Sediments as Sentinels of Pollution Episodes in the Middle Estuary of the Tinto River (SW Spain)

1. Introduction

Numerous anthropogenic activities (e.g., industrial effluents, mining waste, agricultural residues, urban inputs), the natural erosion of geological formations or even high-energy events (river floods, storms, tsunamis, etc.) leave their mark on the geochemical levels of sediments and soils [1,2,3,4,5,6]. One of the most common effects is the increase in concentrations of certain elements, among which heavy metals are commonly used as a reference to measure the level of pollution [7,8,9]. To determine the heavy metal contamination in a soil or sediment, various indices are often used, such as the enrichment factor (EF) or the geoaccumulation index (I_geo), among others [10,11,12,13]. The latter index requires a regional background in order to calculate the degree of contamination, which may be unique for each metal regardless of grain size (e.g., [14,15]), or may include a differentiation between sandy and muddy sediments [16].

Estuaries are often the final repositories of many of these natural processes or human actions, as they often contain the inputs of heavy metals produced in their river basins at present or in the past [17,18,19]. In the last years, numerous investigations have focused on the historical heavy metal pollution of these environments, based on geochemical studies of continuous sediment cores [20,21,22]. Some of them have applied the geocumulation index to detect pollution episodes and, in conjunction with radiocarbon dating and historical data, to pinpoint the time interval in which they occurred [23,24].

The main objective of this short communication is the detection of natural or anthropogenic pollution episodes in a continuous sediment core extracted from the middle estuary of the Tinto River (SW Spain), based on the vertical geochemical variations and the application of the regional geoaccumulation index to five elements. The results obtained will be compared with previous palaeoenvironmental reconstructions and historical mining data to identify the age of these episodes and their importance according to the degree of contamination deduced.

2. Materials and Methods

2.1. Study area

The Tinto River is a small stream that flows through the southwest of Spain and forms a large estuary together with the Odiel river at its mouth in the Atlantic Ocean, surrounded by Neogene-Pleistocene formations (Figure 1A,B). The hydrodynamics of this estuary are tidally dominated, as river discharges are very rare and vary from minimum flows during very dry years to about 350 Hm³ in rainy years [25]. The tidal regime is mesotidal (2.1 m), with a low diurnal inequality [26].

This estuary can be divided into three sectors, depending on the interaction between river and tidal inputs [27]: i) fluvial estuary, composed of numerous braided channels; ii) middle estuary, with numerous marshes that come to form small islands (e.g., Figure 1B: Bacuta Island); and iii) marine estuary, with a central island (Figure 1B: Saltés Island), two main channels (Padre Santo channel, Punta Umbría channel) and two sandy spits (Punta Arenillas, Punta Umbría) that partially protect it. In the marine estuary, wave energy is medium and 75% of the waves dot not exceed 0.5 m in height [28]. The coastal drift currents are oriented towards the east, with a high annual transport of sediment (1.8-3 x 10⁵ m³) [29,30] and have historically favoured the development of the two spits and the progressive closure of the innermost parts of the estuary.

2.2. Historical pollution

Southwest Spain is part of the Iberian Pyritic Belt, one of the most important metallogenic provinces in Europe, which includes some of the largest massive sulphide deposits in the world. These giant deposits have been exploited for 5,000 years within the Tinto River drainage basin [25], with an important mining activity during the Roman period (2100-1700 yr BP) and especially in the last 150 years, with an intensive extraction of pyrite and minor quantities of gold and silver. Currently, the waters of the Tinto River suffer from acid mine drainage, due to this age-old mining activity and to the heavy metal-rich inputs from the washing of the waste dumps generated [31].

In addition, this estuary has been affected by discharges from two industrial concentrations located on its banks since the 1960s (Figure 1B), made up of copper producers, fertiliser factories and refineries, among others. As a final result of these mining-industrial pollutant inputs, the surface sediments of the Tinto-Odiel estuary are among the most polluted in the world, with very high concentrations of As (up to 3,000 mg kg^-1), Cu (up to 4,415 mg kg^-1), Pb (up to 10,400 mg kg^-1) and Zn (up to 5,280 mg kg^-1) in its surface sediments [16,32]. Since 1985, this zone has come under a corrective plan for the control of industrial waste disposal.

2.3. Coring, sampling and Holocene palaeoenvironmental evolution

Core A (Figure 1B,C; 7.5 m depth; 37°14′06′’N; 6°53′54′’W) was extracted in the middle estuary of the Tinto River by usual rotary drilling techniques, with an almost continuous recovery of sediment and a barrel diameter of 11.6 mm. Sixteen samples (2 cm thickness) were obtained, with a distribution of samples linked to: a) the presence of different sedimentary facies; b) the definition of its limits; and c) the visual distribution of bioclasts in the core. The Holocene paleoenvironmental reconstruction of this core was inferred by [33], with basal alluvial sands (Figure 1C: facies F1), bioclastic sands and gravels from the marine flood of this area during the MIS-1 transgression (F2) and upper muddy marsh deposits (F3).

2.4. Chemical analysis

The chemical analysis of sixteen sediment samples were performed and certified by MS Analytical, Langley (Canada). Samples were collected manually by hand picking. They were respectively crushed and subsequently ground into powder using an agate mortar. The concentrations of five elements (As, Co, Cu, Pb, Zn) were obtained by ICP Atomic Emission Spectrometry (total digestion with concentrated acids), with less than 5% variation between different replicate samples and a quality control based on more than 30 references (e.g., OREAS 904). In these samples, detection limits vary between 0.01 mg kg^-1 (e.g., Co) and 2 mg kg^-1 (e.g., Zn).

2.5. Background and geoaccumulation index

A regional background of this estuary was obtained by [34] for both sandy and muddy sediments. This background was applied to obtain the geoaccumulation index of the extracted sediments, according to the following formula:

I_geo= log₂ C_n/1.5 x B_n,

where C_n is the concentration of an element in a given sample and B_n is the background of that element for the grain size of that sample (e.g., sand or mud). This background differentiates sandy (S) and muddy (M) sediments: As (S: 6.8 mg kg^-1; M: 7.3 mg kg^-1), Co (S: 4.2 mg kg^-1; M: 6.8 mg kg^-1), Cu (S: 13.1 mg kg^-1; M: 13 mg kg^-1), Pb (S: 4.3 mg kg^-1; M: 15.4 mg kg^-1) and Zn (S: 15.9 mg kg^-1; M: 39 mg kg^-1).

The pollution classes originally proposed by [35] have been slightly modified as follows: unpolluted (UP; I_geo < 0), very low polluted (VLP; 0 < I_geo < 1), low polluted (LP; 1 < I_geo < 2), moderately polluted (MP; 2 < I_geo < 3), highly polluted (HP; 3 < I_geo < 4), very highly polluted (VHP; 4 < I_geo < 5) and strongly polluted (SP; I_geo > 5).

2.6. Dating

Two dates of core A were produced at the National Center of Accelerators (Sevilla, Spain). These dates were calibrated using CALIB version 8.2 and the final results correspond to calibrated ages using 2σ intervals.

3. Results and discussion

3.1. Facies, geochemistry and dating

The basal alluvial sandy-muddy deposits of this core (F1; 7.5-6.3 m depth; samples A-16 to A-14) have very low concentrations of most of the elements studied, even lower than the regional background (Table 1). They were deposited during the middle Holocene, according to the dating obtained near the boundary between F1 and F2 (Figure 1C and Table 2; sample A-13; ~5.8 cal. kyr BP). The bioclastic sediments of this last facies (6.3-4.9 m depth; samples A-13 to A-11; ~5.8-5 cal. kyr BP) were deposited during the flooding of this area due to the MIS-1 transgression. A first pollution level was detected near the base of this facies (sample A-13: ~5.8 cal. kyr BP) , with a noticiable increase in the concentrations of Cu (from 9.5 mg kg^-1 to 39 mg kg^-1), Pb (from 11.1 mg kg^-1 to 74.1 mg kg^-1) and Zn (from 25 mg kg^-1 to 93 mg kg^-1). The age of this level would indicate an early natural pollution, as mining did not begin in this area until some thousand years later [25,36]. The origin of this first contamination is unknown, although it could come from a natural oxidation of the sulphides and formation of the gossans of the Iberian Pyritic Belt by acid rock drainage processes, which started more than 24 million years ago in this area [37]. Metals from the washing of these deposits would have been transported to the Tinto estuary and deposited in core A when its environment was flooded during the Holocene transgression.

The second pollution level was detected in the marshy muds collected near the base of F3 (Table 1: sample A-9), with some of the highest values of Cu (51.8 mg kg^-1), Pb (46.9 mg kg^-1) and Zn (107 mg kg^-1) of core A and increasing concentrations of As and Co. According to the average sedimentation rates (1.7-2 mm yr^-1) calculated by [38] in this estuary around 5000-4000 yr BP and the aforementioned dating, this new episode would have taken place approximately between ~4.7 kyr BP and ~4.5 kyr BP, coinciding with the development of the first mining and metallurgical works in this area [36].

The concentrations of the five elements decrease between 4.2 m depth and 2.3 m depth, with a further increase of As (from 18.3 mg kg^-1 to 36.9-39.3 mg kg^-1) and, to a lesser extent, Pb (up to 37.9 mg kg^-1) observed between 1.9-1.6 m depth. This third pollution episode is associated with the start of intensive mining in the Tinto river basin from the second half of the 19th century to 1960. This increasing trend is confirmed at 1.4 m depth (e.g., the surface of core A before the recent antropic filling), with the maximum values of all five elements, and especially Cu (284.4 mg kg^-1), Pb (98 mg kg^-1) and Zn (539 mg kg^-1). This fourth episode would correspond to the conjunction of this intensive mining and inputs from nearby industrial complexes during the second half of the 20th century, which have caused a strong pollution in the recent sediments of the Tinto-Odiel estuary [39,40]. These high values confirm this last episode as the most important in the geochemical history of the middle estuary of the Tinto River. Transit to the overlying artificial filling is evidenced by the presence of unpolluted fine sands in the upper 1.3 m of this core (Table 1: sample A-1).

3.3. Geoaccumulation index and pollution episodes

These four episodes are also detected when I_geo is applied to the concentrations of the five elements (Figure 2). The first episode (~5.8 cal. kyr BP) is characterized by the transition from UP (As, Co, Cu) or VLP (Pb) sediments to VLP (Co) or LP (As, Cu, Pb) sediments. These four elements also show similar variations in their pollution class during the second episode (~4.7-4.5 kyr BP), most marked for Pb (from UP to LP) and to a lesser extent for As (from VLP to LP).

The third episode (~1850-1960) is mainly defined by As (from VLP to LP), although an increase in the concentration of most of the other elements is observed (see previous chapter). The fourth episode (~1960-2000) is clearly the most important, with the presence of MP (As, Pb) to HP (Cu, Zn) sediments. Consequently, this last polluting episode is the most important in the geochemical history of the Tinto-Odiel estuary, as evidenced by the analysis of other sediment cores extracted in this area [34,41,42].

5. Conclusions

The joint application of geochemical analysis and the geoaccumulation index has proved to be an effective tool for detecting pollution episodes in the middle estuary of the Tinto River (SW Spain), in conjunction with historical data on mining activities in this area. Four episodes have been distinguished, characterised by an increase in the concentration of the elements studied and their geochemical classes based on I_geo. The first recorded pollution was of natural origin and coincided with the flooding of this estuary during the Holocene transgression. The next three episodes are linked to the start of mining activities, a period of intensive mining and the coincidence of intensive mining and the discharge of highly polluted industrial effluents into the estuary.