Distribution and behavior of many trace elements in the aquatic environment has been well characterized, but little is known about silver (Ag) concentrations in coastal waters, even though this element ranks among the most toxic to marine invertebrates (Calabrese et al., 1977 ; Fisher and Hook, 1997 ; Webb and Wood, 1998). Studies conducted by Flegal et al. (1995), River-Duarte et al. (1999), and Ndung'u et al. (2001), provided the first valuable data on Ag distribution in the oceanic environment, indicating that this element is found in very low concentrations in the dissolved phase. However, although silver concentrations in coastal waters do not reach the nanomolar range (Smith and Flegal, 1993 ; Squire et al., 2002), formation of a stable chloro complex enhances bioavailability and toxicity to biota (Luoma et al., 1995). Experimental studies have shown that Ag is toxic to some living organisms at environmentally realistic levels (Bryan and Langston, 1992). Silver found in the aquatic environment mainly originates in effluents from sewage treatment plants (Rozan and Hunter, 2001). Silver can therefore be used as a tracer of wastewater discharges in coastal waters (Martin et al., 1988 ; Sañudo-Wilhelmy and Flegal, 1992), for instance through the use of sentinel organisms, which concentrate bioavailable contaminants in their tissues (Stephenson and Leonard, 1994 ; Jiann and Presley, 1997 ; Riedel et al., 1998 ; Muñoz-Barbosa et al., 2000). This study concerns biological monitoring as a means of providing a synoptic view of silver contamination in French coastal waters. The National Network for the Observation of Marine Environment Quality (RNO, the French Mussel-Watch) which has been regularly measuring concentrations of various chemical contaminants in oyster and mussel tissues for 25 years (Claisse, 1989), has been monitoring silver levels since 2003. This valuable database including data collected at 80 sampling sites distributed along the French coasts (Fig. 1), is used as a reference to provide the spatial distribution of a given contaminant (Chiffoleau and Bonneau, 1994), identify trends of contamination/decontamination (Chiffoleau et al., 2001), and detect peak concentrations due to accidental events (Chiffoleau et al., 2004). Mussels (Mytilus edulis and Mytilus galloprovincialis) and oysters (Crassostrea gigas) are collected twice a year in February and November. Sample collection (size of samples, size of animals) and treatment (cleaning, depuration, removal of soft parts from the shells, draining, homogenization, and freeze-drying) are performed according to the OSPAR Convention guidelines and the method described by Claisse (1989).

In January and February 2002, the presence of certain agricultural pesticides throughout the coastline of the Caribbean island of Martinique was investigated. The tropical climate of the French West Indies is suitable for banana production, which requires intensive use of pesticides. An inventory of all pesticides used on the island (compounds and tonnage) was compiled. Surveys and analyses revealed the presence of pesticides in the plumes of seven rivers. The organochlorine chlordecone and metabolites of aldicarb were detected at nearly all of the monitored sites, even though the use of chlordecone has been prohibited since 1993. Two triazines (ametryn and simazine) were also identified. The concentrations of carbamates and triazines detected in the water and sediment samples from Martinique are comparable to those reported for mainland France. Chlordecone concentrations in the sediment and particulate matter samples were, however, particularly high in the samples from Martinique. Toxicological implications are discussed. Of particular concern are the high levels of chlordecone (which is bioaccumulating and carcinogenic) and further monitoring of this compound is recommended, especially in fish and other sea-food products.