosis coefficient combinations (Spearman's rank correlation ρ w : 0.250 and 0.243 respectively).Temperature of bottom water and depth were the variables with the best values when each variable was considered alone (ρ w : 0.221 and 0.215 respectively).
Canonical correspondence analysis showed that the first two axes accounted for 37.24% of the total variance of species-environment relation, and 30.7% of the species variance (Fig. 6).Forward selection in this analysis selected bottom temperature, depth, coarse silt, fine sand and very fine sand as the variables explaining most of the variance in the species data (p<0.01).The graphic representation showed an ordination of sites following a gradient in depth, bottom temperature and grain size (Fig. 6).Sampling sites from group A were distributed along positive quadrants of axis I, in shallow bottoms following an increase in content of coarse silt and very fine sand.Groups B and C followed an increase in depth appearing distributed along the negative part of axes I.
According to the results obtained through these analyses (BIOENV, CCA), the distribution of the crustacean fauna in the Ensenada de San Simón was mainly related to depth, temperature of bottom water and sediment grain size.

DISCUSSION
Results showed that distribution of the crustacean fauna in the Ensenada de San Simón was primarily linked to temperature of bottom water and depth.Quantitative sampling showed that the number of species observed in the Ensenada de San Simón (111 taxa) was similar to that found in other Galician rias.Garmendia et al. (1998) found 66 amphipod species in the Ria de Ares-Betanzos, while Lourido et al., 2008, found 125 peracarid species in the Ria de Aldán.Our number of species was high in comparison with those reported from other European estuaries (e.g.20 crustacean species were found in the Ria de Huelva (Cano and García, 1987), 51 in the Arcachon Bay (Bachelet et al., 1996), 23 in the Rance Basin (Desroy and Retière, 2001), 14 in the Bidasoa Estuary (Garmendia et al., 2003) and 50 malacostracean species in the Santander Bay (Lastra et al., 1990)).However, as can be observed in Table 3, some authors sieved organisms with 1 mm mesh, which could involve the loss of small organisms (e.g.Podocopida or Harpacticoida).
Both crustacean diversity and number of species were greater on subtidal bottoms and impoverished towards the brackish intertidal areas, as was also reported from other estuarine suprabenthic communities (Mees et al., 1995;Cunha et al., 1999).Aquatic fauna are submitted to stressful conditions in intertidal zones (e.g.desiccation, fluctuations in salinity; Kikuchi, 1987), derived from changes in freshwater inputs.Intertidal and subtidal soft bottoms of the Ensenada de San Simón were sampled in autumn 1999, the rainy season in these temperate latitudes.In the inlet, changes in salinity are drastic during the year (Saiz et al., 1961;Vilas et al., 1995).Similar areas with more stable salinities and temperatures during the year showed higher biodiversities (e.g.Lourido et al., 2008).Euryhaline species such as Melita palmata or Corophium acherusicum benefit from salinity fluctuations.However, our study referred specifically to the autumn, and the conclusions should be interpreted carefully.Different structures of the crustacean community were found during the year.For most species from temperate latitudes the larval settlement takes place during the summer period (Pfister, 1997;Pallas et al., 2006) or during two annual recruitment cycles (Upchurch and Wenner, 2008), especially in the nursery habitats of seagrass meadows (Heck et al., 2003).The existence of these vegetated bottoms could be relevant in the temporal evolution of the crustacean community.To present a complete study of the assemblage structure, an analysis of the seasonal pattern should be the next step (as in García Muñoz et al., 2008), integrating the life cycle of the fauna and looking for the strongest factors influencing its distribution.Distribution of the crustacean fauna in the inlet was linked to temperature of bottom water and depth, which directly influence the hydrodynamic conditions.The importance of the temperature should be interpreted carefully, since Ysebaert and Herman (2002) pointed out that long-terms averages of environmental variables are more important than values obtained during samplings.The gradual increase in water depth is accompanied by other environmental gradients such as stability of the substrate, which are usually determinant for the macrofauna distribution (Corbera and Cardell, 1995;Cunha et al., 1999).Accordingly with our results, the grain size is one of the most often reported factors in determining the distribution and composition of crustaceans (Robertson et al., 1989).Higher sediment diversity increases the benthic faunas due to a higher diversity of microhabitats (e.g.Simboura et al., 2000;Lourido et al., 2008).The Ensenada de San Simón has a predominance of muddy bottoms, thus limiting the crustacean biodiversity along the inlet.
Intertidal sediments colonized by Zostera noltii and Z. marina in San Simón showed low crustacean species numbers.Usually, seagrass meadows provide a complex habitat that can be colonized by many spe-Fig.6. -Canonical correspondence analysis ordination of environmental variables and sampled sites relative to the axes I and II for the Ensenada de San Simón.Eigenvalues (E) and variance (V) for species data and species-environment correlations are indicated.Ts, sediment temperature, Tb, temperature of bottom water, So, sorting coefficient, Kg, kurtosis, Sk, skewness, GR, gravel, VCS, very coarse sand, CS, coarse sand, MS, medium sand, FS, fine sand, VFS, Vvery fine sand, CSi, coarse silt, FSi, fine silt, C, clay, CO 3 , carbonate content, OM, organic matter content, Q 50 , median grain size.
cies (Somerfield et al., 2002), stabilizing the sediment and providing protection for potential preys.Many ecological studies have shown salinity to be the main factor controlling zonation in the structure of the benthic communities.However, some other authors have emphasized that this is not the main factor in brackish environments (Bazaïri et al., 2003).In our inlet the seagrasses were located in intertidal and/ or shallow subtidal areas subject to abrupt salinity changes.The salinity or other factors related to depth are the major limiting factors for the non-euryhaline species, as suggested by Junoy (1996).Nevertheless, as in Marques and Bellan-Santini (1987), high densities of Melita palmata and Microdeutopus gryllotalpa were recorded within the Zostera meadows.These species can be acting here as detritus feeders (Cunha et al., 1999).M. gryllotalpa has been recorded in other organic enriched environments (Drake et al., 1997).However, M. palmata showed a restricted distribution in San Simón, since it has been classically cited in bottoms down to 50 m depth (Lincoln, 1979;Ruffo, 1982;Hayward and Ryland, 1990).
The highest values of diversity and number of species were found in the stable subtidal sediments situated on the mouth of the inlet, while the lowest values were found in the mouth of the rivers Oitabén-Verdugo and Alvedosa and in the proximities of the harbours (sites 12, 24 and 29).These results confirmed observations made by Conradi et al. (1997), who found a clear differentiation between the amphipods populations living in outer sites of a bay and those living in the inner area.In fact, these authors showed that the amphipod communities reflect the physico-chemical conditions in the area (e.g.Melita palmata or Idotea baltica showing a high tolerance to large physicochemical fluctuations in San Simón).
Multivariate analysis showed three major crustacean assemblages in the Ensenada de San Simón.The group A assemblage, characterized by Melita palmata, Mycrodeutopus gryllotalpa and Idotea baltica, was settled in the intertidal area.Similar fauna was found in intertidal areas along European coasts, e.g. in tidal channels of the Arcachon Bay (Bachelet et al., 1996), the Ria de Foz (Junoy and Viéitez, 1990) and the Mira Estuary (Marques and Bellan-Santini, 1987), and in a coastal lagoon of the NW African coast (Bazaïri et al., 2003).In our case, Melita palmata was the dominant species, as in the communities described by Bachelet et al. (1996) and Bazaïri et al. (2003).Group B, characterized by Corophium cf.runcicorne and Iphinoe tenella, represented a facies of transition.Finally, on the subtidal bottoms a facies  of Harpinia crenulata, H. dellavallei and Tanaopsis graciloides (Group C) was found.This kind of continuous transition between assemblages is characteristic of estuaries and semi-enclosed coastal ecosystems (Bazaïri et al., 2003), and has been cited as evidence of the importance of quantitative studies in differentiating assemblages (García Muñoz et al., 2008).Similar fauna was found in the community described by Desroy and Retière (2001) in the subtidal bottoms of the Rance Basin (western English Channel), although Ampelisca tenuicornis and Microdeutopus versiculatus showed high dominances in that estuary.
The intertidal assemblage displayed some features in common with the Cardium edule-Scrobicularia community described by Thorson (1957).In subtidal areas, the assemblages displayed common features with the Abra alba community described by Petersen (1914).Therefore, the patterns of crustacean distribution are similar to those previously reported for other benthic taxa in the same area (Cacabelos et al., 2008a, b).In conclusion, the most important factor in determining distribution patterns of crustacean communities in the Ensenada de San Simón was the water depth, directly influencing the hydrodynamic conditions and sediment composition of the inlet.

Fig. 1 .
Fig. 1. -Location of the Ensenada de San Simón (Ria de Vigo) with position of the 29 sampling sites and spatial distribution of crustacean assemblages related to sediment grain size (M, mud, SM, sandy mud, MS, muddy sand, CS, coarse sand, VCS, very coarse sand).
expressed as ind.