Introduction and aims

Lagoonal sediments may be polluted with heavy metals owing to natural processes such as weathering and erosion of adjacent materials, and/or as the result of human activity such as ore and metal processing, leaching from dumps, decantation ponds, fertilizer leaching, atmospheric deposition, animal excretion and urban sewage outflows (Carignan and Nriagu 1985; Förstner 1983; Förstner and Wittmann 1979; Galán 2003; Shahidul and Tanaka 2004). Evidences for this recent increase comes from chronological reconstructions of trace metal depositions in dated lake sediments (Nriagu et al. 1982; Schmidt et al. 2004).

Concerning heavy metal pollution, several studies have stressed the relevance of geochemical partitioning on lagoon sediments to understand metal mobilization, associations, availability, etc. Pioneer studies like those of Förstner and Wittmann (1979) and Salomons and Förstner (1984) provide toxic metals’ concentration values (μg/g) for supposedly unpolluted soils and sediments from the Rhine basin. Carignan and Nriagu (1985), Förstner (1983), Boudreau (1999) and Boyle (2001) pointed out four main factors, i.e. pH, redox conditions, organic complex and salinity, which influence the mobility and bioavailability of sediment-bound metals.

Studies carried out on polluted rivers and lakes in several countries (Dinelli et al. 1997; Rasmussen et al. 1998; Soares et al. 1999; Bellucci et al. 2002; Da Silva et al. 2002) show that heavy metals resulting from anthropogenic pollution are either associated with organic matter present in the fine fraction of the sediments, or adsorbed on Fe/Mn oxy-hydroxides, or precipitated as hydroxides, sulphides and carbonates (Manceau et al. 1992; Wen and Allen 1999; Boyle 2001; Bilali et al. 2002; Nikolaidis et al. 2004). Clay minerals have a high capacity for sorbing heavy metals, in particular, Cd, Cu, Pb, Zn and Ni on illite and smectites (Brigatti et al. 1994; Alietti et al. 1995; Galán 2000).

Because of coastal lagoons are the most productive ecosystems in the biosphere and that human activity has given rise to a variety of impacts on these water bodies with the consequent disruption of the natural equilibrium (Shahidul and Tanaka 2004), a better knowledge of the content of pollutants in sediments, their possible sources, and the effect they have on ecosystems are necessary to formulate possible measures for prevention.

This work looks on the environmental problems affecting Nador Lagoon, which is a place of great interest, given the diversity of the aquatic life found there, the fact that it is an area with important economic and social development owing to the fishing, iron and steel and other metallurgical industries located on its shores, in addition to the region’s tourism growth in recent years (Mahjoubi 1991; Hamoumi and Irzi 1998; Hamoumi and Benssaou 2001; Chalouan et al. 2001). It is for these reasons that the possible degradation of the area is of concern to its inhabitants, who depend on it for their livelihoods, and to companies and public institutions, which worry about the decline this would mean in economic investment as well as the resulting social problems this would imply.

The specific objectives of this work were (a) the determination and spatial distribution of trace element in sediments of the lagoon and possible tributary streams, (b) the sediment mineralogical characterization in order to identify the different mineral phases to which they are associated, (c) an approach to the real pollution risk of some toxic elements determining their bioavailability and chemical speciation and (d) the possible pollution origins (geogenic and/or anthropogenic)

General features of the study area

Nador Lagoon is situated on the western Mediterranean coast, between Cape Tres Puntas in Melilla (Spain) and the Moroccan district of Kariat Arekman, on the northern tip of the Bou Areg Plain. Its coordinates are 2°45′W–2°55′W longitude and 35°05′N–35°14′N latitude. Besides being the only lagoon on the Mediterranean coast of Morocco, it is also the largest in extension in the country, with a surface area of approximately 115 km2 (Fig. 1).

Fig. 1
figure 1

Location of the study area

Full size image

Nador Lagoon is located in the Neogene basin of Melilla–Nador, the southern edge of which is formed by a succession of mountain chains belonging to the Rif Mountain Range, which may be of direct influence through its contribution of sediments to the lagoon. In the Rif Mountain Range a series of ultramafic massifs of “alpine type”, essentially made up of lherzolites, are cropping out (Gervilla et al. 1988; Gervilla and Leblanc 1990; Alouan 2001). The multiphase tectonic activity affecting this coast during Neogene and Quaternary times was accompanied by a calc-alkaline vulcanism of shoshonitic affinities (Azdimoussa and Bourgois 1993). The Melilla–Nador Neogene basin is a subsident basin which was individualized in the Upper Miocene after the formation of the outer layers of the Rif range. The coastal strip is characterized by a narrow carbonated reef platform which becomes a marly-diatomitic basin laterally towards the south.

The lagoon complex is made up of a number of differentiated zones (Irzi et al. 1997, 1998; Mahjoubi 1991; Irzi 1987, 2002):

  • The continental fringe (Bou Areg Plain) is characterized by salt marshes, several intermittent streams, or wadis (oued), some of which presently serve as sewage outflows for urban areas upstream, flow into the marshes.

  • Nador Lagoon, still called “Sebkha de Bou Areg” or “Mar Chica”, is a restricted lagoon with a maximum depth of 8 m. It has a semi-elliptic shape and a surface area of 115 km2, measuring 25 km at its longest point and 7.5 km at its widest.

  • The coastal barrier is oriented NW–SE and at no point is over 20 m above sea level. It is 25 km in length and 300–400 m in width, reaching 2 km in its south-easternmost part. There is an entrance channel (the bokhana), which was recently reinforced with two dikes to ensure connection with the Mediterranean Sea, and remains of earlier openings with well-developed internal deltas (Fig 1).

Nador Lagoon is fed by groundwater from the Bou Areg Plain water table, and by intermittent flows from inland areas through wadis, among which the Selouane Oued is the most notable (large amounts of raw sewage enters the lagoon through it), as well as by irrigation canals coming from Nador, Beni Ensar and Kariat Arekman (Fig. 2).

Fig. 2
figure 2

Main watercourses feeding the Nador Lagoon and zone subdivision

Full size image

The presence of a multitude of small, torrential wadis which cut through the massifs of Beni Said, Gourougou and the base of Kebdana is common. Owing to percolation, evaporation and the small amount of water that flows through them, these wadis contribute little water to the lagoon (Inani 1995).

The most important socio-economic activities in the area that are potential sources of pollution are related to the (a) agricultural sector (over 60% of the population), dedicated to intensive cereal farming in the southeast, (b) industrial sector, located mainly in two areas, Selouane (where the largest iron and steel complex in the country is found) and Nador (with its neighbouring metallurgic industrial area), (c) fishing sector, with a fleet of more than 200 ships, several ports and considerable industrial activity related to this sector (fish farming, canning industries) and (d) tourism sector, thriving because of the natural attraction of the district—mountains in Gourougou, “Mar Chica” beaches and spa resorts in Kariat Arekman—as well as its culture—handicrafts, gastronomy, etc. (Quaranta et al. 2004) (Fig. 3).

Fig. 3
figure 3

Most important activities and potential sources of pollution in the Nador Lagoon

Full size image

Materials and methods

Materials

Two sampling campaigns were performed, one in summer and the other in winter. Samples were taken randomly on the perimeter of the lagoon, more densely in those areas where there was a greater possibility for natural and/or anthropogenic pollutants to be found, such as the mouths of wadis and urban and industrial areas. Lastly, sampling transects perpendicular to the longitudinal axis of the lagoon were made in order to determine the distribution of trace elements in the centre of the lagoon.

In the summer of year 2002, a total of 25 lagoonal samples were taken at 0–5 cm (“level 1”) at 25 sampling points. In order to determine the vertical evolution of pollution and possible pollutant sources, a more detailed sampling was carried out in winter of 2003 (Fig. 4). In particular, lagoonal samples were taken at three levels of depth (levels 0, 1 and 2). In order to determine possible trace element sources, stream samples in the proximity of the lagoon (D-I upstream and D-II at the mouth) and some representative soil samples were taken (Table 1).

Fig. 4
figure 4

Location of sampling points of lagoon sediments (C), stream sediments (D), soils, (B) and parent rock (E)

Full size image
Table 1 Sampling and materials studied
Full size table

Methodology

The pH for samples from the summer and winter campaigns was determined with a solid/water ratio of 1:2.5, i.e., 10 g of sediment sample was mixed with 25 ml of distilled water and shaken for 10 min in a magnetic agitator. The sample was left to stand for 30 min before the pH measurement, which was carried out using a previously calibrated Crison-pH-2002 pH meter.

Chemical analyses of major, minor and trace elements were performed on the fraction < 2 mm (the percentage in the whole sample determined by sieve), by Instrumental Neutron Activation (INAA) and Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) at Activation Laboratories Ltd. (1428 Sandhill Drive; Ancaster, Ontario, Canada). Detection limits for the determined elements and the method of analysis used for each element are given in Table 2.

Table 2 Method of analysis and detection limit in ppm, except where indicated
Full size table

The geochemical baseline of Nador Lagoon was determined by statistical analysis of the data obtained in the study of lagoonal sediments. The median, percentile 90 and percentile 95 were used as parameters. The median was used as the baseline, as it is a much more robust indicator than the mean for data populations without a normal distribution. Moreover, the systematic use of this parameter as representative of a population of geochemical data from natural systems is recommended (Rock 1988), given that it is not influenced by extreme values. The percentiles 90 and 95 have been used as standard risk values in Spain. This analysis was performed for As, Cd, Co, Cr, Cu, Ni, Pb and Zn, elements considered to be potentially toxic by the US Environmental Protection Agency (EPA).

According to the results obtained by chemical analysis of trace elements, the samples with the greatest concentrations, collected in both summer and winter, were selected for mineralogical characterization. This was determined by X-Ray Diffraction (XRD) with a Philips PW-1710 device with automatic slit, using the randon powder method on a flat sample slide. Diagrams from 3° to 60° of 2θ angle were obtained. The fraction < 2 μm was studied by oriented aggregate (OA) using different treatments: solvation with ethylene glycol and with dimethyl sulphoxide, and heating at 550° for 2 h. The interval from 3° to 30° was explored. Percentage estimates for the identified minerals were made for the whole sample and for the fraction < 2 μm, using the reflecting powers proposed by different authors and adjusted for automatic grid (Schultz 1964; Barahona 1974; Galán and Martín Vivaldi 1973; Dinelli and Tateo 2001).

On selected anomalous samples (values of some trace elements higher than percentile 90 or 95) metal bioavailability was assessed by extraction of the sample with EDTA 0.05 M (Quevauville et al. 1998) and metal mobility using the Tessier procedure (Tessier et al. 1979).

A correlation analysis between chemical and mineralogical data was carried out in order to assess possible associations between them. The index of geoaccumulation (Igeo) introduced by Muller (1981) was used to measure heavy metal pollution in sediments: Igeo = log 2 (C n /1.5 B n ), where C n is the measured concentration of the element “n” in unpolluted sediments. B n is the average concentration of element n in the background used by Muller (1981). The factor 1.5 is used because of possible variations of the background data due to lithogenic effects. Samples were divided into five groups: unpolluted (Igeo < 1), very low polluted (1 < Igeo < 2), low polluted (2 < Igeo < 3), moderately polluted (3 < Igeo < 4), highly polluted (4 < Igeo < 5) and very highly polluted (Igeo > 5). This index has been used to assess metal enrichment of riverine sediments (Herr and Gray 1997).

Results and discussion

Chemical characterization

For lagoonal sediments, the percentage of fraction < 2 mm is high (>60%) in most of the samples, and pH values (∼8) are quite homogeneous in the whole lagoon, in samples collected in both summer and winter, and for the different levels at which samples were taken. This pH homogeneity implies that trace element mobility must be conditioned by other factors such as the presence of clay minerals and Fe–Mn oxyhydroxides, or by the solid/liquid ratio effect (Dinelli et al. 1997; Bordas and Bourg 2000; Howari and Banat 2000).

For the summer sampling, results of chemical analysis for major elements (data not shown) present low content in aluminium (<10%) and calcium (<6%) in almost all the samples from the northern half of the lagoon. In the south, Ca percentages are slightly higher (8–12%), and on occasions over 25%. The highest Fe contents correspond to samples 5SC (44.8%) and 11SC (10.3%). The high sulphur percentage (1.3 and 2.6%) in samples from the northern zone should be pointed out. Trace element contents are not significant. As values oscillate between 5 and 16 ppm, the highest being from the central part of the lagoon, and Cd content is lower than 1 ppm. Sample 5SC stands out in that it shows very high concentrations of As (84.5 ppm), Cd (6.2 ppm), Co (120 ppm), Cu (466 ppm), Mn (5,141 ppm), Pb (416 ppm) and Zn (1,190 ppm).

The data obtained in the winter sampling have allowed a zone breakdown to be made (Fig. 2) according to major, minor and trace element contents. The highest aluminium contents are found between Beni Ensar and the city of Nador (zones A and B), but in no case is this greater than 15%. In addition, the sulphur is concentrated in the former zone, reaching 2.14% in sample 48°C. This sulphur concentration may be due to organic matter, given that this zone is quite confined. In zone B, iron is the indicator of geochemical difference with a higher percentage than in the rest of the lagoon. Ca content is in the order of ∼20% in samples taken from the zone C, and ∼15% in the SE area of the lagoon (zone D).

The distribution of trace elements at level 0, confirm that the aforementioned zones are maintained or are even more differentiated (Table 3). Between Beni Ensar and Atelouane (zone A), the anomalies encountered are Cd, Co, Cr and Pb. Sample 5SC is once again an exception giving very high concentrations of As (76 ppm), Cd (2 ppm), Co (80 ppm), Cu (384 ppm), Mn (2,520 ppm), Pb (297 ppm) and Zn (821 ppm). The trace element content in this zone is higher than in the rest of the lagoon.

Table 3 Trace element contents and statistical values of lagoon sediments samples calculated at level 0. February 2003
Full size table

In the surroundings of the city of Nador (zone B), the As content is slightly lower than that found in the confined area of NW, although on occasions is over 20 ppm in samples taken from the mouth of the Bouroug wadi (11SC, 12SC and 13SC). The anomalies present in this zone correspond to Mn, Cu and Zn. The coastal barrier and the Kebdana channel (zone C) shows moderately concentrations at specific points of Cd (2.7 ppm), Cr (120 ppm) and Ni (62 ppm) in sample 16SC, and of Cd (2.4 ppm) in sample 30SC. SE of the lagoon (zone D) there are no outstanding values except sample 20SC (2.1 ppm Cd).

Trace element values for level 1 show a similar distribution, but with lower concentrations than in level 0, except samples 46(0–5)C, 47(0–5)C and 48(0–5)C where As is over 30 ppm (Table 4).

Table 4 Trace element contents and statistical values of lagoon sediments samples calculated at level 1. February 2003
Full size table

The level 2 samples were taken from the area designated as zone C. In general, the fine fraction and pH show a slight increase with regard to what has been described for levels 0 and 1. It is to be noted the arsenic values which are above 15 ppm in all the samples, reaching 37.6 ppm in 31(10–14)C.

The statistical study has also allowed the geochemical baseline of the sediments in the lagoon to be defined Table 5). In general, Nador Lagoon does not present important heavy metal or arsenic anomalies; in fact the values determined as geochemical baseline are lower than those found by other authors (Salomons and Förstner 1984;Wright and Mason 1999; Salminen and Gregorauskiene 2000).

Table 5 Statistical values for the samplings of July 2002 and February 2003
Full size table

According to geochemical baseline, in the summer 2002, only sample 5SC can be considered polluted (>percentile 90 value). Between Beni Ensar and the city of Nador (zones A and B) only isolated samples were found to be above the percentile 90 value. Arsenic in sample 31SC from zone C, and Cr and Ni in sample 25SC from zone D are also above this percentile. In general, anomalous values for studied samples from the lagoon are very low.

In winter 2003, As, Cd, Co, Cr, Pb and Zn values are above the median in zone A. The same occurs with Zn, Cu and Mn in the area around the city of Nador (zone B). In both zones values close to percentile 90 predominate and several are even above it. Between Selouane and Arekmane (zones C and D) only three samples are found to be above the percentile 95 for Cd. Values obtained from the majority of samples from these zones are below the median (Tables 3, 4).

In summary, values for trace elements in winter are slightly higher than in summer and with a very similar spatial distribution (Tables 3, 4, 5). This seasonal variation has been described by other authors for estuarine sediments in England (Wright and Mason 1999). Out of the samples collected in July 2002, only sample 5SC can be considered polluted, and in the samples corresponding to February 2003, those from the northern area present As, Cd, Co, Cr, Pb and Zn values above the median. The same occurs with Zn, Cu and Mn around the city of Nador. Sample 16SC, found in an area of intensive agriculture close to Selouane, gives anomalous values for Cr, Ni and Cd.

For stream sediments the pH oscillates between 7.6 and 8.9, and in general the fraction < 2 mm is higher (>70%) than in the lagoon sediments. No significant differences between sampling points I (upstream) and II (close to the mouth) were found in these sediments. Aluminium percentages are over 10% in most of the samples for wadis that enter the lagoon in zones A and B. This is due to the lithology of the Gourougou volcanic complex through which these streams drained. Ca percentages are very low in these zones and rise to 25% in zone C samples owing, in this latter case, to the fact that streams pass through sediments with abundant carbonate materials. Fe is not above 6% in any case and K is over 3% in almost all the samples from zones A and B, the lowest readings being in the zones where the wadis cross Tertiary sediments.

These results may indicate that major elements are brought into the lagoon by streams, and, moreover, are totally linked to the type of lithology through which these streams pass, volcanic rocks in the northern area and marshy sediments to the south.

No significant values of the studied trace elements have been found (Table 6). In zone A there is no high value for any trace element analysed, and given that the heavy metals’ distribution in lagoonal sediments are related to the detritus input owing the nature of the source areas, it can be deduced that the high As, Cd and Co values found in lagoonal sediments in this zone were not carried by these wadis. In zone B isolated anomalies were found for Cd, Cu and Zn (sample 11D-I), and the same occurred in zone C for Cd, Cr and Ni (sample 16D-I). A comparison of these data with the values obtained for lagoonal sediments, allows to deduce that some of these high trace element contents in zones B and C may be due to wadi-borne sediments. A similar moderate Cd pollution also was encountered by Howari and Banat (2000) in Jordan and Yarmouk river sediments. In zone D no outstanding anomalies appear in either lagoonal or stream sediments (Table 6).

Table 6 Trace element content (ppm) in stream sediments. February 2003
Full size table

In soil samples, the pH is higher than 8 and the percentage of fraction < 2 mm is very high. The soil taken from the vicinity of the Gourougou volcanic complex (zone A) differs from the rest in that it presents a very high Fe content (43.4%) and low calcium content, which is logical owing to the lithological characteristics of the area. Agricultural soils 43B, 44B and 45B share a similar composition with Al and Ca percentages between 11 and 14% and low iron content. With regard to trace elements (Table 7), soils from zone A give high values in As (68.2 ppm), Co (96 ppm), Mn (4,254 ppm), Cu (586 ppm), Zn (646 ppm) and Pb (300 ppm). It is therefore possible that erosion releases these elements and may be the source of those found in the lagoon, as occurs in the study by Johansson et al. (1995) where soils and sediments in Sweden show similar contents and distribution for trace element concentration. These results for this zone agree with the existing mineralizations of chromite-Ni arsenide and Ni–Fe–Cu sulphide ores exist in the lherzolites (Gervilla and Leblanc 1990). Soils from zone C present high Cd contents, and they may therefore be deduced as the direct cause for the values found in this area of the lagoon. The same cannot be said for Cr and Ni, as they do not appear in the studied samples from this zone (Table 7).

Table 7 Trace element content (ppm) in soils. February 2003
Full size table

Mineralogical characterization

Samples from the northern zone of the lagoon show an overall mineralogy made up of high percentages of feldspars and iron oxides. The phyllosilicate content varies from 5 to 35%. For sample 5SC, taken from the vicinity of former mine waste decant ponds, a semi-quantitative estimate could not be made owing to the high percentage of amorphous iron oxides. However, quartz, feldspars, phyllosilicates, hematite, magnetite and jarosite were identified.

  • The mineralogical association for zone A is feldspars > phyllosilicates > quartz, with dolomite, ilmenite, hematite, calcite, halite and magnetite as minor components. For the fraction < 2 μm, smectites are the major minerals followed by illite and chlorite + kaolinite in small proportions.

  • Zone B samples show greater feldspar and quartz contents and lower phyllosilicate content than the preceding zone. With regard to mineralogy of fraction <2 μm, the association is illite > smectites > kaolinite + chlorite.

  • Zone C samples show a significant increase in calcite. The mineralogical association is calcite ≥ quartz > phyllosilicates, and feldspars, dolomite, hematite, halite and magnetite as minor components. Illite > chlorite > kaolinite > smectites were found in fraction <2 μm.

  • Zone D shows a high calcite content. Major minerals are calcite  \( \gg \) phyllosilicates > quartz, with minor proportions of feldspars, dolomite, halite, hematite and magnetite. The mineralogy of clays is similar to that of zone C (illite > chlorite > kaolinite > smectites).

Statistical study

In order to make a preliminary assessment of the mineral phases the trace elements are associated with, a correlation analysis between aluminium (essentially contributed by phyllosilicates and feldspars) and iron (contributed by iron oxides, fundamentally) was made. The results obtained indicate that high correlation coefficients exist between Pb, Zn, Co and Cu with Fe in lagoon sediments instead there is no correlation with Al. These correlations are very similar to those obtained by Gobeil et al. (1999) and Bilali et al. (2002) from sediments from the Arctic Ocean basin and Lake Ontario (Canada).

Trace element anomalies in zone A must be related with phyllosilicates and/or iron oxides. Pb, Zn, Cu and Co would be associated with iron oxides, and probably Cr and Co to phyllosilicates (Bilali et al. 2002), but in the latter case the correlation would not be clear owing to the low percentage of these minerals in the zone.

It should be pointed that in none of the studied zones are As and Cd related to Al and Fe, and therefore their origin may be deduced as anthropogenic, or as isolated cases eroded from nearby materials but which have not been adsorbed/absorbed by iron oxy-hydroxides or phyllosilicates.

Metal mobility and bioavailability

Lagoon sediment samples with anomalies of Cu, Pb and Zn show a low availability of these elements because they are bounded to the residual, non-mobile, phase of the sediment (Fig. 5). Only such elements can be considered as available in samples from wadi, where they are associated to organic matter. Pb may be liberated easily from soil sample 42B, because it is bounded to organic matter or Fe-hydroxides. The bioavailability of trace elements is very low for all the cases studied. Only it is to be noted the extraction of Cu from soil sample 42B (≈30%) and Pb from sediment sample 5C (≈39%).

Fig. 5
figure 5

Geochemical partitioning of trace elements

Full size image

It is important to note the binding of As to sediment minerals. Only in samples 5C and 48(0–5) C (zone A) show a notable concentration of arsenic (75 and 33 ppm), but it is bound to the residual fraction. A high As concentration in level 1 (26–37 ppm) can be a problem because it is bound to organic matter and could be leached under degradation. Also Cd could be a problem in sample 16C because it is in the interchangeable fraction (Fig. 5).

In general, in these sediments most trace elements are associated to Fe amorphous oxy-hydroxides, or take part of the residual fraction. A similar distribution was found by Kersten and Förstner (1989) and Howari and Banat (2000) for the Jordan River sediments, and Da Silva et al. (2002) for the Brazil’s river sediments.

Geoaccumulation index

Values obtained for the geoaccumulation index are generally <1 for the whole lagoon, which may indicate that it is not polluted. Sample 5SC is an exception as it reached level 4 (moderately polluted) in both sample collections. Cd is the only element which appears with a geoaccumulation index < 2 (very low polluted) for the summer sampling in zone B, and with an Igeo < 3 (low polluted) for in some samples winter, especially from level 0 (Table 8). In the levels 1 and 2, the Igeo values are lower than 1.

Table 8 Index of geoaccumulation for sampling February 2003 at level 0
Full size table

Conclusions

The studied sediment samples from Nador Lagoon show low concentrations of trace elements. The geoaccumulation indices indicate that the sediments are unpolluted. Only a few samples have been found with anomalous As, Cd, Cu and Pb values.

No significant differences were found between results obtained from sample campaign carried out in summer and winter, nor are there any variations with depth as far as trace element content is concerned.

The distribution of trace elements and chemical and mineralogical composition data, have allowed four areas in Nador or lagoon to be differentiated: Beni Ensar-Atelouane Zone (zone A), Nador Zone (zone B), Centre-Kebdana Channel (zone C) and SE (zone D).

The chemical composition of stream sediments is completely conditioned by the lithology of the drainage area. The streams which enter the lagoon in zone A cannot be considered sources of As, Cd and Co pollution, owing to the insignificant amounts present in samples, but it is possible that they contribute with Cr and Pb, because they are bounded to organic matter. On the contrary, those that enter the lagoon in zones B and C can be considered sources of pollution owing to the Cu–Zn and Cr–Ni–Cd values obtained, respectively, in sediments from wadis.

There is a clear difference between soils from the northern area (Gourougou volcanic complex) and the southern, agricultural area (Tertiary marls). Zone A soils could also be a source of As, Co, Mn and Pb, owing to the high values of these elements in soils, but as they are concentrated in the residual fraction, their availability will be very low.

Concerning possible sources of pollution both a natural and anthropogenic origins occurred. In zone A, As, Co, Cu and Pb derived from materials of Mount Gourougou by soil weathering and erosion, and Cd has an anthropogenic origin owing to the presence of industry and former mines. In zone B, contributions of Cu and Zn, both through soil weathering and river-borne, are anthropogenic pollution, given that urban waste outflows in the area are coming from the diverse human activities carried out on its shores. In zone C, the most important pollutant Cd is deduced to be of anthropogenic origin as the watercourses by which it enters the lagoon pass through areas of industry and intensive agriculture.