Seasonal cycle of copper speciation in Gullmar Fjord, Sweden

The chemical speciation of dissolved Cu was investigated by voltammetric methods in Gullmar Fjord, Sweden, over the course of a year from September 1996 until August 1997. Sampling was carried out on a roughly monthly basis, with an intensive survey carried out in May 1997. Surface water temperatures ranged from 21 to 228C, whereas bottom waters in the fjord were approximately 68C throughout. Macronutrient concentrations in the fjord during the period of the survey were investigated independently by the Göteborgs och Bohus läns Vattenvårdsförbund (Water Quality Association of Göteborg and Bohus). Surface phosphate concentrations were highest in early spring with low levels (,0.1 mmol kg21) over the late spring and summer. Nitrate and silicate showed a similar pattern to phosphate with the exception of high concentrations encountered in surface waters when low salinity plumes caused by runoff were encountered. A period of calm, sunny weather in January 1997 saw the initiation of the spring bloom some 2 months earlier than usual. Dissolved Cu speciation was dominated by organic complexation (over 99.8%) throughout this study. Strong Cu binding ligands (log K . 12.5) were not detected during the winter or early spring and could be related to the temperature-related seasonal appearance of the cyanobacterium Synechoccocus in these waters. The appearance of the strong Cu ligands led to a decrease in the concentration of free copper, resulting in a seasonal cycle for free copper in the fjord. This is the first study to examine Cu speciation over an annual cycle in a coastal environment. The distribution and chemical speciation of trace metals in the upper water column plays an important role in the community structure and physiology of phytoplankton (Sunda 1994). Speciation is important because only particular chemical forms of a given metal may be biologically available. In the upper water column, the speciation of many biologically active trace metals is controlled by complexation with strong organic ligands (Bruland et al. 1991). In general, when a metal is complexed by an organic ligand, the metal becomes less biologically available because the free metal ions are the most labile to the biota (Sunda 1994; Campbell 1995). Organic complexation also can play an important role in the biogeochemical cycling of these elements in the upper ocean. For most elements, organic ligand concentrations are highest in the euphotic zone (Bruland et al. 1991), which suggests a recent biological source. 1 Present address: FB2 Marine Biogeochemie, Chemische Ozeanographie, Institut für Meereskunde, Kiel D-24105, Germany ([email protected]). Acknowledgments The help and patience of the crew of the Arne Tiselius is gratefully acknowledged. Thanks to David Turner, Karen Anderson, Tonia Sands, Frank Linares, Marie Johansson (all at Analytical and Marine Chemistry, Göteborg), for their help during this work. Extra special thanks to Bengt Karlson (Marine Botany, Göteborg University) and Mats Kuylenstierna (KMF) for sharing their time and experience with identifying phytoplankton species. A very special thank you to Odd Lindahl, Peter Tiselius and the Pelagic Monitoring Group at KMF for all their efforts and contributions. The author is indebted to Pege Schelander at the Göteborgs och Bohus läns Vattenvårdsförbund for data on Gullmar Fjord from 1990 to 2000. This manuscript has been greatly improved by the comments of the associate editor Mary Scranton and two anonymous reviewers. This work was supported by Göteborg Marine Research Centre (GMF) for the use of the Arne Tiselius. The author was funded by a New Zealand Foundation for Research, Science, and Technology Postdoctoral Fellowship (GOT501). Complexation is of particular importance for Cu because it is both an essential micronutrient and extremely toxic. Field studies performed in open ocean and in coastal waters have determined that Cu is complexed by low concentrations of strong ligands with further trace metal buffering by higher concentrations of weaker ligands (van den Berg et al. 1987; Coale and Bruland 1990; Sunda and Huntsman 1991; Moffett 1995). These strong ligands frequently are found at concentrations slightly in excess of the ambient copper concentration in seawater and have conditional stability constants of 1012.5 mol L21 or greater. Currently there is no structural or functional chemistry information on these ligands, and they are generally referred to as ligand class 1. The weaker ligands that are also detected in seawater are similarly classified as ligand class 2 and have conditional stability constants ranging from 1010 to 1012 mol L21. Fieldwork in the Sargasso Sea has shown a strong link between the presence of strong Cu complexing ligands (class 1) and the cyanobacterium Synechoccocus (Moffett et al. 1990; Moffett 1995). Laboratory cultures of Synechoccocus isolated from the Sargasso Sea have also been found to produce strong Cu binding ligands when under Cu stress (Moffett and Brand 1996). The growth of Synechoccocus is very sensitive to Cu, with growth rates reduced at free copper concentrations as low as 10 pmol L21 ([Cu]f 5 10211 mol L21) (Brand et al. 1986). Synechoccocus, in ambient openocean seawater (1–5 nmol L21 Cu) if there were no organic complexation of Cu, would be significantly inhibited. Thus Synechoccocus benefits from the production of class 1 ligands in the presence of elevated Cu concentrations. It has been suggested that in the natural environment, Synechoccocus produces class 1 ligands to create conditions more favorable for its growth. Fieldwork in Massachusetts harbors subject to anthropogenic Cu saw a strong inverse relationship between [Cu]f and cell densities of Synechoccocus, as expected from laboratory studies (Moffett et al. 1997). 765 Cu speciation in Gullmar Fjord Fig. 1. Map of study region, Gullmar Fjord, and regional oceanographic setting (inset). Also shown are the positions of the sampling stations occupied during the course of this work. Table 1. Location of sampling stations in and around Gullmar Fjord. Station Maximum depth (m) Latitude Longitude 1 Skagerrak 2 Björkholmen 3 Alsbäck/Djuphålan 4 Ingela’s 5 Sill/Inloppet 6 KMF 350 61 118 63 39–52 50 58819.79N 58823.79N 58819.49N 58816.59N 58815.99N 58827.09N 10829.19E 11837.19E 11832.89E 11829.19E 11822.09E 11827.09E KMF, Kristineberg Marine Research Station. Gullmar Fjord is a relatively pristine fjord situated on the west coast of Sweden (Fig. 1). The fjord is 30 km long and 3 km wide. The main part of the fjord is straight with steep rocky shores. It has a maximum depth of 120 m and a sill depth of 45 m. The water body in the fjord is always stratified and normally consists of three layers, each of different origin (Lindahl and Hernroth 1983): high-salinity water from the North Sea in the deep basin of the fjord (below 50 m depth, salinity 34–35), water from the Skagerrak forming an intermediate layer (20–50 m, salinity 31–33), and water originating from the Baltic Sea comprising the top layer (salinity 18–30). Wind and barometric pressure changes determine the occurrence and vertical distribution of the different water masses and a thermocline is more or less pronounced from May until September, normally at 15 to 20 m (Lindahl and Hernroth 1983). Exchange of the water in the fjord above the sill depth with water from outside is driven by vertical movements of the halocline outside the fjord and occurs within less than a month (Rydberg 1977; Liljebladh and Thomasson 2001). The annual mean freshwater runoff from land into the fjord is 22 m3 s21, with the main river, Örekilsälven, located at the head of the fjord. This runoff can contain a high humic load, which can turn the upper water layer brown. The deep water in the fjord has an Atlantic origin and is normally renewed once a year in one or several pulses during late winter or early spring. Occasionally, this renewal does not take place and the bottom waters can become suboxic or anoxic (Lindahl and Hernroth 1983; Liljebladh and Thomasson 2001). The spring bloom in Gullmar Fjord typically starts in March, lasts 2–3 weeks, and is dominated by diatoms (Lindahl and Hernroth 1983). Prior to the start of the bloom, macronutrient concentrations are at their highest as winter mixing brings nutrients to the surface and land runoff is high. The spring bloom begins despite occasional ice cover in the fjord (Lindahl and Hernroth 1983). In summer, surface waters are depleted of macronutrients, resulting in almost oligotrophic conditions. The autumn bloom is typically dominated by a large dinoflagellate bloom (Lindahl and Hernroth 1983). Synechoccocus-type cyanobacteria are found in the Skagerrak and Gullmar Fjord year round, and show a seasonal abundance with maximum concentrations in summer (Karlson and Nilsson 1991; Kuylenstierna and Karlson 1994). The aim of the present work was to measure Cu speciation over an annual cycle in the Swedish west coast and, in particular, to examine the effect of seasonal Synechoccocus abundance on Cu speciation.