Sewage Pollution and Extinction Risk: an Endangered Limpet As a Bioindicator?

Biodiversity and Conservation (2007) 16:377–397 Springer 2006 DOI 10.1007/s10531-005-3014-3

-1 Sewage pollution and extinction risk: an endangered limpet as a bioindicator?

F. ESPINOSA*, J.M. GUERRA-GARCIÁ and J.C. GARCIÁ-GO´MEZ Laboratorio de Biologıá Marina, Dpto. Fisiologıá y Zoologıá, Facultad de Biologıá, Universidad de Sevilla, Avda Reina Mercedes 6, 41012, Sevilla, Spain; *Author for correspondence (e-mail: [email protected]; phone: +34-954557100; fax: +34-954233480)

Received 6 January 2005; accepted in revised form 9 September 2005

Key words: Harbour, Monitoring, Patella ferruginea, Patellogastropods, Pollution, Sewage outfall, Siphonaria

Abstract. The mollusc Patella ferruginea, endemic to the Mediterranean, is the most endangered marine species on the list of the European Council Directive 92/43/EEC and it is under serious risk of extinction. In spite of the low abundances and restricted distribution of this limpet, important populations have been found in the harbour of Ceuta, north Africa. The main objective of the present study was to characterise, for the first time, the effects of sewage pollution on P. ferruginea and related limpet species, and to evaluate the potential value of these limpet assemblages as bioindicators, using univariate and multivariate analyses. Physicochemical parameters and limpets were sampled in nine stations located at 0, 1, 2, 4, 8, 16, 32, 64 and 128 m away from the discharge point of a sewage effluent in Ceuta harbour. The stations closer to the outfall (0, 1, 2, 4 and 8) were characterised by higher values of turbidity, phosphate and ammonia in the water column, and organic matter, faecal coliforms and faecal Streptococci in sediments. A total of six limpet species were found and studied (Patella ferruginea, P. caerulea, P. nigra, P. rustica, P. ulyssiponensis and Siphonaria pectinata); the number of limpet species increased with increasing distance from the outfall, while diversity and evenness reached the highest values at intermediate sites. Siphonaria pectinata and P. caerulea were the most resistant and abundant species, while P. ferruginea was the most sensitive species to sewage pollution, only found at stations from 32 to 128 m. The distribution of this endangered limpet seems mainly affected by the pollution gradient, and not by the competition with the remaining limpets. The results of this study should be taken into account in future programmes of management and conservation of P. ferruginea.

Introduction

The limpet Patella ferruginea Gmelin, 1791, endemic to the Mediterranean, is the most endangered marine species on the list of the European Council Directive 92/43/EEC on the conservation of Natural Habitats and of Wild Fauna and Flora, 1992 (Ramos 1998), and it is, presently, under serious risk of extinction (Laborel-Deguen and Laborel 1991; Templado and Moreno 1997). Although its relative abundance in Palaeolithic and Neolithic deposits indicates a former distribution in the Western Mediterranean Basin (East coast of Italy, Mediterranean France, Iberian Peninsula, Morocco, Tunisia and the Western Mediterranean islands), today its Mediterranean range has progressively con- tracted to a few restricted areas (Cretella et al. 1994; Templado 1996) probably 378 due to human pressure (Aversano 1986; Laborel-Deguen and Laborel 1991). Although the species is threatened with extinction, important populations have been found recently inside the harbour of Ceuta, Strait of Gibraltar (Guerra- Garcıá et al. 2004a, b); Ceuta harbour is environmentally unique, differing substantially from other conventional harbours, since it is located between two bays connected by a channel which increases the water movement and exchange and contributes to the maintenance of rich and diverse communities of marine invertebrates (Guerra-Garcıá and Garcıá-Go´mez 2005). Little is known about the biology and ecological preferences of P. ferruginea; this species has a very low growth and reproductive rate, reaching sexual maturation at 2– 3 years (Guerra-Garcıá et al. 2004a). It is a protandric species with oocytes of 142–170 lm (Unpublished data) and this limpet feed mainly on cyanobacteria and the algae Ralfsia spp. and Rissoella spp. In the harbour of Ceuta, P. ferruginea coexists with five other limpet species: Patella caerulea Linnaeus, 1758, Patella nigra da Costa, 1771, Patella rustica Linnaeus, 1758, Patella ulyssiponensis Gmelin, 1791 and Siphonaria pectinata (Linnaeus, 1758). Patella nigra is also an endangered limpet, however, unlike P. ferruginea, P. nigra is not suffering a clear regression and populations seem to be increasing (Templado et al. 2004). Recent morphological and molecular studies seems to indicate that P. nigra should be transferred to the genus Cymbula,asCymbula safiana Lamark, 1819 (see Ridgway et al. 1998; Koufopanou et al. 1999). Taking into account that there is a sewage outfall inside the harbour of Ceuta, the present study explores how this sewage effluent is affecting the limpet assemblages in general and Patella ferruginea in particular. Due to their ecological importance, as well as sedentary life, molluscs have assumed a major role in monitoring contaminants worldwide (Boening 1999; Feldstein et al. 2003). They are abundant, sedentary and easy to collect, which makes them ideal for biomonitoring (Bresler et al. 2003a, b). Furthermore, in comparison with other marine environments, access to intertidal ecosystems is usually easier, and these ecosystems are more amenable to management than open ocean and sublittoral benthic habitats. Additionally, the composition of sessile communities is particularly useful as baseline for ecological monitoring because such organisms are unable to avoid disturbances in the marine environment and thus, the composition of the community reflects their common history (Fa et al. 2002). Differences in physical and biological conditions are the main cause of variation in marine communities, both in time and space (Pain and Levin 1981; Dayton 1984; Sousa 1984), and, consequently, environmental gradients are important in determining the structure of assemblages (Bishop et al. 2002). Some of these gradients originate as anthropogenic perturbations. One of the most serious, and increasingly common, sources of disturbance in marine communities is the discharge of sewage effluents (Gerlach 1981; Gray 1982). These effluents are often discharged via outfalls into shallow subtidal habitats, and can result in significant effects on marine biota, involving changes in cytology and physiology at the individual level that are ultimately manifested 379 as alterations in the community structure (Littler and Murray 1975). There are few studies describing the effects of sewage on macro-benthic assemblages living on hard substrata (Terlizzi et al. 2002). Furthermore, according to Underwood (1997), some authors have analysed differences between an im- pacted location and a single control, with evident problems of pseudoreplica- tion and consequently confounding in the logical interpretation of results. Additionally, along the Mediterranean coasts, despite a large amount of domestic and industrial sewage discharged to the sea (UNEP 1989), there are few published accounts of the effects of sewage effluents on the macrobenthos (see Terlizzi et al. 2002). The objectives of the present study are to characterise the effect of a sewage effluent on density, size and assemblages of the patellid community inhabiting the rocky intertidal shores of the harbour of Ceuta, North Africa, with special attention to the endangered P. ferruginea, and to explore the value of these patellid molluscs as bioindicators of sewage pollution.

Material and methods

Study area

The harbour of Ceuta (3553¢ N, 518¢ W) is one of the most important in the Strait of Gibraltar, because of its situation relative to Europe and Africa (Figure 1). There are no river discharges directly into the harbour and, unlike other harbours in this region of the Mediterranean, industrial activity adjacent to the harbour of Ceuta is absent. Consequently, the contamination inside the harbour is derived from the sewage effluents of urban influence, antifouling paints and accidental oil spills during the loading and dumping involved in shipping operations. Although the sediments are moderately polluted in the harbour of Ceuta (Guerra-Garcıá et al. 2003), this harbour has a unique design with two opposing entrances and a channel, which increases water exchange maintaining high oxygen levels in the water column. Under these unusual conditions (polluted sediments but oxygen saturation), the communities of the harbour of Ceuta are characterised by rich and diverse macrobenthic communities. The number of species is unusually high when compared with other more conventional enclosed harbours of nearby areas which do not have two opposing entrances or a channel (see Guerra-Garcıá and Garcıá-Go´mez 2004a–c, 2005; in press). Inside the harbour there is a sewage outfall located close to a leisure resort. The study was conducted during summer (August 2002), which coincides with the maximum peak of tourist activity. Raw sewage is discharged onto the shore area through a pipeline with a rectangular section measuring 135 · 75 cm. Although the outfall is discharging throughout the whole year, it is especially 380

Figure 1. Map of the study area in the harbour of Ceuta, North Africa. Arrow indicates the sewage discharge point, and the gross continuous line the 128 m transect. Sampling sites from 0 to 128 m along the transect are indicated. active during the summer months, discharging every day. The outfall terminus is located in the intertidal area, composed of sandstone artiﬁcial rocks.

Sampling methods

A total of nine sites were selected for this study. The first station (Site 0) was the ‘outfall site’ and the remaining stations were located at 1, 2, 4, 8, 16, 32, 64 and 128 m away along a transect (see Figure 1). The presence of the sewage effluent involves a gradient, and sampling is usually conducted in spatial scale and regularly (May 1985; Foe and Knight 1987; Lopez-Gappa et al. 1990; Zmarzly et al. 1994); however, sampling at regular intervals from a disturbance using a single spatial scale is not necessarily appropriate (Bishop et al. 2002). We used herein a logarithmic scale (log2) to explore more carefully the area close to the outfall since communities inhabiting this area are the most affected by the sewage pollution.

Limpet sampling In each station, the density of limpets was measured by counting the number of specimens found in a 1 m wide transect located from the zero tidal level to the upper level of the intertidal in which limpets were present. At site 0, coincident with the outfall, no limpets were found. The size of each limpet was estimated by measuring the shell length (see Guerra-Garcıá et al. 2004a). The species 381 were identified ‘in situ’ when possible; if the species could not be properly visually identified due to epiphytic alga or fauna, the limpet was removed for identification in the laboratory.

Physico-chemical analysis In addition to the limpet study, several physico-chemical parameters were measured in the water column and sediments in each sampling stations. Water samples were taken in the nine stations. Free and total ammonia, nitrate, nitrite, phosphate and pH were measured using colorimetric methods based on analytical test Seachem, and dissolved oxygen using the kit D.O.Azoo. Turbidity was measured in nephelometric turbidity units (ntu) using a turbidimeter Hanna HI 93703. Sediments were also collected at the nine sites, at 4 m depth and 20 m away from the coast. Analysis were conducted using the finest fraction of the sediment (diameter less than 0.063 mm) (see Guerra-Garcıá and Garcıá-Go´mez 2005)). Sediment samples were conserved in sterile containers until their arrival at the laboratory where they were frozen. The organic content was analysed by ashing samples of sediment to 500 C for 6 h and re-weighing (mean values of three replicates of 2 g each) (see Estacio et al. 1997). Phosphate and nitrate were measured using UV visible spectrophotometry following Me´todos oficiales de ana´lisis para aguas residuales y suelos (1986). Hydrocarbons were measured by extraction and FT-IR spectrophotometry (see Estacio et al. 1997). The faecal coliforms were measured as the coliforms which fermented lactose in a medium M-FC with production of acid and gas at 44.5 C during 24 h. For measuring the faecal Streptococci, agar m Enterococcus at 35 C during 48 h was used. The results of microbiological analyses were expressed as colonial formers units (cfu) per gram of sediment.

Statistical analysis

The total number of species, the Shannon–Wiener diversity index (Shannon and Weaver 1963) and the Pielou’s evenness index (Pielou 1966) were calcu- lated for each station based on the abundances of patellid species. The affinities among stations were established through cluster analysis using the UPGMA method (unweight pair–group method using arithmetic averages) (Sneath and Sokal 1973), based on the Bray–Curtis similarity index for the limpet species matrix and on the Euclidean distance for the environmental matrix. Abundance data of the limpet species were double square root transformed so that the ensuing classification and ordination were not determined only by the most dominant species (Clarke and Green 1988); environmental data were transformed using log (x+1) (see Guerra-Garcıá and Garcıá-Go´mez 2001). In order to confirm the results of the cluster, Principal Component Analyses were used for the ordination of stations based on the physicochemical data measured in water and sediment, and a MDS (non-metric 382 multidimensional scaling) was used for biological data. To test the ordination, the stress coefficient of Kruskal was employed (Kruskal and Wish 1978). To explore the relationships among environmental measures and the patellid assemblages, a canonical correspondence analysis (CCA) was applied. Rela- tionships between multivariate biological structure and environmental variables were also examined using the BIO-ENV procedure (Clarke and Ainsworth 1993). Percentage of similarity analysis (SIMPER) (Clarke 1993) was used to determine the species involved in grouping of the different stations. Possible differences in limpet abundances and sizes between the groups defined by ordination and classification analyses based on physicochemical parameters, were tested using Kruskal–Wallis test, since normality (Shapiro– Wilk test) and/or homogeneity of variances (Levene test) were not passed, and consequently ANOVA parametric tests could not be carried out. Univariate and multivariate analyses were carried out using BMDP (Dixon 1983), PRIMER (Clarke and Gorley 2001) and the PC-ORD (McCune and Mefford 1997).

Results

Physicochemical parameters

Table 1 includes the values of physicochemical parameters measured in water and sediments. In the water column, a clear trend could be observed for nutrients. The concentrations of ammonia and phosphate were at least one order of magnitude higher in the stations located near the outfall than in the remaining stations further away. The highest values of turbidity and pH were also recorded near the sewage outfall. On the other hand, oxygen showed the lowest values in the first 16 m, while in stations located at 32, 64 and 128 m away from the outfall, the oxygen concentrations were higher, reaching 9.0 mg/l. With regard to the sediment, both faecal coliforms and Streptococci were more abundant in the stations closer to the outfall; the same trend was obtained for hydrocarbons (Table 1) and for organic matter (Figure 2). However, phosphate and nitrate did not show a clear pattern, and concentrations of phosphate were often higher in the stations more distant from the outfall. The PCA plot obtained using the physicochemical parameters of the water column reflected similar gradients to the PCA elaborated with the sediment variables, although the former showed more clearly the gradient produced by the outfall (Figure 3). The first axis of the PCA ordination using water column measures, explained 59.1% of the total variance and the percentage of variance explained by axis 1 of the PCA based on the sediment measures was 43.4%. The PCA analyses, consequently, confirmed the trend of a pollution gradient from the outfall point. Cluster analyses also supported this ordination and showed clearly that stations could be separated into two groups according to Table 1. Values obtained for variables measured in the water column and sediment at the nine sampling sites along the transect. n.a: data not available.

Sites Water column Sediment

Free ammonia Total ammonia Nitrate Nitrite Phosphate Oxygen pH Turbidity F. coliforms F. Streptococci Phosphate Nitrate Hydrocarbons (ppm) (ppm) (ppm) (ppm) (ppm) (mg/l) (ntu) (cfu/g) (cfu/g) (ppm) (ppm) (ppm)

0 2.50 3.50 0.2 0.10 3.5 6.0 8.2 35.42 4 16.94 268.36 1 2.50 3.50 1.0 0.50 3.5 6.5 8.2 15.06 9 134 5.35 7.13 76.67 2 0.50 3.00 0.2 0.10 2.5 5.0 8.7 10.69 3 119 4.89 20.37 78.82 4 0.15 2.00 0.2 0.10 2.0 7.0 8.0 5.19 18 230 5.20 3.19 119.78 8 0.15 2.00 0.2 0.10 3.0 5.0 7.9 2.19 2 108 9.24 6.35 85.69 16 0.01 0.15 0.2 0.05 1.0 5.0 8.1 0.85 0 15 8.38 8.59 23.51 32 0.01 0.15 0.2 0.05 0.2 9.0 8.1 1.10 0 15 17.02 17.92 32.64 64 0.01 0.15 0.2 0.05 0.2 9.0 8.1 0.86 0 7 14.53 14.26 52.70 128 0.01 0.15 0.2 0.05 0.2 9.0 7.9 0.88 1 n.a. 9.40 16.77 70.69 383 384

Figure 2. Percentage of organic matter measured in the nine sampling sites along the transect. Mean values and standard deviations are included.

Figure 3. PCA (left) and Cluster (right) plots, based on the physicochemical parameters measured in the water column and sediment. the physicochemical parameters both in the water column and sediments. One group (from now on referred to as near sites) was formed by the sites located in the ﬁrst 8 m of the transect, and the second group (from now on referred to as 385 further away sites) included the remaining sites (from 16 to 128 m) (Figure 3). These two groups of stations shown by the physicochemical data (near sites vs further away sites) were used in further comparisons between density and size of specimens, as well as for SIMPER analysis.

Biological data

The limpets Siphonaria pectinata (Basommatophora) and Patella caerulea (Patellogastropoda) were present in all the sampled sites, although the densities were significantly higher in the further away sites than in the near sites (S. pectinata, K = 5.4, p < 0.05; P. caerulea K = 5.3, p < 0.05). There were also significant differences in the size for these two species, and specimens located near the outfall were larger than the specimens located further away (Figure 4) [S. pectinata: K = 120.3, p < 0.001 (near sites: 2.52 ± 0.65 cm (mean ± SD), n =71; further away sites: 1.49 ± 0.34 cm, n = 458); P. caerulea: K = 10.92, p < 0.001 (near sites: 3.24 ± 1.21 cm, n = 44; further away sites: 2.63 ± 0.99 cm, n = 378)]. In the case of Patella nigra, the highest densities and sizes were found at the intermediate stations (4–32 m away from the outfall). Patella rustica, P. ulyssiponensis and P. ferruginea were absent in the stations closer to the outfall. The endangered mollusc Patella ferruginea was the most sensitive species being present further away from the outfall (from 32 to 128 m). The relative contribution in percentage of individuals of each limpet species is shown in Figure 5. Although there was an increase in the number of species at the further away sites, S. pectinata and P. caerulea were the species which most contributed to the total densities in all the sampled sites. Although the number of species increased from 1 to 128 m along the transect, Shannon diversity and Pielou evenness values were higher at intermediate stations (Figure 6). The multivariate analysis (Cluster and MDS) using limpet abundances also showed a separation between the near sites and the further away sites (Figure 7). According to the SIMPER, the total average similarity inside the near sites group (stations 1, 2, 4 and 8) was 72.05%, and inside the further away sites group (stations 16, 32, 64 and 128) was 87.37%. The total average dissimilarity between the two groups of stations was 39%, and the species which most contributed to the dissimilarity was P. rustica (Table 2).

Relationships between limpet assemblages and environmental variables

The results of the correlation analysis between the biotic and environmental similarity matrices indicated that the best correlation (0.96) occurred with the group of variables free ammonia, total ammonia, turbidity and organic matter 386

Figure 4. Number of individuals and size of each limpet species at the sampled stations along the transect. Mean values and standard deviations are included for size. in sediment, according to the BIO-ENV. Including extra variables the signifi- cance of the correlation decreased, although the values were also significant (Table 3). Taking into account that eight stations only were sampled, the maximum number of variables which we could included in the CCA analysis was seven; consequently, we selected the seven variables which, independently, showed the best correlation coefficients with the biotic data in the BIO-ENV analysis: turbidity, free ammonia, total ammonia, nitrite, phosphate, organic 387

Figure 4. Continued matter in sediment, and phosphate in sediment. The ﬁrst axis of the CCA analysis (Figure 8, Table 4) mainly correlated with the organic matter of the sediment, but also with the concentrations of total ammonia, turbidity and phosphate (both in water and sediment). Axis 1 separated Patella ferruginea and P. rustica from the remaining species. These two limpets were associated with the further away sites (32, 64 and 128 m), characterised by lower concentrations of ammonia, phosphate in water, organic matter in sediment and lower values of turbidity. On the other hand, the axis 2 separated the stations 1 388

Figure 5. The relative contribution in percentage of individuals of each limpet species at the sampling sites. See also Figure 4.

Figure 6. Species richness (S), Shannon–Wiener diversity (H¢) and Pielou evenness (J) of the patellid community in each station. and 2 -and the species Siphonaria pectinata and Patella caerulea-, from the stations 4, 8 and 16 -and the species P. nigra and P. ulyssiponensis-, based on the concentrations of free ammonia, which correlated signiﬁcantly with the axis 2.

Discussion

Physico-chemical data

The present study showed moderate concentrations of nutrients (mainly ammonia and phosphates) in the water column close to the outfall. These 389

Figure 7. Cluster and MDS analysis based on the abundance of patellid species.

Table 2. Average abundances of the species from sites located closer to the outfall (N) and further away (F).

Species Abund. N Abund. F Av. Dis. Ratio Dis. (%) Cum. Dis (%)

Patella rustica 0.75 18.00 9.13 2.18 23.40 23.40 Patella caerulea 11.00 94.00 7.77 2.49 19.90 43.30 Siphonaria pectinata 17.25 115.25 7.01 4.59 17.97 61.27 Patella ferruginea 0.00 5.50 6.75 1.64 17.29 78.57 Patella ulyssiponensis 0.50 2.50 4.23 1.13 10.83 89.39 Patella nigra 6.50 5.75 4.14 1.10 10.61 100.00 Species are listed in order of decreasing contribution to the average dissimilarity (Av. Dis.) between the two groups. The ratio indicates Dis/Standard deviation. The total average dissimilarity between groups is 39%. concentrations were considerably higher than those measured in other areas of the harbour of Ceuta and adjacent waters (see Guerra-Garcıá 2001), which indicates the urban influence of sewage effluent. The oxygen concentrations near the outfall (5–7 mg/l) were lower than in the sites 32, 64 and 128 m (9 mg/l). Several studies have recently pointed out the importance of oxygen in the water column on macrofaunal assemblages (Saı´z- Salinas 1997; Guerra-Garcıá and Garcıá-Go´mez 2005). Our data are in agreement with Littler and Murray (1975) that recorded values of 5.5 mg/l 390

Table 3. Summary of results from BIO-ENV analysis. k Best variable combination

1 Turbity 0.871 F. ammonia 0.870 T. ammonia 0.866 Nitrite 0.828 2 Turbidity 0.944 Turbidity 0.923 F. ammonia 0.893 T. ammonia 0.893 Phosphate T. ammonia T. ammonia Cu 3 T. ammonia 0.946 T. ammonia 0.929 F. ammonia 0.907 F. ammonia 0.907 Turbidity Phosphate T. ammonia T. ammonia O.M. Turbidity Turbidity O.M. 4 F. ammonia 0.960 T. ammonia 0.951 T. ammonia 0.946 F. ammonia 0.927 T. ammonia Nitrite Nitrate T. ammonia Turbidity Turbidity Turbidity Phosphate O.M. O.M. O.M. Turbidity 5 F. ammonia 0.956 T. ammonia 0.953 T. ammonia 0.945 F. ammonia 0.939 T. ammonia Nitrite Nitrate T. ammonia Phosphate Phosphate Phosphate Oxygen Turbidity Turbidity Turbidity Turbidity O.M. O.M. O.M. O.M. 6 F. ammonia 0.950 T. ammonia 0.944 T. ammonia 0.943 F. ammonia 0.937 T. ammonia Nitrite Nitrate T. ammonia Phosphate Phosphate Phosphate Phosphate Turbidity Turbidity Turbidity pH O.M. O.M. O.M. O.M. Phosphate (sed) Phosphate (sed) Phosphate (sed) Phosphate (sed) Combinations of variables, k at a time, giving the highest rank correlations between biotic and environmental similarity matrices are shown with the value of the weighted Spearman rank correlation coefficient; text in bold type indicate the best combination overall. in an outfall point located in San Clemente Island, California, and values of 7–8 mg/l, 30 m away. Mean values of turbidity measured inside Ceuta’s harbour are about 0.62 ntu (CEDEX 2000); however, the values measured in this study near the effluents reached 35 ntu. These high values of turbidity, due to the presence of solids and organic matter in suspension may have direct negative effects since these can clog the gills of marine organisms, such as limpet molluscs. In this sense, Marshall and McQuaid (1989) compared the survival rate of a pulmonate limpet (Siphonaria capensis) with a prosobranchia limpet (Patella granularis) in the presence of sand and low oxygen concentrations, and found a higher survival of the pulmonate limpet possibly due to the different breathing strategy. This fact could explain the presence of S. pectinata very close to the effluent outfall in the present study. Furthermore, Branch (1981) pointed out that siltation may be more important than salinity in excluding limpets from estuaries.

Biological data

There is a lack of studies dealing with the abundance of patellids in areas aﬀected by sewage disposal. Bishop et al. (2002) found higher densities of the 391

Figure 8. Graphic representation of the stations, limpet species and environmental measures with respect to the ﬁrst two axes of the canonical correspondence analysis (CCA).

Table 4. Summary results of the canonical correspondence analysis.

Axis 1 Axis 2 Axis 3

Eigenvalue 0.11 0.06 0.01 Species–environment correlation 0.99 0.99 0.99 Percentage of species variance 58.3 30.8 6.7 Correlation with environmental variables Free ammonia (ppm) – 0.72* – Total ammonia (ppm) 0.79* –– Nitrite (ppm) – – – Phosphate (ppm) 0.89** –– Turbidity (ntu) 0.76* –– OM sed. (%) 0.98*** –– Phosphate sed. (ppm) À 0.89** ––

*p < 0.05, **p < 0.01, ***p < 0.001. 392 limpet Patelloida latistrigata in sites further away from an outfall; Tablado et al. (1994) found significant lower densities of Siphonaria lessoni near an outfall point. In the present study, for the six limpet species recorded, the lowest abundances were found close to the outfall, and sites 1 and 2 were dominated by Siphonaria pectinata, which was the most resistant species to the sewage pollution. On the other hand,P. ferruginea was most sensitive to the effluent, and was absent in regions of low oxygen concentrations, and high concentrations of ammonia, phosphate and turbidity in the water column. This limpet is the most endangered marine invertebrate in the Western Mediterra- nean rocky shores. It has been traditionally considered a K-strategist, associated with clean waters, a long life cycle and a lower reproductive rate (see Guerra-Garcıá et al. 2004a). Although several important populations of P. ferruginea have been found recently in the harbour of Ceuta (Guerra-Garcıá et al. 2004a, b), the presence of this species on the Iberian Peninsula is ex- tremely reduced; the most recent cites of P. ferruginea in the Iberian Peninsula are reported by Garcıá-Go´mez (1983) who recorded several individuals in Algeciras Bay, and Moreno (1992), who found two specimens in Cabo de Gata. It is especially interesting that other species, such as P. nigra and P. ulyssiponensis were more abundant in intermediate sites. These higher abundances could be due to enrichment effects (sensu the Intermediate Disturbance and Dynamic Equilibrium Hypotheses) and should be considered at least as potential components of community structure through e.g. competitive effects. It is also interesting that P. rustica was the greatest contributor to dissimilarity. It is an upper-shore species, usually found well above the other limpet species studied, and may indicate the possible influence of dessication pressures along the vertical gradients. For the dominant species, S. pectinata and Patella caerulea, the specimens near the discharge point were significantly larger than the individuals located further away. A combination of several factors (such as the higher inputs of organic matter and the lower intraspecific and interspecific competition due to lower densities) could explain this pattern. High growth rates in areas affected by organic enrichment have been recorded for Patella vulgata (Hatton 1938; Fischer-Piette 1948; Lewis and Bowman 1975) and S. pectinata (Voss 1959). Bastida et al. (1971) measured a higher growth rate for S. lessoni inside a harbour area. Liu and Morton (1998) found larger specimens of Patelloida saccharina and Patelloida pygmaea in polluted areas in comparison to unpolluted sites, and Tablado et al. (1994) recorded higher growth rates in specimens of S. lessoni placed near sewage outfall than specimens placed in non-polluted areas. Multivariate analysis showed that turbidity, ammonia and phosphate in the water column, together with the organic matter in sediment, were the parameters that mainly explained the limpet assemblages. Recent studies focused on marine communities of Algeciras Bay, also located in the Strait of Gibraltar, have shown that turbidity is one of the factors negatively affecting the distribution of benthic assemblages (Sańchez-Moyano et al. 1998). 393

Eﬀects of sewage eﬄuents

The results of the present study, based on the limpet assemblages, showed a clear gradient of environmental perturbation from the sites near the sewage outfall to the further away sites. As pointed out by Pearson and Rosenberg (1978), the effects of a sewage effluent are most pronounced in the vicinity of the outlets and decrease progressively with increasing distance from the discharge points. Raffaelli and Hawkins (1996) pointed out that most of the small volumes of effluent discharge have ecological effects in the first 10 m away from the outfall. Littler and Murray (1975), in a study conducted in San Clement Island, Cali- fornia, indicated pollution effects inside the first 30 m from the effluent, and normal communities from 90 m away. Terlizzi et al. (2002) provided evidence that sites located about 100–300 m apart from an outfall located in Apulian coast, Italy, were similar to each other and they differed from the site located in close proximity to the sewage. Lopez-Gappa et al. (1993) found pollution effects on intertidal communities 50–100 m away from effluents from the cities of Necochea and Quequeń (Argentina). In the present study, natural conditions were completely re-established at 64 m from the discharge point.

Conclusions and environmental implications

The physicochemical parameters measured both in the water column and sediment showed a clear perturbation due to the presence of the sewage effluent inside the harbour. However, the effect on limpet assemblages is restricted, and from 32–64 to 128 m from the point of discharge ‘normal’ conditions are re-established. From a methodological point of view, the use of log2 can be useful to select and locate the sampling sites in this kind of study, focused on environmental gradients. Furthermore, limpet assemblages can be used as a tool to evaluate the environmental state of the coast, without the need of using the whole community which would be more time consuming and would require higher costs. In this sense, authorities and government institutions are increasingly demanding quick and effective environmental studies in coastal areas. This fact precludes developing series of data based on complex matrices of various taxonomic groups whose identifications requires great effort and time. A short- term spatial study dealing with only several easy-to-identify limpet species can yield similar results to those obtained with costly physico-chemical analysis. Although further studies are necessary to investigate the ecological preferences and responses to pollution of the endangered mollusc Patella ferruginea, this study represents a first approach, showing that this limpet is sensitive to sewage pollution, specially to the increase of turbidity, ammonia and phosphate, and the decrease of oxygen concentrations in the water column. These initial results should be taken into account in future studies of management 394 and conservation of P. ferruginea. Furthermore, the present work showed that the highest densities of P. ferruginea were measured at 32 and 64 m away from the outfall, coinciding with the highest densities of other limpets; therefore, this trend indicates that the main problem for P. ferruginea is the degree of pollution, and not interspecific competition with other limpets. This fact should be also taken into consideration in future programmes of reintroduction or resettling of this endangered limpet.

Acknowledgements

Special thanks are due to our colleagues Alexander Roi Gonza´lez, Cristina Huertas and Aurora Ruiz for helping during sampling. The present work was funded by the Asamblea de Ceuta, Autoridad Portuaria de Ceuta, and a PhD scholarship FPU AP-3556-2001 (to F. Espinosa) from the Ministry of Edu- cation and Culture of Spain.

References

Aversano F.R. 1986. Esperimento di insediamento artificiale di Patella ferruginea Gmelin, 1791 nelle acque del Golfo di Arzachena (Sardegna settentrionale). Boll. Malacol. 22: 169–170. Bastida R., Capezzani A. and Torti M.R. 1971. Fouling organisms in the port of Mar del Plata (Argentina). I. Siphonaria lessoni: ecological and biometric aspects. Mar. Biol. 10: 297–307. Bishop M.J., Underwood A.J. and Archambault P. 2002. Sewage and environmental impacts on rocky shores: necessity of identifying relevant spatial scales. Mar. Ecol. Prog. Ser. 236: 121–128. Boening D.W. 1999. An evaluation of bivalves as biomonitors of heavy metals pollution in marine waters. Environ. Monit. Assess. 55: 459–470. Branch G.M. 1981. The biology of limpets: physical factors, energy flow and ecological interac- tions. Oceanogr. Mar. Biol. Ann. Rev. 19: 235–380. Bresler V., Abelson A., Fishelson L., Feldstein T., Rosenfeld M. and Mokady O. 2003a. Marine molluscs in environmental monitoring. I. Cellular and molecular responses. Helgoland Mar. Res. 57: 157–165. Bresler V., Mokady O., Fishelson L., Feldstein T. and Abelson A. 2003b. Marine molluscs in environmental monitoring. II. Experimental exposure to selected pollutants. Helgoland Mar. Res. 57: 206–211. CEDEX 2000. Ana´lisis y medidas complementarias para la gestioń del material dragado en el muelle de cruceros turı´sticos adosado al Muelle de Espan˜ a. Informe especı´fico, Ministerio de Fomento, Madrid, pp. 1–52 Clarke K.R. 1993. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 18: 117–143. Clarke K.R. and Ainsworth M. 1993. A method of linking multivariate community structure to environmental variables. Mar. Ecol. Prog. Ser. 92: 205–219. Clarke K.R. and Green R.H. 1988. Statistical design and analysis for a ‘biological effects’ study. Mar. Ecol. Prog. Ser. 46: 213–226. Clarke K.R. and Gorley R.N. 2001. PRIMER (Plymouth Routines In Multivariate Ecological Research) v5: user manual/tutorial. PRIMER-E, Plymouth. Cretella M., Scillitani G., Toscano F., Turella P., Picariello O. and Cataudo A. 1994. Relationships between Patella ferruginea Gmelin, 1791 and the other Tyrrhenian species of Patella (Gastro- poda: Patellidae). J. Molluscan Stud. 60: 9–17. 395

Dayton P.K. 1984. Processes structuring some marine communities: are they general? In: Strong D.R.J., Simberloff D., Abelle L.G. and Thistle A.B. (eds), Ecological Communities: Conceptual Issues and the Evidence. Princeton University Press, Princeton, pp. 181–197. Dixon W.J. 1983. BMDP Statistical Software. University California Press, Berkeley. Estacio F.J., Garcıá-Adiego E.M., Fa D., Garcıá-Go´mez J.C., Daza J.L., Hortas F. and Go´mez- Ariza J.L. 1997. Ecological analysis in a polluted area of Algeciras Bay (Southern Spain): external ‘versus’ internal outfalls and environmental implications. Mar. Pollut. Bull. 34: 780–793. Fa D.A., Finlayson C., Garcıá-Adiego E., Sańchez-Moyano J.E. and Garcıá-Go´mez J.C. 2002. Influence of some environmental factors on the structure and distribution of the rocky shore macrobenthic communities in the Bay of Gibraltar: preliminary results. Almoraima 28: 73–88. Feldstein T., Kashman Y., Abelson A., Fishelson L., Mokady O., Bresler V. and Erel Y. 2003. Marine molluscs in environmental monitoring. III. Trace metals and organic pollutants in ani- mal tissue and sediments. Helgoland Mar. Res. 57: 212–219. Fischer-Piette E. 1948. Sur les e´lements de prosperite´des Patelles et sur leur spećificite´. J. Conchol. 84: 45–96. Foe C. and Knight A. 1987. Assessment of the biological impact of point-source discharges employing asiatic clams. Arch. Environ. Contam. Toxicol. 16: 39–51. Garcıá-Go´mez J.C. 1983. Estudio comparado de las tanatocenosis y biocenosis malacolo´gicas del Estrecho de Gibraltar y a´reas pro´ximas. Iberus 3: 75–90. Gerlach S.A. 1981. Marine Pollution. Diagnosis and Therapy. Springer-Verlag, Berlin. Gray J.S. 1982. Effects of pollutants on marine ecosystems. Netherlands J. Sea Res. 16: 424–443. Guerra-Garcıá J.M. 2001. Ana´lisis integrado de las perturbaciones antropogeńicas en sedimentos del Puerto de Ceuta. Efecto sobre las comunidades macrobentońicas e implicaciones ambien- tales. Ph.D. thesis. University of Sevilla, Sevilla, 346 pp. Guerra-Garcıá J.M. and Garcıá-Go´mez J.C. 2001. The spatial distribution of the Caprellidea (Crustacea: Amphipoda): a stress bioindicator in Ceuta (North Africa, Gibraltar area). PSZNI Mar. Ecol. 22: 357–367. Guerra-Garcıá J.M., Gonzalez-Vila F.J. and Garcia-Gomez J.C. 2003. Aliphatic hydrocarbon pollution and macrobenthic assemblages in Ceuta harbour: a multivariate approach. Mar. Ecol. Prog. Ser. 263: 127–138. Guerra-Garcıá J.M. and Garcıá-Go´mez J.C. 2004a. Soft bottom mollusc assemblages and pollution in a harbour with two opposing entrances. Estuarine Coast. Shelf Sci. 60: 273–283. Guerra-Garcıá J.M. and Garcıá-Go´mez J.C. 2004b. Crustacean assemblages and sediment pollution in an exceptional case study: a harbour with two opposing entrances. Crustaceana 77: 353– 370. Guerra-Garcıá J.M. and Garcıá-Go´mez J.C. 2004c. Polychaete assemblages and sediment pollution in a harbour with two opposing entrances. Helgoland Mar. Res. 58: 183–191. Guerra-Garcıá J.M., Corzo J., Espinosa F. and Garcıá-Go´mez J.C. 2004a. Assessing habitat use of the endangered marine mollusc Patella ferruginea (Gatropoda, Patellidae) in northern Africa: preliminary results and implications for conservation. Biol. Conserv. 116: 319–326. Guerra-Garcıá J.M., Corzo J., Espinosa F., Fa D. and Garcıá-Go´mez J.C. 2004b. Extinction risk and harbours as marine reserves? J. Mollus. Stud. 70: 116–118. Guerra-Garcıá J.M. and Garcıá-Go´mez J.C. 2005. Oxygen levels versus chemical pollutants: do they have similar influence on macrofaunal assemblages? A case study in a harbour with two opposing entrances Environ. Pollut. 135: 281–291. Guerra-Garcıá J.M. and Garcıá-Go´mez J.C. 2005. Assessing pollution levels in sediments of an exceptional harbour with two opposing entrances. Environmental implications. J. Environ. Manage. 77: 1–11. Hatton H. 1938. Essais de bionomie explicative sur quelques especes intercoidales d’algues et d’animaux. Annal. l’Inst. Oceanogr. Monaco 17: 241–348. Koufopanou V., Reid D.G., Ridgway S.A. and Thomas R.H. 1999. A molecular phylogeny of the patellid limpets (Gastropoda: Patellidae) and its implications for the origins of their antitropical distribution. Mol. Phylogenet. Evol. 11: 138–156. 396

Kruskal J.B. and Wish M. 1978. Multidimensional Scaling. Sage Publications, Beverley Hills, California. Laborel-Deguen F. and Laborel J. 1991. Statut de Patella ferruginea Gmelin en Mediterranee. In: Boudouresque C.F., Avon M. and Gravez V. (eds), Les spe` ces marines a` prote´ger en Me´di- terraneé. GIS Posidonie Publishers, Marseille, pp. 91–103. Lewis J.R. and Bowman R.S. 1975. Local habitat-induced variations in the population dynamics of Patella vulgata L. J. Exp. Mar. Biol. Ecol. 17: 165–203. Littler M.M. and Murray S.N. 1975. Impact of sewage on the distribution, abundance and community structure of rocky intertidal macro-organisms. Mar. Biol. 30: 277–291. Liu J.H. and Morton B. 1998. The impacts of pollution on the growth, reproduction and population structure of Hong Kong limpets. Mar. Pollut. Bull. 36: 152–158. Lopez-Gappa J.J., Tablado A. and Magaldi N.H. 1990. Influence of sewage pollution on a rocky intertidal community dominated by the mytilid Brachidontes rodriguezi. Mar. Ecol. Prog. Ser. 63: 163–175. Lopez-Gappa J.J., Tablado A. and Magaldi N.H. 1993. Seasonal changes in an intertidal community affected by sewage pollution. Environ. Pollut. 82: 157–165. Marshall D.J. and McQuaid C.D. 1989. The influence of respiratory responses on the tolerance to sand inundation of the limpets Patella granularis L. (Prosobranchia) and Siphonaria capensis Q. et G. (Pulmonata). J. Exp. Mar. Biol. Ecol. 128: 191–201. May V. 1985. Observations on algal floras close to two sewage outlets. Cunninghamia 1: 385–394. McCune B. and Mefford M.J. 1997. PC-ORD. Multivariate Analyses of Ecological Data. Version 3.0. Mjm Software Design, Gleneden Beach, U.S.A. Me´todos oficiales de ana´lisis para aguas residuales y suelos 1986. Tomo III. MAPA. Direccioń general de polı´tica alimentaria, Madrid. Moreno D. 1992. Presencia de Patella ferruginea Gmelin 1971 en el Cabo de Gata (Almerıá, SE Espan˜ a). Cuadernos de Investigacioń Biolo´gica, Bilbao 17: 71. Paine R.T. and Levin S.A. 1981. Intertidal landscapes: disturbances and the dynamics of pattern. Ecol. Monogr. 51: 145–178. Pearson T.H. and Rosenberg R. 1978. Macrobenthic succession in relation to organic enrichment and pollution of the marine environment. Oceanogr. Mar. Biol. Annl. Rev. 16: 229–311. Pielou E.C. 1966. The measurement of diversity in different types of biological collections. J. Theor. Biol. 13: 131–144. Raffaelli D. and Hawkins S.J. 1996. Intertidal Ecology. Chapman and Hall, London, pp. 356 Ramos M.A. 1998. Implementing the Habitats Directive for mollusc species in Spain. J. Conchol. Special Publicat. 2: 125–132. Ridgway S.A., Reid D.G., Taylor J.D., Branch G.M. and Hodgson A.N. 1998. A cladistic phylogeny of the family Patellidae (Mollusca: Gastropoda). Philos. Trans. Biol. Sci. 353: 1645–1671. Sańchez-Moyano J.E., Estacio F.J., Garcıá-Adiego E.M. and Garcıá-Go´mez J.C. 1998. Las pra- deras submarinas de la Bahıá de Algeciras. Evolucioń histo´rica y planes para su restauraciońy conservacioń. Almoraima 19: 173–180. Saı´z-Salinas J.I. 1997. Evaluation of adverse biological effects induced by pollution in the Bilbao Estuary (Spain). Environ. pollut. 96: 351–359. Shannon C.E. and Weaver W. 1963. The Matemathical Theory of Communication. University of Illinois Press, Urbana, Illinois, pp. 117 Sneath P.H.A. and Sokal R.R. 1973. Numerical Taxonomy. The Principles and Practice of Numerical Classification. WH Freeman and Company, San Francisco, pp. 573 Sousa W.P. 1984. Intertidal mosaics: patch size, propagule availability and spatial variable patterns of sucession. Ecology 65: 1918–1935. Tablado A., Lopez-Gappa J.J. and Magaldi N.H. 1994. Growth of the pulmonate limpet Siphonaria lessoni (Blainville) in a rocky intertidal area affected by sewage pollution. J. Exp. Mar. Biol. Ecol. 175: 211–226. Templado J. 1996. Patella ferruginea. In: Ramos M.A. (ed.), Inventario de las especies de invertebrados no artro´podos incluidas en los anejos de la Directiva 92/43/CEE del Consejo. ICONA, Spain, pp. 11–15. 397

Templado J. and Moreno D. 1997. La lapa ferrugıńea. Biolo´gica 6: 80–81. Templado J., Calvo M., Garvıá A., Luque A.A., Maldonado M. and Moro L. 2004. Guıáde invertebrados y peces marinos protegidos por la legislacioń nacional e internacional. Organismo Autońomo Parques Nacionales, Ministerio de Medio Ambiente, Madrid. Terlizzi A., Fraschetti S., Guidetti P. and Boero F. 2002. The effects of sewage discharge on shallow hard substrate sessile assemblages. Mar. Pollut. Bull. 44: 544–550. Underwood A.J. 1997. Experiments in Ecology: Their Logical Design and Interpretation using Analysis of Variance. Cambridge University Press, Cambridge. UNEP 1989. State of the Mediterranean marine environment. MAP Tech. Rep. Ser. 28: 34–39. Voss N.A. 1959. Studies on the pulmonate gastropod Siphonaria pectinata (Linnaeus) from the southeast coast of Florida. Bull. Mar. Sci. 9: 84–99. Zmarzly D.L., Stebbins T.D., Pasko D., Duggan R.M. and Barwick K.L. 1994. Spatial patterns and temporal succession in soft bottom macroinvertebrate assemblages surrounding an ocean outfall on the southern San Diego shelf: relation to anthropogenic and natural events. Mar. Biol. 118: 293–307.