SciELO - Scientific Electronic Library Online

 
vol.39 número3Estudios del bentos de arrecifes naturales y artificiales de Mar del Plata, Argentina, Atlántico sudoccidentalAsentamiento postlarval en la langosta espinosa, Panulirus argus (Latreille, 1804) (Decapoda: Palinuridae) en la costa Caribe de Costa Rica índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Latin american journal of aquatic research

versión On-line ISSN 0718-560X

Lat. Am. J. Aquat. Res. vol.39 no.3 Valparaíso nov. 2011

 

Lat. Am. J. Aquat. Res., 39(3): 567-574, 2011
DOI: 10.3856/vol39-issue3-fi1lltext-16

Research Article

 

Dissolved trace metals in the water column of Reloncaví Fjord, Chile

Metales trazas disueltos en la columna de agua en el fiordo Reloncaví, Chile

 

Ramón Ahumada1, Anny Rudolph1, Elizabeth González1, Gary Fones2 Gonzalo Saldías3 & Ramón Ahumada-Rudolph3

1Department of Environmental Chemistry, Facultad de Ciencias Universidad Católica de la Santísima Concepción, P.O. Box 297, Concepción, Chile
2School of Earth and Environmental Sciences, University of Portsmouth, UK
3
Escuela de Postgrado, Magíster en Oceanografía. Universidad de Concepción, Concepción, Chile

Dirección para correspondencia


ABSTRACT. We analyzed the concentration of dissolved trace metals (Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb) in the water column of Reloncaví Fjord. Sampling was performed during the CIMAR 12 Fiordos cruise in 2006. A total of 36 passive samplers or DGTs (diffusion gradient in thin films) were anchored at four stations along the longitudinal axis of the fjord. The DGTs were deployed at three depths per station and left there for 48 h. The metal contents on each thin film were analyzed using inductively coupled plasma atomic emission spectroscopy. Concentrations were highest in the surface layer at the head of the estuary, which is directly influenced by Petrohué River. Characteristic sequences of the studied metals were defined in the area with the greatest continental influence (Z(5-25m) = Cu >Mn> Fe > Ni >Pb> Cr > Cd > Co) and in the area with a marine or coastal influence (Z(5-25m) = Fe > Cu>Mn> Ni >Pb> Cr > Cd > Co). A similar metal sequence was found in the deepest layer: Z(40-m) = Fe >Mn> Cu >Pb> Ni > Cd > Cr > Co. The passive sampling technique using DGTs to determine dissolved trace metals in the sea water provided robust information on the concentrations of the ten metals analyzed.

Keywords: dissolved metals, passive samplers, fjords, southern Chile.


RESUMEN. Se analiza la concentración de metales trazas disueltos (Cd, Co, Cr, Cu, Fe, Mn, Ni y Pb), en la columna de agua del fiordo Reloncaví. El muestreo se realizó durante la campaña CIMAR 12 Fiordos, 2006. Para ello se fondeó en cuatro estaciones y en tres profundidades, un total de 36 muestreadores pasivos o DGT (láminas de gradiente de difusión) a lo largo del eje longitudinal del fiordo, durante 48 h. El contenido de metales en cada lámina fue analizado mediante espectroscopía de emisión atómica con acoplamiento inductivo de plasma. Las mayores concentraciones se observaron en la superficie de la columna de agua, en la cabeza del estuario, directamente influenciada por el río Petrohué. Se definió para el área una secuencia de los metales estudiados característica, para la zona de mayor influencia continental: Z(5-25m) = Cu > Mn > Fe > Ni > Pb > Cr > Cd > Co y para la zona de influencia marina o costera: Z(5-25-m) = Fe > Cu > Mn > Ni > Pb > Cr > Cd > Co, observándose una secuencia semejante para los metales analizados en la capa de mayor profundidad: Z(40 m) = Fe > Mn > Cu > Pb > Ni > Cd > Cr > Co. La técnica de muestreo pasiva, con DGTs, para la determinación de metales traza disueltos en agua de mar, permitió obtener información robusta de las concentraciones de los diez metales analizados.

Palabras clave: metales disueltos, muestreadores pasivos, fiordos, sur de Chile.


INTRODUCTION

them as natural enrichment or pollution (Regnier & Wollast, 1993; Salamanca & Jara, 2003; Navarro et al., 2006). Trace metals include a highly toxic group that requires monitoring. These metals are most easily absorbed and bioaccumulated by living organisms when the metals are in their dissolved (ionic) form.

Baselines of metals are established to determine the natural concentrations and ranges of variations for these elements or compounds, in order to elucidate any changes that may be produced and to identify Sea water contains low concentrations L-1 or ng L-1) of most metals in solution. Their solubility depends on the solvent characteristics and the solubility constant for each elements, which determines its proportionality and abundance. These can be studied both in the water column and in interstitial waters (Davison et al., 1991, 1994, 1997, 2000; Zhang & Davison, 1995; Fones et al., 1998; Pizarro et al., 2001).

Metal concentration in sea water can be modified regionally through volcanic activity, metheorization and/or erosion of the detritus phases of the rocky basement, and fluvial contribution (Humbord et al., 1997). Although rivers transport small amounts of dissolved metals, these can alter the natural proportions in local sea water by dilution or enrichment of some major and minor cations (Ca, Na, B, Mg, Ba, etc.) (Libes, 2009).

The variability of metal concentration in water is highly uncertain, due to the fact that samples are obtained through instantaneous sampling. At present, a wide gamut of continuous recording systems is available for physical, biological, and chemical measurements. However, the study of the chemistry of dissolved metals is highly complex, involving difficulties of sampling (i.e., obtaining, transporting, storing, and pre-treating the samples), low concentrations (near the detection limit of the analytical equipment) of dissolved ions in this environmental matrix, and the risk of contamination. Every stage of such research offers the potential for sample contamination and the possibility of altering the results of the analytes under study (González et al., 2004).

The fjords of southern Chile are highly dynamic systems, whether due to their dimensions, morphology, tidal activity, and/or important freshwater contributions from the continent. The input of freshwater stratifies the water column, creating a low salinity surface layer and resulting in the dynamics of estuarine behavior (Silva et al., 1998; Dávila et al., 2002; Sievers & Silva, 2008). A fast-moving surface layer moves out of the basin and a slower, deep layer moves inward to the head (Salinas & Hormazábal, 2004; Cáceres & Valle-Levinson, 2004; Valle-Levinson et al., 2007). This generates changes in the physical and chemical variables of the water column (Sievers, 2008; Silva, 2008), particularly organic matter contents, temperature, pH, Eh, and/or ionic strength, amongst others.

Dissolved trace metal information for the Chilean coastal waters has been scarcely reported on the scientific literature, and only for few metals (Jara & Salamanca, 2003; Valdés et al., 2006; Ahumada et al., 2008). The first information on dissolved metal contents in the fjords of southern Chile was obtained in 1995, with the beginning of the CIMAR Fiordos Program. The analytic difficulties did not allow to make the analyses in all cruises to the fjords areas. The present work analyzed the dissolved fraction of eight elements (Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb), determining their vertical distribution and variability along the Reloncaví Fjord water column. A passive sampling technique was used in which devices with diffusion gradients in thin film (DGT) were anchored at different depths in the water column for a known period of time (48 h) in order to produce the in situ pre-concentration of the metal ions.

Study Area

Reloncaví Fjord is located at 41°38'22''S, 72°22'40''W, and has a west-east orientation (Fig. 1). This fjord averages 56 km in length and 3 km in width, with a water surface area of ~170 km2. The first third of the estuary (nearest to the head) is 20 to 100 m deep, dropping rapidly to 150 m in the second third and 400 m in the last third; the depth at the mouth is ~450 m. The interior tides fluctuate between 6 and 7 m at syzygies tides and are as much as ~1 m lower during quadrature (León, 2005; Valle-Levinson et al., 2007). Local winds blow from the south (southeast and southwest) in spring and summer and from the north in autumn and winter (Molinet et al., 2003).


Figure 1. Geographic location of the Reloncaví Fjord and positions of the DGT anchoring stations during the CIMAR 12 Fiordos cruise in October 2006.

Figura 1. Ubicación geográfica del Estuario Reloncaví y posicionamiento de las estaciones de anclaje de los DGT, durante el crucero CIMAR 12 Fiordos en octubre 2006.

The fjord has three important tributaries - the Petrohué River at the head (Q = 278 m3 s-1) the Cochamó river (Q = 20 m3 s-1) and Puelo river (Q = 670 m3 s-1) in the middle - as well as other, smaller tributaries (Anuario Estadístico de Chile, 2007) (Fig. 1). Sediment grain size is mostly silt-mud (63%) and clay (37%) at the head, silt in the central part, and silt and fine sand with a high percentage of fine sand (48%) near the mouth. Organic matter contents in the fjord fluctuate between 5.1 and 6.9% (Pineda, 2009; Silva et al., 2009). Important salmon and mussel farming activities are developed along the fjord coastline, and mollusks are also harvested from natural banks.

MATERIALS AND METHODS

Sampling was performed during the oceanographic CIMAR 12 Fiordos cruise in October 2006, on board the "AGOR Vidal Gormaz" of the Chilean Navy. The experimental design of sampling considered four stations and 36 passive samplers or DGTs (diffusion gradient in thin films) (Davison, 2000; Pizarro et al., 2001; Richir et al., 2011). The four stations were located along the longitudinal axis of the fjord, from its head to its mouth (Fig. 1). For this, we prepared and dropped anchor lines having a dead weight at the bottom, a subsurface buoy, and a surface buoy (Fig. 2a). Three unit of DGTs were inserted in a plate (Fig. 2b), and then three plates were located along the anchor line at three diferent sampling depths i.e., surface (5-10 m), mid-water (25-35 m), and deep water (40-50 m).


 

Figure 2. a) Diagram of DGT set-up and anchoring, b) plate detail, c) DGT capsule.

Figure 2. a) Diagrama de anclaje de los DGTs, b) detalle de una placa, c) cápsula de un DGT.

The deployment and recovery operations were done according to Neshyba & Fonseca (1981), and a Zodiac boat was used for anchoring the four lines between 9 and 11 October 2006. The DGTs were left in place for 48 h and reference data of salinity and temperature were obtained with a CTD, at each station during the anchoring and when DGTs were recovered (Madrid & Zayas, 2007). CTD data was used in the computation of metal concentrations.

The anchoring at station 4 was lost due to bad weather conditions. Only three stations were completed successfully, stations 5, 6 and 7. Station 7 covered the first fourth of the fjord length at its head; station 6 covered the second fourth (the Cochamó and Puelo rivers empty into this section) and station 5 covered the third fourth of the fjord length.

Each plate contained three thin sheets or DGTs, which were considered to be replicates per depth (Fig. 2b). The DGTs were recovered, removed (Fig. 2c), washed with MilliQ water, and stored individually in labeled Ziploc® bags at 4°C until their analysis. The analytical determination of the 8 metals in the sea water, i.e., Fe, Mn, Cd, Co, Cr, Cu, Ni, Pb, was done using atomic emission spectroscopy with inductively coupled plasma (AES-ICP) in clean laboratories of the University of Portsmouth (UK).

These data were tabulated, and the statistical software package Statistica version 6.0 (StatSoft. Inc., 2001) was used for their statistical description.

RESULTS

Table 1 shows the results of the average concentration of the metals in solution measured at three of the sampling stations. The study shows that the surface layer of the water column at the head of the estuary (St. 7) was directly influenced by contributions from the Petrohué River, with a temperature of 10°C and salinity < 20 psu. The water column showed strong vertical stratification and, below 20 m depth, temperatures of 12°C and higher salinity (28 psu) than in the surface layer were found. In the central part of the fjord, despite contributions from the Cochamó (St. 7) and Puelo (St. 6) rivers, temperatures were near 12°C at the surface and 10,5°C at 40 m. Salinity was 28 psu, at station 7, increasing along a gradient with depth until reaching 32 psu at surface salinity near station 5.


Data showed that most of the metals studied (Cd, Co, Cr, Cu, Mn, Ni, Pb) were higher in the surface layer (5 m) at the head of the fjord (St. 7), which is influenced by the Petrohué River. The only exception to these findings was Fe. The concentration of metals at the head of the estuary was ~5 times greater than the rest of metals, showing a decreasing gradient towards the mouth.

The Cochamó and Puelo rivers enter the fjord around St. 7 and 6, respectively, but did not have the same effect on metal concentrations. Instead, these rivers contributed to dilute rather than increase metal concentrations. This suggests that the behavior of the metals is different along the longitudinal axis of the fjord. Likewise, metal concentrations in the intermediate water layer (25 and 35 m depth), which had intermediate salinity values, were lower than in the upper and deeper layers, as found at all three stations (Table 1).

DISCUSSION

The main forcing factors in the Reloncaví Fjord are freshwater contributions (Petrohué, Puelo, Cochamó rivers), tidal differences (6-7 m), and strong local winds. These factors result in high variability within the fjord in terms of the spatial and temporal distribution of physical, chemical, and biological properties of its waters (Silva et al., 1998; Dávila et al., 2002; Salinas & Hormazábal, 2004; Cáceres & Valle-Levinson, 2004; Valle-Levinson et al., 2007). Given this context, it was decided to use a passive technique that would allow to obtain an in situ concentration of the dissolved metals over a 48 h time period. The disadvantage of this system is the high probability of losing the anchor lines, as it occurred at the station four at the mouth of the estuary.

In terms of absolute abundance, Co, Cr, and Cd showed the lowest concentrations in all the stations sampled at the estuary. The concentrations of Cu and Mn were higher than other metals in the surface layer at the head of the fjord, showing a characteristic abundance sequence pattern. The concentration of Fe increased towards the fjord mouth and as well as in depth.

The temperature and salinity were used to identify the thickness of the stratified layer in the fjord (Table 1).The upper 25 m correspond to a layer of estuarine influence with salinities between 22-30 psu. The deeper layer corresponds to a layer of marine influence, with salinity near 30 psu and reaching 32 psu at the mouth of the fjord (St. 4). This vertical salinity gradient was observed in all the stations and a horizontal gradient was observed from the head of the fjord to its mouth. Given the relative abundance of metals and the stratification of the fjord, we defined areas of concentration distribution.

The metal concentrations in the deep layer were lower than in the surface one, but higher than in the intermediate layer, with the exception of Fe, which was first in abundance at 40 m depth. This concentration distribution, allowed to identify two groups according to their relative abundance:

1. The area of continental influence covered the fjord's first quarter (St. 7). The sequence at the surface (5 and 25 m) was Z(5-25 m) = Cu > Mn > Fe > Ni > Pb > Cr > Cd > Co and at depth (40 m) was Z(40-m) = Fe > Mn > Cu > Pb > Ni > Cd > Cr > Co.
2. The area of marine influence included two quarters at the center of the estuary (St. 5 and 6). The sequence at the surface (5 and 25 m) was Z(5-25 m) = Fe > Cu > Mn > Ni > Pb > Cr > Cd > Co and at depth (40 m) was Z(40 m) = Fe > Mn > Cu > Pb > Ni > Cd > Cr > Co.

The secuences pattern of metals abundances in the deep layer (40 m) were similar in both of these areas (continental and marine influence). Dissolved Fe was the most abundant element for the area, in which salinity was higher and velocity lower in this deep layer, which circulates towards the head of the estuary (Salinas & Hormazábal, 2004; Cáceres & Valle-Levinson, 2004).

Table 2 shows the whole column average data for the continental or estuarine area and for the marine influence or coastal area (the fjord empties into the Interior Sea of Chiloé). The greater concentrations of Cu in the area of continental influence (less salty) could be associated with the formation processes of hydroxides colloids or soluble complexes in this area, which has higher pH (differences of pH from freshwater to seawater).


The concentrations of Ni and Cu in the sediments indicate a risk zone at the head of the fjord (St. 7), unlike the situation observed in the water column.

Finally, for comparative purposes and reference, it is presented the concentration values for dissolved metals reported by other authors at different sites (Table 3). The values of Cd, Co, Cr, Mn, Ni, and Pb found in the Reloncaví Fjord were near to or lower than those found in the other sites studied. Nonetheless, the concentrations of Cu, were greater in the surface layer at the stations near the head of the fjord, and Fe was more concentrated towards the mouth and in the deeper parts of the fjord. The marked differences in the concentrations of Cu suggest an additional input of these metals from the continent.


CONCLUSIONS

The dissolved metal concentrations were high for Cu, Mn and Fe, and low for Co, Cr, and Cd. Moreover, the concentrations were the greatest in the surface layer of the water column and at the head of the fjord, a lower-density area due to the influence of the Petrohué River.

The dissolved metal abundances revealed a characteristic sequence for the area of continental influence (Z(5-25 m) = Cu > Mn > Fe > Ni > Pb > Cr > Cd > Co) different from that of the area of marine or coastal influence (Z(5-25 m) = Fe > Cu > Mn > Ni > Pb > Cr > Cd > Co). The two areas shared a similar sequence in the deep layer: Z(40 m) = Fe > Mn > Cu > Pb > Ni > Cd > Cr > Co.

The use of DGTs devices for sampling and the AES-ICP analyses shown to be adequate; its results were not significantly different from those obtained by DPASV with HMDE. Thus, our data are consistent with the values that fluctuate according to the characteristics and dynamics of the sampling site.

The passive samplers (DGTs) allowed to determine dissolved trace metal contents in the water column in an area of high variability and to obtain robust information on the concentrations of the ten metals analyzed.

ACKNOWLEDGEMENTS

The authors thank the Comité Oceanográfico Nacional for financing the project CONA C12F 06-01; the captain, crew, and divers of the "AGOR Vidal Gormaz" for help during sampling; and the analysis laboratories: Laboratory of Geochemistry (School of Earth and Environmental Sciences, UK) and Laboratorio de Metales Traza. Facultad de Ciencias, Universidad Católica de la Santísima Concepción (UCSC) and Dra. Gina Rojas for the revision of english text. Finally we thank to the Dirección de Investigación of UCSC by the facilities offer to the final translation of the paper.

 

REFERENCES

Ahumada, R., L. Gerli, E. Fernández & I. Sánchez. 2008. Trace metals in the coastal waters and organisms of austral Chilean chanels and fjords. In: N. Silva & S. Palma (eds.). Progress in the oceanographic knowledge of Chilean interior waters, from Puerto Montt to Cape Horn. Comité Oceanográfico Nacional, Pontificia Universidad Católica de Valparaíso, Valparaíso, 45-49.         [ Links ]

Baeyens, W., M. Elskens, R. Van Ryssen & M. Leermakers. 1998. The impact of the Scheldt input on the trace metal distribution in the Belgian coastal area (results of 1981-1983 and 1995-1996). Hydrobiología, 366: 91-108.         [ Links ]

Beltrame, M.O., S.G. de Marco & J.E. Marcovecchio. 2009. Dissolved and particulate heavy metals distribution in coastal lagoons. A case study from Mar Chiquita Lagoon, Argentina. Estuar. Coast.Shelf. Sci., 85: 45-56.         [ Links ]

Botté, S.E., R.H. Freije & J.E. Marcovecchio. 2008. Dissolved heavy metal (Cd, Pb, Cr, Ni) concentration in surface water and porewater from Bahía Blanca estuary tidal flats. Bull. Environ. Contam. Toxicol., 79: 415-421.         [ Links ]

Cáceres, M. & A. Valle-Levinson. 2004. Transverse variability of flow both sides of a sill/contraction combination in a fjord like inlet of southern Chile. Estuar. Coast. Shelf. Sci., 60: 335-338.         [ Links ]

Dávila, P., D. Figueroa & E. Müller. 2002. Freshwater input into the coastal ocean and its relation with the salinity distribution off austral Chile (35-55°S). Cont. Shelf Res., 22: 521-534.         [ Links ]

Davison, W., H. Zhang & G.W. Grime. 1994. Performance characteristics of gel probes used for measuring pore waters. Environ. Sci. Technol., 28: 1623-1632.         [ Links ]

Davison, W., G.R. Fones & G.W. Grime. 1997. Dissolved metals in surface sediment and a microbial mat at 100 mm resolution. Nature, 387(6636): 885-888.         [ Links ]

Davison, W., G.R. Fones, M. Harper, P. Teasdale & H. Zhang. 2000. Dialysis, DET and DGT: in situ diffusional techniques for studying water, sediments and soils. In: J. Buffle & G. Horvai (eds.). In-situ monitoring of aquatic systems: chemical analysis and speciation. IUPAC, Wiley, New York, pp. 495-569.         [ Links ]

Davison, W., G.W. Grime, J.A.W. Morgan & K. Clarke. 1991. Distribution of dissolved iron in sediment pore waters at sub-millimetre resolution. Nature, 352: 323-325.         [ Links ]

Fang, T-H. & C-L. Lin. 2002. Dissolved and particulate trace metals and their partitioning in a hypoxic estuary: the Tanshui estuary in northern Taiwan. Estuaries, 25: 598-607.         [ Links ]

Fones, G.R., W. Davison & G.W. Grime. 1998. Development of constrained DET for measurements of dissolved iron in surface sediments at sub-mm resolution. Sci. Total Environ., 221: 127-137.         [ Links ]

González, E., R. Ahumada, V. Medina, J. Neira & U. González. 2004. Thermospray flame furnace atomic absorption spectrometry (TS-FF-AAS): application for total determination of Cd, Pb and Zn in fresh waters, seawater and marine sediments. Quím. Nova, 27(6): 873-877.         [ Links ]

Hatje, V., S. Apte, C. Hales & G.F. Birch. 2003. Dissolved trace metal distributions in port Jackson Estuary (Sydney Harbor) Australia. Mar. Pollut. Bull., 46: 719-730.         [ Links ]

Humbord, C., V. Itteekkot, A. Cociasu & B.V. Bodungen. 1997. Effect of Danube River dam on Black Sea biogeochemistry and ecosystem structure. Nature, 386: 385-388.         [ Links ]

Jara, B. & M. Salamanca. 2003. Disponibilidad ambiental de cobre en aguas de Bahía Coliumo. Cienc. Tecnol. Mar, 26(1): 3-11.         [ Links ]

León, J.E. 2005. Influencia del caudal del río Puelo sobre la salinidad y la concentración de oxígeno disuelto en el estuario Reloncaví, Llanquihue, Chile. Tesis de Magíster, Universidad Austral de Chile, Valdivia, 80 pp.         [ Links ]

Libes, S.M. 2009. Introduction to marine biogeochemistry. Elsevier, New York, 910 pp.         [ Links ]

Madrid, Y. & Z. Pedrero-Zayas. 2007. Water sampling: traditional methods and new approaches in water sampling strategy. Trends Anal. Chem., 26(4): 293-299.         [ Links ]

Mendiguchía, C., C. Moreno & M. García-Vargas. 2007. Evaluation of natural and anthropogenic influences on Guadalquivir River (Spain) by dissolved heavy metals and nutrients. Chemosphere, 69: 1509-1517.         [ Links ]

Molinet, C., A. Lafont, G. Lembeye & C.A. Moreno. 2003. Patrones de distribución espacial y temporal de floraciones de Alexandrium catennela (Whedom & Kofoid) Balech 1985, en aguas interiores de la Patagonia noroccidental de Chile. Rev. Chil. Hist. Nat., 76: 681- 698.         [ Links ]

Navarro, E., L.W. Dásele, V.F. Camaño, M.C. Ortiz-Hernández & E.A. Gutiérrez-Galindo. 2006. Geoquímica de Fe, Ti y Al como indicadora de la sedimentación volcanicoclástica en la laguna costera de San Quintín, Baja California. Cienc. Mar., 32(2): 205-217.         [ Links ]

Neshyba, S. & T. Fonseca. 1981. Corrientes costeras. Manual de mediciones y análisis. Invest. Mar. Valparaíso, 7: 1-132.         [ Links ]

Pizarro, J., M. Rubio & A. Lira. 2001. Optimisation of an in situ sampling technique: analysis of Cu and Fe in aquatic system. Bol. Soc. Chil. Quim., 46(3): 281-285.         [ Links ]

Regnier, P. & R. Wollast. 1993. Distribution of trace metals in suspended matter of the Scheldt. Estuary Mar. Chem., 43: 3-19.         [ Links ]

Richir, J., L.N. Serpe, P. Lefèbvre, L. Deraikem, A.L. Gilles, B. Renzo & G. Sylvie. 2011. Combined utilization of DGTs and bioindicators to trace chemical contamination threats on coastal ecosystems. 43rd Colloquium on Ocean Dynamics. Université de Liège, Belgique. http://modb.oce.ulg.ac.be/colloquium.         [ Links ]

Salamanca, M.A. & B. Jara. 2003. Distribución y acumulación de plomo (Pb y 210Pb) en sedimentos de los fiordos de la XI Región. Chile. Cienc. Tecnol. Mar, 26(2): 61-71.         [ Links ]

Salinas, S. & S. Hormazábal. 2004. Capacidad de transporte de la constricción de Meninea para un flujo de dos capas y el efecto de la corriente de marea. Cienc. Tecnol. Mar, 27: 5-15.         [ Links ]

Sievers, H. 2008. Temperature and salinity in the austral chilean channels and fjords. In: N. Silva & S. Palma (eds.). Progress in the oceanographic knowledge of Chilean interior waters, from Puerto Montt to Cape Horn. Comité Oceanográfico Nacional - Pontificia Universidad Católica de Valparaíso, Valparaíso, pp. 31-36. Book on line at http://www.cona.cl/.         [ Links ]

Sievers, H. & N. Silva. 2008. Water masses and circulation in austral Chilean channels and fjords. In: N. Silva & S. Palma (eds.). Progress in the oceanographic knowledge of Chilean interior waters, from Puerto Montt to Cape Horn. Comité Oceanográfico Nacional - Pontificia Universidad Católica de Valparaíso, Valparaíso, pp. 53-58. Book on line at http://www.cona.cl/        [ Links ]

Silva, N. 2008. Dissolved oxygen, pH, and nutrients in the austral Chilean channels and fjords. In: N. Silva & S. Palma (eds.). Progress in the oceanographic knowledge of Chilean interior waters, from Puerto Montt to Cape Horn. Comité Oceanográfico Nacional, Pontificia Universidad Católica de Valparaíso, Valparaíso, pp. 37-43. Book on line at http://www.cona.cl/.         [ Links ]

Silva, N., C. Calvete & H. Sievers. 1998. Masas de agua y circulación general para algunos canales entre Puerto Montt y laguna San Rafael Chile (Crucero CIMAR Fiordo 1). Cienc. Tecnol. Mar, 21: 17-48.         [ Links ]

Silva, N., J. Haro & R. Prego. 2009. Metals background and enrichment in the Chiloé Interior Sea sediments (Chile). Is there any segregation between fjords, channels and sounds? Estuar. Coast. Shelf Sci., 82(3): 469-476.         [ Links ]

StatSoft. Inc. 2001. Statistica (data analysis software system). www.statsoft.com.         [ Links ]

Valdés, J., D. Román, P. Dávila, L. Ortlieb & M. Guíñez. 2006. Variabilidad estacional de cadmio en un sistema de surgencia costera del norte de Chile (Bahía Mejillones del Sur, 23°S). Rev. Chil. Hist. Nat., 79: 517-535.         [ Links ]

Valle-Levinson, A., N. Sarkar, R. Sanay, D. Soto & J. León. 2007. Spatial structure of hydrography and flow in a Chilean fjord, Estuario Reloncaví. Estuar. Coast. Shelf. Sci., 30(1): 113-126.         [ Links ]

Vásquez, G.F., R.A. Díaz & L.G. Salvador. 1998. Dissolved metals in Alvarado Lagoon, México. Environ. Int., 24: 721-727.         [ Links ]

Zhang, H. & W. Davison. 1995. Performance characteristics of diffusion gradients in thin films for the in-situ measurement of trace metals in aqueous solution. Anal. Chem., 67(19): 3391-3400.         [ Links ]

 

Received: 7 June 2010; Accepted: 12 October 2011

Corresponding author: Ramón Ahumada (rahuma@ucsc.cl)