Peracarida populations of hard substrate assemblages in ports of the NW Aegean Sea (eastern Mediterranean)

Abstract

This study deals with the structure of Peracarida populations in four ports in the NW Aegean Sea, Greece, and with the degree this structure is influenced by the particular biotic and abiotic conditions that prevail in the ports. Quantitative samples were taken during summer and winter in two successive years from artificial hard substrates and were analysed using common biocoenotic methods. The examination of approximately 81,250 specimens revealed the presence of 24 Peracarida species, the most dominant of which were Corophium acutum, Leptochelia savignyi and Elasmopus rapax. All species are very common and have been reported from many sites and assemblages in the N Aegean Sea. The ratios of certain Peracarida genera are discussed as possible indicators of environmental health that may be used in long-term biomonitoring programmes on the impact of pollution in harbours.

Introduction

The estimation of marine pollution has become a priority subject in recent years (Thomas 1993). Approaches to this problem have been manifold. In polluted ports, high abundance of species usually coincides with low diversity and vice versa. According to Bellan-Santini et al. (1994), this is particularly obvious for oligotrophic systems such as the Mediterranean Sea, where low abundance and high diversity are normally observed.

The biodiversity in ports is not yet fully understood, because knowledge of the structure and function of artificial hard substrate assemblages in polluted harbours is limited (Baxevanis and Chintiroglou 2000). This is due both to the difficulties involved in quantitative sampling, and to the complexity of the abiotic factors which affect the assemblages established on artificial substrates (Bellan and Pérès 1994). Furthermore, biodiversity seems to be increased in ports, which causes additional difficulties in drawing reliable conclusions (Baxevanis and Chintiroglou 2000). Generally, assemblages in ports (e.g. Marseilles) are considered to be assemblages of invertebrates in heavy polluted waters (Pérès and Picard 1964; Bellan-Santini 1968, 1969). Relevant information has been given by Bellan-Santini (1981, 1983), Bellan-Santini and Desrosiers (1976, 1977), Bellan-Santini et al. (1994), Zavodnik and Zavodnik (1978), Tursi et al. (1982, 1984), Hargrave and Thiel (1983) and Leewis and Waardenburg (1989). The characteristics of these assemblages are related to the environmental conditions as well as to the organisms’ adaptability (Baxevanis and Chintiroglou 2000; Damianidis and Chintiroglou 2000).

Owing to their ecological importance, numerical abundance and sensitivity to a number of pollutants, Peracarida have been known as sensitive environmental indicators (Hart and Fuller 1979) and thus have been used as primary biomonitors in various regions (Reish and Barnard 1979; Meijering 1991; Thomas 1993). The aim of this study was to provide information on the structure of Peracarida (Crustacea) populations of artificial hard substrate assemblages in four ports in the NW Aegean Sea, in order to set up a basis for future biomonitoring activities.

Methods

Study area

Four stations were selected in the NW Aegean Sea: ST1 = port of Thessaloniki, ST2 = fishing port of Kalamaria, ST3 = fishing port of N. Michaniona (largest in Greece) and ST4 = tourist marina of Porto Karras (Fig. 1). The choice was based mainly on the stations’ location in the bay of Thessaloniki and their distances from the main sources of pollution in Thermaikos Gulf. ST1 and ST2, located on the NW coasts of the gulf, are influenced by sewage produced by the neighbouring industrial and urban area of the city. ST3, located on the NE side of Thermaikos Gulf, receives only local urban waste. However, organic pollution is evident, originating from material of the estuaries system of the Axios, Loudias and Aliakmonas rivers, and also from fishing boat waste (Baxevanis and Chintiroglou 2000; Kamba et al. 2000). Lastly, ST4, located at Toroneos Gulf, is less affected by human activities, which are restricted to the summer tourist season: the station served as a reference point.

Fig. 1

Map showing sampling stations: ST1 port of Thessaloniki; ST2 EOT marina in Kalamaria; ST3 fisheries port in N Michaniona; ST4 Porto Karras Marina in Chalkidiki

There is much information on the concentration of chemicals (nitrates, nitrites, ammonium, phosphates, silicates) in Thermaikos gulf (Samanidou et al. 1987; Nikolaidis and Moustaka-Gouni 1990; Stergiou et al. 1997). Nitrates and ammonium, in particular, show significant seasonal fluctuations, with a maximum during winter and a minimum in summer. According to Samanidou et al. (1987), nutrients are highly concentrated near Thessaloniki harbour (ST1 and ST2) where the majority of urban sewage was discharged untreated until 1998.

The assemblages at ST1 and ST2 have been classified as hard substrate assemblages of heavily polluted ports (Pérès and Picard 1964; Bellan-Santini et al. 1994). The flora is dominated by Ulva rigida C.Ag, while the dominant faunal species are the ascidians Ciona intestinalis Linnaeus 1767 and Clavelina lepadiformis Muller 1776, and the polychaete Schistomeringos rudolphii Delle Chiaje 1828 (unpublished data). The assemblage at ST3 resembles that of photophilic algae and also that of semi-polluted ports (see Pérès and Picard 1964; Bellan-Santini 1969, 1981; Bellan et al. 1988). The presence of the sea anemone Aiptasiogeton pellucidus (Hollard, 1848), the high number of syllid polychaetes, and also the increased diversity of the flora [Bryopsis sp., Ceramium sp., Cladophora prolifera (Roth) Kutz., Dictyota dichotoma (Huds.) Hauck, Lomentaria sp.], indicate that the assemblage at ST3 represents an intermediate state between the assemblages at ST1 and ST2 on the one hand and at ST4 on the other. Finally, a typical facies of the photophilic algae assemblage (facies of Stypocaulon scoparium Kutzing) has been found at ST4 (see Pérès and Picard 1964; Bellan-Santini 1969; Taramelli-Rivosecchi 1969; Tsuchiya and Bellan-Santini 1989).

Sampling techniques

Sampling was carried out by scuba diving using the quadrate/sampler designed by Chintiroglou and Koukouras (1992). The surface area covered by the sampler was 400 cm². For each sampling site and period, three to five replicates were taken in order to achieve the minimum necessary area for a statistical investigation of hard substrates (Weinberg 1978; Stirn 1981; Chintiroglou and Koukouras 1992). The samplings were performed in winter (W) and summer (S) over two successive years (1994–1995). All samples were collected from the cement piers of the port facilities, which had been constructed more than 15 years previously. Samples were taken at a depth which corresponded to the average depth of the underwater cement columns of the piers (5 m at ST1, 2 m at ST2 and ST4, and 3 m at ST3). Samples were preserved in 10% formalin and transferred to the laboratory for further investigation. Specimens of the class Peracarida were counted and identified to species level.

Abiotic parameters

Physico-chemical factors (salinity, conductivity, water clarity, dissolved oxygen and temperature) were measured monthly at each sampling site. They were made using micro-electronic equipment [WTW salinity-conductivity-O₂ meter (Hanna Instruments, Vila do Conde, Portugal), Lovibond Checkit pH-meter (Hanna Instruments)]; water clarity was determined using the Secchi disc (KC Denmark, Research Equipment, Silkeborg, Denmark). Furthermore, the total hydro-dynamism was measured, based on the corrosion of plaster in the water (Kaandorp 1986; Kluijver and Leewis 1994).

Data analysis and statistics

Common biocoenotic methods were employed to analyse the data (Bellan-Santini 1981; Damianidis and Chintiroglou 2000). Hence, the numerical abundance (N) on a scale of 1 m², the mean dominance (D), and the frequency (f) were estimated. Also, Shannon-Weaver’s (H′) and Pielou’s Evenness (J′) were calculated on a log₂ basis (Daget 1979). The pollution indicator ratio of the abundance of certain genera of Peracarida, as proposed by Bellan-Santini (1981), was calculated in order to test its efficiency in the present case study.

The data obtained from different sampling stations were analysed using cluster and multidimensional scaling techniques, based on the Bray-Curtis similarity and log transformed numerical abundances, using the PRIMER package (see Clarke and Green 1988; Clarke and Warwick 1994). The significance of the multivariate results was assessed using the ANOSIM test. SIMPER analysis was applied in order to identify the percentage contribution of each species to the overall similarity within a site and to the dissimilarity between sites (Clarke 1993). The above analyses were performed to examine the similarity degree of samples in both space and time.

Results

Abiotic factors

Hydro-dynamism at the stations was measured by checking the corrosion of plaster. The stations ranked as follows: ST3(50.8%)>ST1(32.3%)>ST2(19.6%)>ST4(10%). Water clarity was always higher in winter than in summer and the stations ranked as follows: ST4>ST3>ST1>ST2.

At all stations, dissolved Ο₂ showed no significant fluctuations over time. The range was 6.5–8 ppm, with the exception of ST2: as a result of minor exchange and high temperatures during summer, the values at ST2 were lower (4–5 ppm). The pH values were almost the same at all stations in both sampling periods (pH≈8). Other abiotic factors showed a more or less marked seasonal pattern (Fig. 2).

Fig. 2

Temperature and salinity at the sampling stations over the study period

Composition of peracarid populations

A total of 81,250 peracarid individuals were counted and identified as 24 species belonging to four taxa (18 Amphipoda, two Isopoda, three Tanaidacea and one Cumacea). The most abundant species were Corophium acutum, Leptochelia savignyi and Elasmopus rapax, followed by Ericthonius brasiliensis, Corophium sextonae and Cymodoce sp. The rest of the species occurred only sporadically (Tables 1, 2).

Table 1 Peracarid species found at the sampling stations during 1994. f Frequency, D mean dominance. Rows in bold refer to the most dominant species

Table 2 Peracarid species found at the sampling stations during 1995. f Frequency, D mean dominance. Rows in bold refer to the most dominant species

Ratios between peracarid genera as pollution indicators

According to the relevant literature, species of the genera Corophium, Erichthonius and Leptochelia often dominate under polluted conditions. They are commonly referred to the characteristic of organic rich environments. In contrast, species of the genera Tanais and Elasmopus usually occur in clear waters. All of the above genera, which are characteristic inhabitants of harbours, proved to be unevenly distributed to our sampling stations. Bellan-Santini (1981) proposed that the ratio of the abundance (or dominance) of certain peracarid genera might represent a reliable indicator of pollution. Specifically, the author suggested that the ratio of the mean dominance of the genera Jassa and Hyale reflects the degree of pollution (the value is higher under increased pollution), at least for the western Mediterranean Sea. At our study sites, however, Hyale was not recorded at all and Jassa only sporadically. It seems that the composition of peracarid populations in the NW Aegean is quite different from that in the western Mediterranean. As Barnard (1958) stated: “different harbours may include different species, even when they are only few miles down-coast”. Instead of the ratio proposed by Bellan-Santini (1981), we calculated the ratio of the sum of the mean dominance (D) of the genera Corophium, Erichthonius and Leptochelia, and the sum of D of Elasmopus and Tanais, to test the hypothesis that this ratio reflects the water condition at the four sampling sites. The ratios were as follows: 1.4 at ST1; 7.93 at ST2; 5.66 at ST3; and 1.3 at ST4, suggesting increased pollution in the following order: ST2>ST3>ST1>ST4. This result, however, does not correspond to the actual situation. The value at ST1 was strongly influenced by the presence of E. rapax during 1995. The occurrence of this species is not clearly related to pollution: the species has been recorded from both ports and photophilic algae assemblages (Kocatas 1978). Its presence is related with low hydro-dynamism (Barnard 1958) which favours nontubicolous species. In 1994, all the Peracarida collected were tubicolous, demanding turbid water that carries along organic and mineral particles for tube construction. The ratio at ST4 was higher than expected, due to the high numbers of C. acutum in winter 1994, after a successful reproductive period leading to a temporary peak in population density (Gouvis 1988).

Diversity-abundance: similarity of peracarid populations

The largest number of species occurred at ST3 (N _S=15), followed by ST4 (N _S=11), while fewer species were recorded at ST1 and ST2 (N _S=5 and N _S=6, respectively). The number of species decreased from 1994 to 1995 at ST3 and ST4 but remained stable at ST1 and ST2. At ST1, ST2 and ST3, the number of individuals was generally higher in summer, while the opposite was observed at ST4.

Shannon (H′) values varied between 0.39 (ST2 W94) and 2.39 (ST3 W94), while evenness ranged from 0.15 (ST2 W94) to 0.96 (ST1 W94). The above indices showed a similar pattern and were strongly influenced by the increased numerical abundance of some species.

The affinity of samples is shown in Fig. 3. Both analyses [cluster and non-metric multidimensional scaling (mds)] indicate a separation of the samples into three main groups of about 50% similarity degree. Group A contains all samples from ST4, group B the 1994 winter samples from ST1, and group C the remaining samples from ST1 as well as all samples from ST2 and ST3. However, at about 60% similarity degree, five groups can be distinguished: group A comprising the 1995 summer samples from ST4, group B the remaining samples from ST4, group C the 1994 winter samples from ST1, group D the remaining samples from ST1 as well as all samples from ST2, and group E all samples from ST3. The stress value for the two-dimensional mds configuration is 0.06, indicating a very good ordination of samples (Clarke and Warwick 1994). The performance of a one-way ANOSIM test gave global R of 0.798 at a significance level of P<0.001. Thus, the four sites are separate. Further examination in order to localize the differences between the stations using a pair-wise test confirmed the separation of ST3 and ST4 (R ranges from 0.89–1, P=0.029), while ST1 proved to be related to ST2 (R 0.58, P=0.029).

Fig. 3

Dendogram using group-average linking on Bray-Curtis and multidimensional scaling analysis from standardised species numerical abundance data of 16 samples. W Winter, S summer; numbers refer to year

Group B showed an average similarity of 34.9%. According to SIMPER analysis, two species alone (L. savignyi and Tanais dulongi) were responsible for 60% of the average similarity and four species (the above plus C. acutum and Amphithoe ramondi) for 90%. Group D reached an average similarity of 29.4%. E. rapax and L. savignyi alone covered 60% of this value while four species (the above plus C. acutum and E. brasiliensis) covered 90%. Group E reached an average similarity of 44. 9%: C. acutum and L. savignyi were responsible for 60% of this value, whereas four species (the above plus E. rapax and C. sextonae) covered 90%. For groups A and C, relevant results could not be obtained, because these groups contained only one sample each.

As regards the divergence among groups, we found that group A had an average dissimilarity percentage of 75.2% with group B, 88.4% with group C, 97.8% with group D and 96.2% with group E. The species C. acutum, L. savignyi, T. dulongi and A. ramondi contributed 90% of the dissimilarity percentage between groups A and B; T. dulongi, L. savignyi and E. rapax between groups A and C; L. savignyi, E. rapax, C. acutum, E. brasiliensis and T. dulongi between groups A and D; and C. acutum, L. savignyi, E. rapax, C. sextonae and T. dulongi between groups A and E. Group B had an average dissimilarity percentage of 96.6% with group C, 82.2% with group D and 65.4% with group E. C. acutum, L. savignyi, T. dulongi and A. ramondi were responsible for 90% of the dissimilarity with group C; L. savignyi, C. acutum, E. rapax, E. brasiliensis and T. dulongi with group D; and C. acutum, L. savignyi, E. rapax, C. sextonae, T. dulongi and E. brasiliensis with group E. Group C had an average dissimilarity of 97.1% with group D and 96.3% with group E. L. savignyi, E. rapax, C. acutum and E. brasiliensis accounted for 90% of the above dissimilarity with group D; and C. acutum, L. savignyi, E. rapax and C. sextonae with group E. Finally, groups D and E showed an average dissimilarity of 68.9%. L. savignyi, C. acutum, E. rapax, E. brasiliensis and C. sextonae contributed 90% of this value. It is quite clear that only a few species are important in characterizing or differentiating sites.

Discussion

All the species found during this study had been reported as common fouling organisms and also as members of sublittoral hard substrate assemblages, having a wide geographical range across the Mediterranean Sea (Bellan-Santini 1969; Stephanidou and Voultsiadou-Koukoura 1995). However, some of them show ecological preferences in some Mediterranean regions which differ from those reported in the literature (e.g. Saldanha 1974; Bellan-Santini and Desrosiers 1977; Kocatas 1978; Bellan-Santini 1981; Bellan Santini et al. 1982, 1994; Hong 1983; Topaloglou and Kihara 1993; Lantzouni et al. 1998; Baxevanis and Chintiroglou 2000). For instance, the main habitats of C. acutum are sponges and the benthic assemblages in ports (Isaac et al. 1994). The typical representatives of the genus Corophium in the assemblages of photophilic algae are Corophium volutator and C. acherusicum, as reported by Kocatas (1978), Saldanha (1974) and Topaloglou and Kihara (1993). The genus Corophium has a wide ecological distribution (Ruffo 1998), favoured by the presence of organic particles which are used for tube-building and as food (typical filter feeders). Ericthonius brasiliensis was reported to live in the polluted waters of Turkish ports and in ports of Los Angeles, USA (Barnard 1958; Kocatas 1978). Elasmopus rapax was reported as a fouling organism (Ruffo 1998), but another species of the same genus, E. poecillimanus, is characteristic of clear waters (Bellan-Santini 1981; Tsuchiya and Bellan-Santini 1989). T. cavolinii is a detritivorous species which is very common in ports and in clear water conditions (Tsuchiya and Bellan-Santini 1989). This species was present in only five out of 16 samples. Its presence at ST4 (D=37.42) was indicative of the water conditions, whereas its presence at ST2 was very limited (D<0.01). Leptochelia savingyi is a species characteristic of environments rich in organic matter: it can grow fast under eutrophic conditions (Tsuchiya and Bellan-Santini 1989). This species has also often been reported as a member of assemblages of soft photophilic algae, sponges and Posidonia oceanica meadows (e.g. Bellan-Santini 1969; Kocatas 1978; Isaac et al. 1994). Cymodoce sp. was significantly present only at ST1 and ST2, a fact which is probably related to organic pollution. These facts imply that there are different species of Peracarida in different ports. Therefore, their composition as a pollution indicator must be treated separately in each case. An example for this is the port of Marseilles, where Bellan-Santini (1981) and Desrosiers et al. (1982) reported Jassa falcata as a pollution indicator, a species absent from the NW Aegean Sea.

We generally agree with Bellan-Santini (1981) that the ratio of certain peracarid genera can be a reliable indicator of environmental health. This presupposes, however, that

• 1. There is both a pollution indicator genus and a clear water indicator with a wide geographical range (or at least a local replacement of such genera by others with similar ecological demands); and that

• 2. Long-term data are available which allow the influence of biotic interactions to be determined.

In any case, the data must be explained with caution because the biological cycle of most of the species is not fully understood and a population outbreak of one of them would alter the results.

The feeding habits of Peracarida play a decisive role in their distribution in benthic assemblages, because the animals’ migratory activities are quite limited (Barnard 1958, 1963). The dominant species at our study sites were tube-dwellers. This behaviour is usually connected with filter feeding or feeding on detritus (Isaac et al. 1994), and migration is largely restricted to reproductive purposes (Barnard 1958, 1963). The environment of semi-polluted and polluted harbours favours the dominance of such species, due to the accumulation of organic material (Tsuchiya and Bellan-Santini 1989).

The diversity values measured in the present study are similar to those reported from other Mediterranean regions (e.g. Marseilles). According to Bellan-Santini (1981) and Desrosiers et al. (1982), the Η′ index ranged from 2.4 to 4.6 in non-polluted, and from 0.16 to 1.96 in polluted areas. The amphipod fauna of Mytilus galloprovincialis assemblages in NE Thermaikos Gulf showed a diversity (H′) close to 1.0 (Lantzouni et al. 1998). Variations in diversity indices (H′ and J′) seem to be affiliated to various biotic and abiotic factors. For example, the numerical abundance of Peracarida had its maximum in summer (with the exception of ST4). Lantzouni et al. (1998) presented similar findings for M. galloprovincialis assemblages, which were attributed to the reproductive efforts of certain species. Additionally, the variability of microhabitats of certain biotopes may also increase the diversity (Tsuchiya and Nishihira 1986; Damianidis and Chintiroglou 2000).

The multivariate analysis shows a clear ranking of the stations with respect to their pollution status. ST1 and ST2 share some common features, and are both exposed to heavy pollution. This is strongly supported by Shannon (H′) and Evenness (J′) indices and also by the number of species (N _S). On the other hand, ST3 and even more ST4 differ from ST1 and ST2, particularly with respect to the number of species found. This is probably due to the stations’ greater distance from the main sources of pollution in Thermaikos Gulf. The clustering of the samples did not reveal any pattern with respect to the period of sampling (clustering of summer and winter samples). The assemblages at all stations were dominated by a few species, resulting in low biodiversity (Dahl and Dahl 2002). An area is usually considered as heavily affected by pollution when the community shows no seasonal pattern or variations in its normal faunal composition. In such cases, some species are very abundant, a fact that indicates some kind of disturbance in the area (Gray and Mirza 1979). Methods such as ΑBC (abundance/biomass comparison) by Warwick (1986) or k-dominance (Clarke 1990) are not always applicable to organisms such as Peracarida whose biomass is very low (Clarke and Warwick 1994). Many authors consider crustaceans to be excellent objects for biomonitoring studies (Bellan-Santini 1981; Wenner 1988; Thomas 1993). The present study supports this opinion, as it has demonstrated that peracarid communities generally respond to water conditions, and that their composition can play an important role as a pollution indicator. However, for biomonitoring ports of different historical background and different use, methods have to be adjusted adequately for each special case, on the basis of long-term studies.