Vertical profiles of cobalt and zinc in marine sediments of the Santa Rosalía mining region , Gulf of California , Mexico

Core sediments that were sampled from the coastal zone of the Santa Rosalía mining region and the adjacent deeper area of the Gulf of California were analysed by an instrumental neutron activation analysis (INAA). The levels of Co, Sc and Zn were used to assess the effect of pollution from the solid wastes of a copper smelter on the sediment composition in the study area. The Co/Sc and Zn/Sc ratio vertical distributions were compared to Co/Sc and Zn/Sc ratios of the earth’s crust and to a lower layer of the sediment cores and therefore less affected by pollution. The ratios of Co/Sc and Zn/Sc in cores from a predominant pollution “hot spot” near Santa Rosalía port are very high (40-150 and 150-350, respectively), suggesting that the thickness of the polluted layer exceeded the length of the cores (75-93 cm). The values of Co/Sc and Zn/Sc ratios decrease drastically in the cores collected outside the main “hot spot”. In this case, the values obtained from core depths of 20-34 cm approached the regional coastal surface sediment background (approximately 2 and 5-10 for Co/Sc and Zn/Sc, respectivelly). The Co/Sc ratios for the sediments of the deeper and further from the coast cores are slightly variable, but at sediment depths below 15 cm they are nearly constant and approximately equal to the Co/ Sc ratio of the earth ́s crust. In the upper part of these cores (0-15 cm), the Zn/Sc ratios are higher (10-20) than the Zn/Sc ratio of the earth ́s crust, which was probably due to the additional input of incompletely mineralised biogenic particulate Zn from the photic layer of the water column. The shape of the vertical profile the Zn/Sc ratio of SR22 sediment core (depth 360 m) in front of the main pollution “hot spot” shows that the maximum levels of pollution occurred in the past, and self-purification/restoration of the natural pollutant levels in the fine deep-water sediments can be observed during recent years.


Introduction
The metal mining and smelting on the sea shore is quite common because of the close proximity of mineral deposits to the tectonically active areas of land-sea interfaces where ancient hydrothermal vents have been uplifted and exposed on the continental margins.The most well-known cases include copper deposits and mines along the Pacific coast of Canada (e.g., the "Brittania" mine in British Columbia), along the coast of northern Chile (near Antofagasta) and in West Greenland (Elberling et al., 2002), in addition to a Pb smelter in the East of Russia on the shore of the Sea of Japan (i.e., near Dalnegorsk in the Primorye region) (Shulkin, 1998).The dumping of mining tailings and solid smelting wastes into the adjacent sea used to be a widespread practice in the past.
One of these sites where sediments are strongly contaminated is a coastal zone along the Santa Rosalía mining region on the eastern shore in the centre of the Baja California peninsula in Mexico (Fig. 1).This area displays high levels of metals that are associated with former copper mining and smelting activities that were performed for nearly a century (Wilson and Rocha, 1955;Rodríguez Figueroa, 2004).There is clear evidence of the pollution of surface sediments by such solid waste dumping (Shumilin et al., 2000;Rodríguez Figueroa, 2004, 2010); however, recent data were somewhat encouraging, as the heavy metals from highly polluted sites were shown to be mainly concentrated in the residual fraction of the surface sediments (Shumilin et al., 2011).Additionally, an analysis of the metal levels in brown seaweeds showed that most toxic metals were retained in the sediments and were not released to the water column (Choumiline et al., 2006;Rodríguez Figueroa, 2010;Rodríguez-Figueroa et al., 2010).However, various severe environmental problems are still very real.For example, while the mining and smelting activities in Santa Rosalía closed more than twenty years ago, the self-cleaning/restoration dynamics of the pollution "hot spot" have not Zn/Sc de la corteza terrestre y con las obtenidas en la capa inferior de los testigos de sedimentos menos afectados por la polución.Las relaciones de Co/Sc and Zn/Sc en los testigos de sedimentos del área con mayor polución cerca del puerto de Santa Rosalía son muy altas, oscilando desde 40 hasta 150 y de 150 hasta 350 respectivamente, sugiriendo que el espesor de la capa altamente contaminada excede la longitud de los testigos (75-93 cm).Los valores de las relaciones Co/Sc y Zn/Sc disminuyen drásticamente en los testigos muestreados fuera de la principal área de polución.En este caso, los valores obtenidos en la profundidad de 20-34 cm del testigo se aproximan a los valores regionales de fondo de las relaciones para los sedimentos superficiales costeros (cerca a 2 para Co/Sc y a 5-10 para Zn/Sc).Las relaciones Co/Sc para los sedimentos de los testigos, muestreados a mayor profundidad y distancia de la costa son ligeramente variables, pero por debajo de 15 cm son casi constantes y aproximadamente iguales a la relación de Co/Sc para la corteza terrestre.En la parte superior (0-15 cm) de estos testigos de sedimentos, las relaciones Zn/Sc son más altas (10-20) que las relaciones Zn/Sc para la corteza terrestre, debido, probablemente, al aporte adicional del Zn biogénico particulado desde la capa fótica de la columna del agua.La forma del perfil vertical de las relaciones Zn/Sc del testigo de sedimentos SR22 (360 m de profundidad) frente al área de mayor polución muestra que los niveles máximos de polución ocurrieron en el pasado, y una autopurificación/restauración de los niveles naturales de los contaminantes en los sedimentos finos profundos se puede observar durante los últimos años.
yet been characterised.Cobalt and zinc are environmentally important trace metals (Smith and Carson, 1981).The concept of geochemical provinces has been considered in relation to their regional availability and their transfer through the food chain to humans (Hamilton, 1994).Both cobalt and zinc are essential micronutrients for marine phytoplankton growth.They are supplied to the ocean mainly through estuarine systems from land (Tovar-Sánchez et al., 2004).Interestingly, these two elements can replace one another metabolically in marine phytoplankton that has been cultivated under laboratory conditions (Sunda and Huntsman, 1995).Co and, to a lesser extent, Zn are significantly enriched in the marine sediments of the Gulf of Iskenderen (Turkey) relative to their average crustal abundance.This enrichment is mainly due to the input of the weathered products of basic and ultrabasic source rocks that are present on the adjacent coast and hinterland (Ergin et al., 1996).Sometimes, Co and Zn are greatly enriched in the environment, for example, in fjord sediments near a former mine in West Greenland (Elberling et al., 2002) or in the "Boleo" copper mining district on the eastern coast of the Baja California peninsula, near the town of Santa Rosalía (Wilson and Rocha, 1955;Shumilin et al., 2000;Rodríguez Figueroa, 2004).
Because of natural Cu-Co-Zn mineralization, the Santa Rosalía mining region was extensively used for copper mining and smelting throughout the past century.As a result, local beach sands and coastal marine sediments are strongly contaminated by Cu, Co, Zn and other heavy metals (Shumilin et al., 2000;2005).However, the impacted zones and the penetration depth into the sedimentary column are ostensibly limited in size (Rodríguez Figueroa, 2010;Rodríguez et al., 1998;Shumilin et al., 2000aShumilin et al., , 2000b)).
The objective of this study was to determine the vertical distribution of the anthropogenic contaminants at certain sites that were selected within the impacted area and a supposedly unaffected zone of the adjacent part of the Gulf of California to determine the depth at which the sedimentary column is affected by the pollution and to discover any evidence of self-cleaning processes.Our previous studies of smelting wastes, arroyo sediments, beach sands and surface marine sediments showed that there is a significant presence of Cu, Co, Mn and Zn in contaminated sedimentary materials.However, Co and Zn are especially convenient elements for accomplishing this type of environmental assessment because instrumental neutron activation analysis (INAA) for these elements is a very effective tool to analyse a considerable quantity of subsamples rapidly and inexpensively.Other contaminants, such as Cu and Mn, cannot be measured by the traditional INAA technique and require a laborious strong acid digestion to obtain solutions that can finally be measured by flame atomic absorption spectrophotometry or ICP-AES.Moreover, there have been some indications that the complete digestion of materials originating from a smelter in Santa Rosalía is difficult to achieve (Rodríguez Figueroa, 2010).To minimise any possible effects of grain size and dilution by biogenic calcium carbonate or biogenic silica on the levels of Co and Zn in the sediments, the normalisation of their absolute values with scandium was applied (Dias and Prudêncio, 1998;Grousset et al., 1995;Monna et al., 2004).Scandium belongs to a group of terrigenous indicators (i.e., Al, Cs, Ti, Li, Sc and Th) that are strongly associated with aluminosilicates in soils and sediments, and it was selected for our study because it is the only element of this group that can easily be measured by INAA in a solid matrix with high precision.Moreover, the linear regressions of the levels of many elements versus the levels of Sc in the sedimentary materials from the coastal environments of southern Baja California have displayed the best correlation coefficients when compared with Al, Ti or Fe (Rodríguez Castañeda, 2008).The Co/Sc and Zn/Sc ratio vertical profiles were prepared and taken into consideration to assess the depth of penetration of the contaminants into the marine sediments.

Sampling
The study area and the locations of the sampling stations are shown in Figure 1.The SR1-SR-5 and SR-10 sediment core samples were collected in April 2006 using a manual corer (length 100 cm) by scuba divers from a small plastic motorboat that was used as a platform for separating the core sediment samples.
The coastal core samples were thoroughly extracted from the core and, depending on the results of visual ob- servation of the opened core directly after sampling, were separated into layers (1 cm, 2.5 cm and 5 cm thick) using a plastic knife and a ruler.Then, they were transferred with a plastic spoon into pre-cleaned polyethylene packets and stored in a freezer until processing.In the laboratory, the samples were thawed and split into subsamples before treatment.Two deep-water cores, SR22 and SR62, were sampled in 1994 by a box corer in the Gulf of California in front of the Santa Rosalía port during the "El Puma" cruise (Fig. 1) and were sliced into 2 cm thick sediment subsamples.Related data on the composition of the sediments of SR22 and SR62 cores have been described previously (Shumilin et al., 2000a;Choumiline et al., 2006).

Analysis
The levels of Co, Sc and Zn in the samples of finely ground and dried sediments were determined using INAA (Shumilin et al., 2000a).The samples were dried in an electrical oven at 60°C for 24 h, and homogenised in an agate mortar.Then, 20 mg subsamples, in addition to certified standard reference material IAEA-356 (polluted marine sediment), homemade Russian standard reference materials (RCC-1, SARM-7, ST, SGD, KH, RUS, TB) and the stone meteorite "Allende" were irradiated simultaneously in an experimental nuclear reactor with a flux of thermal neutrons of 2.8×10 13 n s −1 cm −2 at 150°C.After "cooling", the gamma-spectrometric measurements of the irradiated samples were conducted with a semiconductor Ge (Li) detector that was coupled to a 4096-channel high resolution pulse analyser LP-4900 ("Nokia", Finland).
The elemental composition calculations of the irradiated samples were performed using statistical software that was developed ad hoc.This procedure confirmed that the precision (less than 4.6 % for Co, less than 4 % for Sc and less than 13 % for Zn) and accuracy of the determination of the concentration of the elements in the sediments were generally acceptable (Table 1).The detection limits were 0.05 mg kg -1 for Co and Sc and 10 mg kg -1 for Zn.
line sediment quality value at certain layers of the core (7.5-10 cm, 12.5-15 cm, 17.5-20 cm and 22.5-25 cm), but the Zn concentrations were generally lower than the ERL guideline values.

Discussion
The grain size of the coastal sediments from Santa Rosalía mining region is very heterogeneous due to its specific lithology and the occurrence of anthropogenic solid waste dumping (Rodríguez Figueroa, 2010).Generally, mixtures of particles of different sizes and origins are observed on spatial and temporal scales (Rodríguez Figueroa, 2010).To minimise possible effects of grain size variability and dilution of the contaminants in the sediments by inert quartz, by biogenic calcium carbonate or silica, the concentrations of Co and Zn were normalised with Sc.The regional background levels of the Co/Sc and Zn/Sc ratios in the study area and the shapes of the Co/Sc and Zn/Sc ratios vertical profiles were useful to evaluate the depth of penetration of tailing and smelting solid waste particles into the coastal marine sediments.
General information about the values of the Co/Sc and Zn/Sc ratios in the coastal and deep marine sediments in front of the Santa Rosalía mining region is presented in Table 3.The vertical profiles of Co/Sc and Zn/Sc ratios in the sediment cores of this study (cores SR1-5, SR10) as well as SR22 (from Shumilin et al., 2000a) and SR62 (from Choumiline et al., 2006) are shown in Figures 2  and 3.
As shown in Table 3, high mean values of Co/Sc ratio that ranged from 52 to 106 were detected for cores SR1, SR2 and SR4, and they were much lower for cores SR3, SR5 and SR10 (a range of 1.9-2.9).The lowest, most likely background, values were found in cores SR22 and SR62, where they ranged from 1.11 to 1.26.The highest Zn/Sc ratios were also found in SR1, SR2 and SR4 cores (a range of 214-262), followed by the open-sea core SR62 (a range of 11.7-24.1)and then cores SR3, SR5 and SR10 (a range of 8.9-10.1).
The observed Co/Sc ratios in the lower depths of cores SR3, SR5, SR10 and SR22 most likely corresponded to the local background conditions.Surface coastal sediments outside the study area showed slightly higher values of Co/Sc ratio that ranged from 1.6 to 6.0 (mean 3.4±1.2) along the northern margin and from 0.9 to 8.7 (mean 3.8±1.8)along the southern margin, supposedly reflecting the shore littoral transport of the contaminants from the highly polluted areas located near the town of Santa Rosalía.
The shape of the vertical profile Co/Sc ratio of SR22 sediment core (depth 360 m) in front of the main pollution "hot spot" showed that the maximum levels of pollution occurred in the past (see increased Co/Sc ratio values for samples collected from about 5-15 cm subsurface depth interval in Fig. 2), and self-purification/restoration of the natural pollutant levels in the fine grained deep- tively).The Co/Sc ratios for the sediments of the deeper cores were slightly variable, but they were nearly constant and were approximately equal to the Co/Sc ratio of the earth´s crust at sediment depths below 15 cm.The Zn/Sc ratios in the same cores were higher (10-20) than the Zn/Sc ratio of the earth´s crust, which was probably due to the additional input of incompletely mineralised biogenic particulate Zn from the photic zone of the water column.Co/ Sc ratios along core SR22 showed that the maximum levels of pollution occurred in the past, and self-purification/ restoration of the natural pollutant levels in the fine deepwater sediments has occurred during recent years.
water sediments has occurred during recent years (see decreasing Co/Sc values for samples collected from about <5 cm depth in Figure 2).
The Zn/Sc ratios of the surficial sediments from the northern and southern margin did not exhibit the same tendency for the Co/Sc ratios as for the Zn/Sc ratios as was the case for the deeper cores SR22 and SR62.This situation could be due to the lower influence of contamination of the deep sediments in the central portion and slightly elevated contributions of accumulated biogenic particulate Zn from the water column in cores SR22 and SR62.
Cobalt is not a typical contaminant found in the coastal marine sedimentary environment, as it predominantly appears at background levels (Cobelo-García and Prego, 2003).Being mainly of terrigenous origin in coastal sediments, it is usually incorporated into natural resistant aluminosilicate matrices of marine sediments, and for that reason, cobalt was proposed as an additional terrigenous indicator or normaliser to identify the anthropogenic impacts of Cu, Pb and Zn on coastal marine sediments from the continental margin adjacent to Sydney, Australia (Matthai and Birch, 2001).However, certain levels of non-detrital Co in surface sediments from the Bay of Bengal have previously been attributed to anthropogenic contamination (Selvaraj et al., 2004).In contrast, marine sediments formed in the areas of high biological productivity and driven by coastal upwellings are often depleted in Co due to the remobilisation of this element from particles in oxygen-depleted water and sedimentary environments (Brumsack, 2006).
Somewhat higher Zn/Sc ratio values in open-sea sediment cores (Shumilin et al., 2000a;Choumiline et al., 2006) relative to coastal "background" cores are most likely a result of the bioaccumulation of Zn in plankton organisms with the subsequent inclusion of their debris into the sea floor sediments.Additionally, it is hard to say to what extent the nearby Guaymas Basin with its hydrothermal sources influences the Zn concentrations, as direct observations over a hydrothermal area have previously indicated elevated levels of dissolved and particulate Zn no higher than 600 m above a 2000 m deep sea floor (Tambiev and Demina, 1992).

Conclusions
The ratios of Co/Sc and Zn/Sc in cores from the predominant pollution "hot spot" near Santa Rosalía port were very high (40-150 and 150-350, respectively).The values of Co/Sc and Zn/Sc ratios decreased drastically in the cores that were collected outside the main "hot spot" below a core depth of 20-34 cm and approached the regional coastal surface sediment background ratios (approximately 2 and 5-10 for Co/Sc and Zn/Sc, respec-

Fig. 1 .
Fig. 1.-Study area and locations of the sampling sites for sediment cores in the coastal zone of the Santa Rosalía copper mining region in the Gulf of California.Fig. 1.-Área de estudio y localizaciónde los sitios de muestreo de los testigos de sedimentos en la zona costera de la región minera cuprífera de Santa Rosalía en el Golfo de California.