High resolution monitoring of dissolved Cu and Co in coastal surface waters of the Western North Sea

doi:10.1016/S0278-4343(03)00003-7

Continental Shelf Research

Volume 23, Issue 6, April 2003, Pages 611-623

https://doi.org/10.1016/S0278-4343(03)00003-7 Get rights and content

Abstract

This paper discusses the distribution of dissolved Cu and Co in coastal surface waters of the western North Sea. High resolution data (more than 5500 (Cu) and 12000 (Co) data points), obtained using on-line flow-analysis techniques, clearly showed enhanced trace metal levels near the coast and estuarine outflows. In an offshore direction, the high metal levels were gradually diluted with metal-depleted seawater originating in the North Atlantic. The data indicated benthic sources for Cu and Co due to diagenetic remobilisation, sediment resuspension with injection of interstitial waters into the overlaying water and/or desorption from resuspended particles. Some evidence of biological removal of dissolved Cu was apparent in offshore coastal waters with enhanced chlorophyll concentrations. The study demonstrates how high resolution trace metal data allows us to deconvolute complex metal input and removal processes in dynamic coastal waters.

Introduction

The processes that determine the concentrations and distributions of dissolved trace metals in coastal waters include: mixing of river water (high metal concentrations) with cleaner seawaters, atmospheric inputs, benthic exchange and particle–water interactions, including uptake by planktonic organisms (Lapp and Balzer, 1993; Rivera-Duarte and Flegal, 1994; Tappin et al (1993), Tappin et al (1995)). These various processes may occur simultaneously and are of variable magnitude. The factors are typically poorly quantified, and little is known about the spatial and temporal variability of input and removal processes. Improved knowledge on trace metal concentrations and distributions in coastal seas is important because of the implications for ecosystem functioning. Enhanced Cu inputs into coastal waters have, for example, been implicated in shifts in phytoplankton species composition (Moffett et al., 1997).

The dynamic and complex nature of many coastal waters (including the North Sea), complicates the separation of biogeochemical processes from simple mixing. A high spatial and temporal resolution in measurements is therefore required to resolve the key controlling processes of trace metal behaviour. The commonly used approach of discrete sampling from survey vessels using Go-Flo or Niskin samplers from a hydrowire or rosette frame does not provide the required resolution. Recently, underway sample collection and automated voltammetric shipboard analysis has been used to study dissolved trace metals in estuarine and coastal waters (van den Berg and Achterberg, 1994; Braungardt et al., 1998; Whitworth et al., 1998; Achterberg et al., 1999, Achterberg and Braungardt, 1999). The methods were based on automated batch-analysis procedures, with a rate of analysis of ∼3 samples h⁻¹.

In this study we used a novel flow-analysis system, combined with cathodic stripping voltammetry (CSV) (Colombo et al., 1997). This approach allows high-resolution on-line determination of dissolved trace metals in seawater pumped on-board a ship, at a rate of > 30 samples h⁻¹. At this rate of data collection it is possible to compare trace metals directly with nutrient and physical data collected during the cruises. In this contribution, high-resolution data are presented for dissolved Cu and Co in coastal surface waters of the western North Sea. The data were obtained during research cruises in October 1994, April 1995 and June 1995 on-board R.R.S. Challenger (cruises CH115a, CH118a and CH119a, respectively) as part of the Land Ocean Interaction Study (LOIS) programme of the National Environmental Research Council (NERC).

Section snippets

Study area

Two different transects were undertaken along the east coast of England during the LOIS surveys: grid 1 was north of the Humber Estuary (53.8°N) and was sampled in a northerly direction to 56°N (Fig. 1A), and grid 2 was sampled in a southerly direction from 53.8° to 52.8°N (Fig. 1B). The track followed a zigzag pattern in a zone between 1 and 20 mile offshore in grid 1, and a box-like pattern in grid 2.

On-line analyses

Surface seawater was pumped on-board ship at a rate of 4 l min⁻¹ using a peristaltic pump and a

Copper and Co distributions

Circa 3100 data points were collected for Cu at a frequency of 30 measurements h⁻¹ during CH115a (October 1994), and 2500 data points at a frequency of 40 measurements h⁻¹ during CH119a (June 1995). A total of 12000 Co data points were recorded at a frequency of 60 measurements h⁻¹ during cruise CH118a (April 1995). Grid 1 was repeated twice and grid 2 three times, during CH118a. Measurements were therefore obtained at the same geographical locations but in different waters due to tide and wind

Concentrations and distributions of Cu, Co and ancillary variables

Lowest Cu and Co concentrations were apparent in offshore areas and northeastern parts of the coastal grid (Fig. 2), corresponding with higher salinities as a result of the inflow of Atlantic Ocean waters into the North Sea (Hydes and Kremling, 1988). When approaching the nearer shore and estuarine regions, metal, nitrate and SPM concentrations increased indicating the importance of near coastal sources. These sources were also apparent during the zigzag cruise track pattern of the ship in the

Conclusions

Estuarine and coastal waters are highly dynamic systems, with strong temporal and spatial variations in constituent concentrations. In order to unravel trace metal input and removal processes in the waters of the western North Sea, high resolution measurement techniques were used in this study. With this novel approach a thorough investigation of relationships between trace metals and other oceanographic variables was possible, at a level of detail approaching that used by physical

Acknowledgements

British Gas plc financially supported the research of C.C. The British Oceanographic Data Centre provided the physical data of the cruises. The cruises were financially supported by the LOIS programme of the NERC, and our research was supported by a research grant (GST/02/751) of the NERC. The assistance by the Principal Scientists (R. Howland, PML; D. Prandle, POL; R. Lowry, BODC) and the officers and crew of RRS Challenger are greatly appreciated. The authors thank Dr. A.D. Tappin and two