Introduction

The Piedras River marshland and El Rompido spit bar Natural Site is located in Huelva Province, SW Spain (Fig. 1) and extends for 2530 ha. The area was declared a Natural Site in 1989 and also a Special Area of Conservation of Wild Birds in 2002. The area has been proposed as a European Site of Community Importance. This Natural Site is located at the mouth of the Piedras River containing a tidal marshland protected from the Atlantic Ocean by a ~12 km long spit bar. This estuary is located between the Guadiana River to the west and the Tinto and Odiel Rivers to the east. These rivers drain the Iberian Pyrite Belt, the largest repository of volcanogenic massive sulphide deposits in the world, and as a result of mining within the catchment areas both rivers have high levels of pollution (Borrego et al. 2004; Delgado et al. 2009).

Fig. 1
figure 1

Location of the study area and cores

Full size image

The geology of the drainage basin is represented in Fig. 2. To the north, the Hercinian deposits are represented by Carboniferous dolomites, dolomitic limestone’s and marls. Miocene deposits consist mainly of calcarenites and yellow silts while Pliocene deposits consist of grey–yellow sands and silty sand. The Quaternary deposits are represented by river terraces in the upper areas and eolian and marsh deposits in the lower areas.

Fig. 2
figure 2

Geological map of the study area

Full size image

The special geological characteristics of this area have led to research on the trace metals present in the sediments of the estuaries located in this area to evaluate the degree of contamination (Ruiz 2001; Santos Bermejo et al. 2003; Borrego et al. 2004; Lozano-Soria et al. 2005; Sainz and Ruiz 2006; Ruiz et al. 2009; Delgado et al. 2010; Carretero et al. 2011). Most of this research has been focussed on the Guadiana and Tinto-Odiel estuaries, while research in the Piedras River estuary has been less extensive and only using surface samples. Sainz and Ruiz (2006) studied the influence of the very polluted Tinto-Odiel River and concluded that the littoral fringe extending up to 50 km is classified as moderately to strongly polluted. Due to dominant longshore currents along this coast much of this pollution is transported to the east, some pollution is though transported to the west, reaching the mouth of the Piedras River estuary. Due to the high interest in this area as a Natural Site, the major objective of this study is to evaluate the potential heavy-metal pollution in surface and deep sediments of the marshland through the last 7000 years. This is the first time core samples have been recovered to study the pollution history of the area enabling the determination of the local background values for the heavy metals in the Piedras River estuary. In addition an attempt is made here to evaluate whether the anthropogenic activity resulting in the pollution recorded in the sedimentary record of the two adjacent river estuaries is present in the sedimentary record of the Piedras River estuary.

To study the heavy metal pollution in estuarine sediments indices such as contamination factor (CF), geoacumulation index (I geo) and enrichment factor (EF) have been used in other coastal areas (Rubio et al. 2000; Reddy et al. 2004; Chen et al. 2007; Praveena et al. 2008; Dasaram et al. 2011; Feng et al. 2011) and are still used to assess contamination in soils, river environments, estuaries, lakes, reservoirs or urban sediments (Kumar et al. 2013; Zahra et al. 2014; Diop et al. 2015; Elkady et al. 2015; El-Sayed et al. 2015; Palma et al. 2015; Rivera et al. 2015; Ma et al. 2016). To assess sediment contamination at this location the indices cited above have been calculated with comparisons made between sediments deposited pre and post anthropogenic activity in the area. These indices have been correlated with other factors such as particle size and total organic carbon (TOC) values of the sediments.

Methodology

Sampling and analyses

A total of 18 manual gouge and percussion boreholes were completed in the study area. For this study three percussion cores were selected that included a full record of the Holocene infilling sequence (cores RP-11, RP-15 and RP-16, Fig. 1). These cores were sampled into plastic pipes and stored in a fridge at −4 °C, cores were then cut longitudinally and sampled at intervals of 20 cm resulting in a total of 91 samples. Contamination was avoided by discarding the outer layer of each core section. Dates from cores were established by Lario et al. (2009) through radiocarbon dating. The ages and sedimentation rates agree with data available for others estuaries in the Gulf of Cadiz (Lario et al. 2002).

The bulk content of Cr, Cu, Ni, Pb and Zn present in the sediments was studied as they are the main heavy metals present in the area (Leblanc et al. 2000; Ruiz et al. 2008) in addition they are the trace metals typically determined in heavy metal contamination studies (Kanallopoulos et al. 2006; Feng et al. 2011). To determine the total concentration of metals in the sediment samples, 0.5 g of the <63 µm fraction (dry weight) obtained by dry sieving was used. The <63 µm fraction was utilised as the content of heavy metals is normally accumulated in the fine fractions associated to clay minerals and oxides (Soares et al. 1999; Fan et al. 2002; Cuong and Obbard 2006). The samples were digested with a mixture of acids (9 ml of concentrated HNO3 + 3 ml of concentrated HCl) in Teflon® PTFE (polytetrafluoroethylene) beakers in a microwave unit (CEM MARS 5, Matthews, USA), according to EPA 3051A method (USEPA 1996). Dissolved samples were then diluted to 40 ml with milli-Q water. Following digestion, the heavy metal concentration in the sample solutions and reagent blanks was measured by ICP-AES in a Thermo iCAP 6500 spectrometer and by ICP-MS in a Thermo X-Series II spectrometer (Thermo Electron, Cambridge, UK). All reagent blanks and matrix interference were monitored throughout the analyses, and were below the instrument detection limit. The analytical method was assessed using the 2711 Standard Reference Material (Montana Soil, from LGC Promochem, Barcelona, Spain) with an agreement of 93.3–107.7 % between the certified values and the concentrations obtained by this study. Al and Fe were determined by X-ray fluorescence spectrometry, using a Phillips Spectrometer PW 1410/20 with a PW 1730 generator.

Particle size distribution was analysed by laser diffraction using a Coulter® LS 230 and Carbon content (total, organic-TOC- and inorganic-IC) of the sediments was analysed using a Shimadzu® TOC-VCSH analyser.

Use of pollution indices

The metal content in sediments is the result of an addition of metals originating from natural sources and anthropogenic activity. It is estimated that the contribution of metals incorporated in sediments as a result of human causes is higher than the contribution from natural processes (Nriagu and Pacyna 1988). In a long-term study such as this across the Holocene the identification of phases of raised metal concentration in sediments would typically be related with significant changes in human activities in the catchment area. Establishing the pollution history of the estuary and thus human activity in the catchment area is thus value to archaeological study of the area.

In this study, the possible sediment contamination was assessed using various pollution indices, including contamination factor and degree of contamination, index of geoaccumulation and normalized enrichment factor, these are outlined below.

Contamination factor (CF)

The assessment of sediment contamination has been carried out using the contamination factor (CF) and contamination degree (CD). The approach suggested by Hakanson (1980) is applied which enables an assessment of sediment contamination through using pre-industrial levels as a reference of the background concentrations enabling comparison with post-industrial sediments.

Calculation will be simplified as:

$${\text{CF}} = Cm/Cb$$

where Cm is the mean content of an individual metal in the sample and Cb is the pre-industrial concentration of an individual metal (background concentration). This should then identify a direct relationship between the concentration of a specific metal and how this corresponds to the background level.

Hakanson (1980) define four categories of contamination factor (CF)

CF < 1:

low contamination factor

1 < CF < 3:

moderate contamination factor

3 < CF < 6:

considerable contamination factor

CF > 6:

very high contamination factor

The contamination factor described above is a single element index. The sum of contamination factors for all elements examined represents the contamination degree (CD) of the environment and four classes are recognized (Hakanson 1980)

CD < 8:

low degree of contamination

8 < CD < 16:

moderate degree of contamination

16 < CD < 32:

considerable degree of contamination

CD > 32:

very high degree of contamination

Geo-accumulation index (I geo)

The geoaccumulation index (I geo) has been used since the late 1960s, and has been widely employed in trace metal studies. Müller (1969) first used this index in basal sediments but it has also been used largely to evaluate contamination in soils (Loska et al. 2003, 2004; Yaqin et al. 2008; Rivera et al. 2015), rivers (Ma et al. 2016), lakes (Zahra et al. 2014; Elkady et al. 2015), beaches (Diop et al. 2015), mangroves (Praveena et al. 2008), Ria type embayment coasts (Rubio et al. 2000) or even harbours (Chen et al. 2007) and reservoirs (Goher et al. 2014; Palma et al. 2015). The Igeo enables the assessment of contamination by comparing current and pre-industrial concentrations, although it is not always easy to reach pre-industrial sediment layers (Yaqin et al. 2008). In this study, the I geo for selected sediments was calculated using

$$I_{\text{geo}} = \log_{2} \left( {Cm/1.5Cb} \right)$$

where Cm is the measured concentration of the element in the sediment and Cb is the geochemical background value. The constant 1.5 is used to analyse natural fluctuations in the content of a given substance in the environment and to detect very small anthropogenic influences. It is also a factor used for lithological variations of trace metals.

The geoaccumulation index consists of seven grades or classes (Müller 1979, 1981)

I geo < 0 :

I geo Class = 0 practically uncontaminated

0 < I geo < 1 :

I geo Class = 1 uncontaminated to moderate contaminated

1 < I geo < 2 :

I geo Class = 2 moderate contaminated

2 < I geo < 3 :

I geo Class = 3 moderate to strong contaminated

3 < I geo < 4 :

I geo Class = 4 strong contaminated

4 < I geo < 5 :

I geo Class = 5 strong to very strong contaminated

I geo > 5 :

I geo Class = 6 very strong contaminated

Enrichment factor (EF)

The use of EF has been widely applied as a means of identifying and quantifying the anthropogenic origin of certain elements since the 1970s (Chester and Stoner 1973; Duce et al. 1975), it is also a convenient measure of geochemical trends and is used for making comparisons between areas (Sinex and Helz 1981) and it is still widely used (e.g. Puig et al. 1999; Rubio et al. 2000; Karageorgis and Hatzianestis 2003; Loska et al. 2003, 2004; Selvaraj et al. 2004; Vazquez and Sharma 2004; Kanallopoulos et al. 2006; Radakovitch et al. 2008).

The extent of metal contamination compared to the background of the catchment area was assessed using the enrichment factor (EF) (Woitke et al. 2003; Selvaraj et al. 2004). The EF was also based on the standardization of a tested element against a reference one. A reference element is the one characterized by low occurrence variability such as Sc, Mn, Ti, Al and Fe (Pacyna and Winchester 1990; Quevauviller et al. 1989; Schiff and Weisberg 1999; Reimann and De Caritat 2000; Sutherland 2000).

Some authors used Al to normalize the metals in sediments since it represents the aluminasilicates, the predominant content of coastal sediments (Chen et al. 2007; Delgado et al. 2010) but other observations support that either Al or Fe can be used for metal normalization if there is a significant correlation between them (Feng et al. 2011). Lozano-Soria et al. (2005) found a high positive correlation between Al and Fe in Piedras River sediments and thus in this study Al and Fe were used as reference elements for metal normalization.

The enrichment factor was calculated using the formula based on the equation suggested by Buat-Menard and Chesselet (1979):

$${\text{EF}} = \left[ {Cm} \right)/Cr)\left] / \right[Cmb/Crb]$$

where Cm is the content of the metal in sample, Cr is the content of the reference element in the sample, Cmb is the content of the metal in the background sample and Crb is the content of reference element in the background sample.

In this case

$$\begin{aligned} {\text{EF}}_{\text{Al}} & = \left[ {Cm} \right)/C_{\text{Al}} )\left] / \right[Cmb/C_{\text{Al}} b] \\ {\text{EF}}_{\text{Fe}} & = \left[ {Cm} \right)/C_{\text{Fe}} )\left] / \right[Cmb/C_{\text{Fe}} b] \\ \end{aligned}$$

Five contamination categories are recognized on the basis of the enrichment factor (Sutherland 2000)

EF < 2:

deficiency to mineral enrichment

2 < EF < 5:

moderate enrichment

5 < EF < 20:

significant enrichment

20 < EF < 40:

very high enrichment

EF > 40:

extremely high enrichment

Analysis of results

A contrast independence test was carried out with the aim of determining whether significant relationships exist between variables. In addition a study of the data distribution was completed to establish the tests (parametric or non-parametric) that could be applied. A Kolmogorov–Smirnov (K–S) test was used to check the normality of the variables. Whilst Pearson Correlation Coefficient (R) and Spearman Correlation Coefficient (rs) tests were used for Normal and Non-Normal distribution, respectively. SPSS software was used for the calculations. This methodological approach has been applied in other geoacummulation indexes studies (Rubio et al. 2000; Diop et al. 2015).

Results and discussion

Sedimentary framework

From the cores studied it is possible to interpret the Late Pleistocene and Holocene evolution of the Piedras River estuary, and thus the pollution during this period. The river is incised into coarse-grained Quaternary sediments deposited during the last major lowstand ca. 18 ka, when sea level was ~120 m lower and the coastline lay 14 km seawards from the present (Dabrio et al. 2000). Some studies of estuaries and lagoons on the coast of south-western Iberia have shown that the post-glacial transgression reached larger valleys such as the Tagus (Vis et al. 2008), Guadiana (Boski et al. 2002, 2008; Delgado et al. 2012), Tinto-Odiel (Dabrio et al. 1999, 2000) and Guadalquivir (Dabrio et al. 1999, 2000) estuaries between 13,000 and 10,000 calBP. Smaller valleys were only inundated when sea level reached its post-glacial maximum at about 7500–6500 calBP (Dabrio et al. 1999; Freitas et al. 2002; Schneider et al. 2010). In the Piedras River estuary the sediments underlying the Holocene estuary sequence are Miocene Pliocene (Fig. 2). According to the interpretation (Lario et al. 2009) and the radiocarbon data obtained (Fig. 3) the transgressed estuary basins changed from brackish to more open marine conditions as the sea rose until ca. 6500 calBP, when it reached the maximum height and the sandy estuarine barriers ceased to prograde toward the muddy central basins. Then, as it has been observed in other estuaries from the Gulf of Cadiz, rates of sedimentation and the eustatic sea-level rise decreased significantly, and the estuarine infilling was dominated by lateral progradation (Lario et al. 2002). At ca. 4000 calBP the fluvial input exceeded the rate of sea-level rise, causing partial emergence of tidal flats in the mostly filled estuarine basin. Dominance of coastal progradation upon vertical accretion at ca. 2800–2200 calBP favoured the spreading out of tidal flats and sandy barriers. Some high energy (erosional) episodes have been recorded in the upper sequence as have been interpreted in others areas of the Gulf of Cadiz (Lario et al. 2010, 2011). Also, the development of the El Rompido spit barrier during the last two centuries has partially enclosed the estuary for the first time during the Holocene.

Selecting geochemical background values

Through all the indices calculated, the Cb parameter in the equations represents the geochemical background value. The equations indicate that the index will be affected by the content of the samples and the geochemical background values.

Some authors have proposed as background reference the world average shale (Rubio et al. 2000) or the Earth’s crust (Loska et al. 2003, 2004). However, these levels tend to be very general and may distort the results. Rubio et al. (2000) recommended the use of regional background values. While the geochemical background values are constant, the levels of contamination vary with time and space.

In this study, to evaluate the possible contamination of the area during the last 7000 years, samples have been selected as background value samples that relate to ‘natural’ conditions, pre-dating the human activity. According to archaeological data, the extraction of minerals from the catchment area dates back to the age of the Iberians and Tartessos, some 5000 years ago (Davis et al. 2000; Leblanc et al. 2000; Nocete et al. 2005). Samples of that age also represent a significant point in the depositional history of the Piedras River estuary at ca. 6500 calBP, when open marine were recorded representing the Holocene sea-level maximum. Thus, samples corresponding to this age are appropriate to be used as a background level because they represent the beginning of infilling of the estuary after the maximum Holocene transgression and are unpolluted by anthropogenic activity. Importantly the environmental reconstruction identifies that the sedimentary environment has remained relatively consistent since ca. 6500 calBP, as recorded in other nearby estuaries (Lario et al. 2002), enabling reliable comparison of contamination levels through this time period.

Therefore, to estimate the background values, five samples from core RP-15 dated ca.6500 calBP were analysed and the mean values calculated. The values for the average background in the Piedras River estuary produced from this analysis were Cr: 45.8 mg/kg, Cu: 16.1 mg/kg, Ni: 27.2 mg/kg, Pb: 13.7 mg/kg and Zn: 62.2 mg/kg (Table 1).

Table 1 Background geochemical values calculated in the different estuaries of the Southwestern Spanish Coast (element concentrations in mg/kg)
Full size table

Different background values have been calculated for the different estuaries in the south-western Spanish coast (Ruiz et al. 1998; Delgado et al. 2008, 2012; Carretero et al. 2011; Table 1). All of them are based on samples of Holocene age except the one calculated for the Tinto-Odiel estuary in which samples of Neogene and Holocene age have been used. The background values for the Tinto-Odiel estuary are slightly lower than for the other three estuaries, probably reflecting the larger time-span of samples used for the calculation. The other three estuaries show very similar background values, particularly the Doñana and Piedras River estuaries. Only the value of Cr for the Doñana estuary is significantly higher than in the other three, indicating the necessity of calculating the background values for each studied area, as local geological variations can lead to variations in the geochemistry of sediments and thus the background values among areas with similar geology.

Contamination indices

A total of 91 samples from the three cores have been analysed to evaluate the possible record of pollution through time at the site. Particle size distribution, organic carbon content and metal concentration values are presented in Tables 2, 3 and 4. For the three cores the proportion of the clay fraction in the sample was low, generally under 10 %. In cores RP-15 and RP-11 the silt fraction predominated while in RP-16 the sandy dominated. TOC values were not very high ranging between 0 and 1.76 %, typically increasing towards the top of the three cores due to the organic carbon decomposition in older sediments (Kanallopoulos et al. 2006). Mean TOC values decreased along the profile towards the more marine locations in the estuary (RP-11 > RP-15 > RP-16, Figs. 1, 2).

Table 2 Particle size distribution, organic carbon content and metal concentration values in RP-11b core
Full size table
Table 3 Particle size distribution, organic carbon content and metal concentration values in RP-15 core
Full size table
Table 4 Particle size distribution, organic carbon content and metal concentration values in RP-16 core
Full size table

For the study of the contamination indices, 14 samples have been selected (Table 5) representing different periods of time through the Holocene from the time of maximum sea-level. The samples represent the estuarine filling sequence and therefore any natural changes (e.g. flooding, increase in erosional rates…) or human induced changes (e.g. minning, farming…) during this time should be recorded. In addition, some older samples (up to ca. 9000 calBP) have been analysed to compare the estuarine background values with previous natural values and to detect any previous input of metals to the basin due to alluvial episodes or leaching of metallic deposits.

Table 5 Metal concentration in selected time slices
Full size table
Fig. 3
figure 3

Sedimentology of the cores and cross section, with indication of the Holocene estuary sequence studied

Full size image

Table 6 shows the CF values for the selected samples. All values were in the low contamination factor or low values of moderate contamination factor categories. Therefore, all samples showed a low degree of contamination except for the surface samples, which showed a moderate degree of contamination.

Table 6 Contamination factor (CF) and contamination degree (CD) in the selected time slices
Full size table

The I geo (Table 7) shows the same tendency as the CF index with uncontaminated levels in most samples. I geo values are lower than 1, which classifies the sediments as unpolluted, similar to the values calculated by Ruiz (2001) in surface samples. The enrichment factors (EFAl and EFFe, Table 8) both showed similar trends and most samples indicate deficiency to mineral enrichment.

Table 7 Geoaccumulation index (I geo) in the selected time slices
Full size table
Table 8 Enrichment factors (Al and Fe) in the selected time slices
Full size table

Statistical analysis

Tables 9, 10 and 11 show the Pearson correlation matrix for the particle size fractions, TOC and the heavy metals studied. It can be seen that both the grain size and total organic carbon are important factors affecting the metal distribution in the sediments with high correlation factors. Generally there is a strong correlation between the quantity of the fine fractions and trace metal concentrations (Wang and Chen 2000; Huang and Lin 2003) while in this study the TOC values were relatively more important than grain size in controlling the distribution of trace metals in the sediments as it correlates with higher values with all metals than with the clay content. These results demonstrate that organic matter content was a more important factor affecting the trace metal concentrations than the grain size distribution. The Cr, Ni, Pb, Zn and Cu correlate positively for all the sections, which indicate a common geogenic origin for the assemblage.

Table 9 Pearson correlation matrix showing the coefficients of correlation among grain size fractions
Full size table
Table 10 Pearson correlation matrix showing the coefficients of correlation among grain size fractions
Full size table
Table 11 Pearson correlation matrix showing the coefficients of correlation among grain size fractions
Full size table

TOC has a positive correlation with clay and silt fractions and negative with the sand fraction (Tables 9, 10, 11) which indicates that organic matter in the sediments was fine to very fine grained.

The study of all the contamination indices clearly shows a consistent trend indicating that the Piedras River estuary sediments display no signs of contamination as a result of human activity during the Holocene. This greatly contrasts with the situation of the adjacent estuaries.

The Iberian Pyrite Belt (IPB), with over 1600 million tons of polymetallic massive sulphide deposits originally in place and about 2500 million tons of mineralization in the stockwork, is one of the most significant metallogenic areas in the world and possibly the largest concentration of sulphides worldwide. Also, the Iberian Pyrite Belt is one of the oldest mining districts in the world with mining activity recorded over the last 5000 years. Mining commenced in the Chalcolithic period (third millennium BC) and continued during the Bronze Age and Tartessian civilization. During Roman times industrial mining extracted more than 20 million tons. After the abandonment of the mines in the fourth century BC, the mines were exploited only on a minor scale. In the late nineteenth century the mining was operated by British and French companies, and large-scale, open-pit operations prevailed until the deposits had been essentially depleted about a century later. Due to this activity the Tinto and Odiel Rivers represent an extreme case of acid mine drainage with high concentrations of toxic elements. The many sources of pollution, along with the low buffering capacity of the rocks that form the substrate of the basins of the Tinto and Odiel Rivers, result in high levels of contamination, unparalleled worldwide. (Ruiz 2001; Sainz and Ruiz 2006; Olías and Nieto 2012).

In the Doñana estuary relatively polluted sediments (3800–3000 calBP) have been found derived from old mining activities located in its catchment area (Carretero et al. 2011). In the Guadiana estuary Delgado et al. (2012) determined background levels from core sediments unaffected by human activities and determined that the surface elements of the estuary with ages younger than 4500 calBP contain significant levels of contamination associated with the mining activities in the IPB. Firstly the Guadiana Estuary sediments record a general increase in elements such as Co, Cr and Ni in the Copper Age, in the Bronze age and in the Roman period greater increases of heavy metals were recorded. Finally the most recent period of modern mining activities are also recorded in the heavy metal contents of the estuary sediments.

The absence of pollution in Piedras River estuary is interpreted as being due to the fact that in the Piedras River catchment area there are no mineral ore deposits and thus no historic mining activity. Sainz and Ruiz (2006) found that the presence of pollution at the mouth of the Piedras River indicates the influence of flow from the Tinto and Odiel Rivers. Absence of this pollution in the estuarine sediments indicate than the inner estuary has not been influenced by discharge from the Tinto and Odiel Rivers during the Holocene and, therefore, the dominant longshore currents in this area have remained the same (towards the east) throughout the Holocene. There is no evidence of the transfer of contaminants between the closed basins by wind or by the sea, and thus sedimentation in the Piedras River is dominated by fluvial input rather than from extra-estuary marine sources.

Data obtained in this study supports that of studies of surficial pollution in the Piedras River estuary (Ruiz 2001; Lozano-Soria et al. 2005), that found very low or unpolluted sediment. Furthermore we found that this characteristic extends throughout the last 7000 years confirming that the area has not been influenced by pollution from human activity during the Holocene. Whilst similarities exist between this study and the previous studies in the region it must be stressed that previous work was limited to surficial samples and used background values from the nearby Tinto-Odiel estuary. Therefore, this study represents the first investigation where the research has included the study of the entire estuarine sequence and has used local geochemical background values to evaluate the record of the trace metals in the estuary trough the time.

Conclusions

This study evaluated the degree of heavy-metal contamination of the Piedras River Natural Site using a variety of pollution indices. Through analysing sediments derived from deep coring it has been possible to recover the entire Holocene sedimentary sequence of the estuary and for the first time local geochemical background values of uncontaminated sediments has been established.

The comparison of background geochemical values from different estuaries in the area reveals the necessity of calculating the background values for an estuary using local data and an adequate time-span. Local geological variations can result in key changes in the geochemistry of sediments producing changes in the background values among areas with similar geology.

For the first time in the Piedras River estuary this work uses the study of the entire estuarine sequence and local geochemical background data, to evaluate the evolution of the trace metals in the estuary trough the time. Pollution indices reveal low or unpolluted sediments, extending through the last 7000 years, and confirming that the area has not been influenced by pollution from human activity during the Holocene. The results presented here agree with previous work on this estuary that was limited to surficial samples, and using regional background values (from the nearby Tinto-Odiel estuary, Ruiz 2001; Lozano-Soria et al. 2005).

Results show that all of the pollution indices calculated dismiss any trace of heavy-metal pollution through the last 7000 years and therefore there is no evidence of any minning activity in this catchment area. The heavy-metal concentrations are of geogenic origin and appear to be linked to the amount of fine grained organic matter in the sediments rather to clay fraction. The results also indicate that the estuary has not been influenced by extra-estuary pollution sources from the nearby polluted Guadalquivir Rivers, Guadiana River or Tinto and Odiel Rivers. It can also be inferred that the dominant longshore currents in this area have remained the same during the Holocene and that sedimentation in the Piedras River estuary is dominated by fluvial input rather that extra-estuary marine sources. This study provides new pollution data and detail of estuary evolution during Holocene; such details are relevant for the preservation of this Natural Site and Special Area of Conservation of Wild Birds.