Speciation and origin of particulate copper in runoff water from a Mediterranean vineyard catchment

Speciation and origin o particulate copper in runoff water rom a
Mediterranean vineyard catchment

Fungicide treatments have led to large copper contents of the topsoils of most vineyards. This paper examines the contamination of surface waters by copper in a Mediterranean wine-growing catchment. Its aims were to characterise the forms of copper associated
with suspended matter during a heavy autumn storm event and to identify which soils contribute the most to the copper exports.

1. Introduction
Copper has been used to protect vines against fungus disease since the end o the eighteenth century ( affor-gue,1928). Even today, a high proportion of ungicides
are in the form of copper compounds. It is known that such inputs tend to accumulate in the surface layers of the soil, and in turn the topsoils of most vineyards contain large amounts of copper (Flores Velez, 1996).
However, copper is only partially trapped in the surface horizons since many vineyards are subject to water erosion phenomena (Vogt et al., 1986; itzler et al., 1988). These can wash the copper to downstream crops or ecosystems. Depending on the speciation of copper in runoff water, the downstream physicochemical conditions (organic complexants, pH, ionic strength, etc.) are likely to favour its solubilisation (Serpaudet
al., 1994; Dawson and Macklin, 1998), and thereby its toxicity. It is therefore of interest to characterise the speciation o copper transported by runoff water leaving areas with highly contaminated soils. This is especially the case in south France which presents the largest vineyard area in the world and a Mediterranean rain all regime leading to intense sur ace runoff and erosion processes (Gomer, 1994; Albaladejo et al., 1995). Besides, it must be pointed out that copper seems to be predominantly transported by suspended matter in runoff water (e.g. Gilbin et al., 2000; Xue et al., 2000), which can be related to the affinity of copper or sorption on some components o suspended sediments like organic matter, clay minerals and hydrous metal oxides (Flemming and Trevors, 1989).

In this context, this study intends to:
1. Characterise the amounts and forms of copper transported by suspended matter during storm flow in a catchment that is representative of the soils and the agricultural practices of the Mediterranean wine-growing region; and
2. Identify which soils contribute the most to the particulate copper load of runoff water.

Total copper content is an inadequate variable to use to characterise the behavioural differences of copper in soil, suspended matter or sediment. Copper in soil is known to be distributed among several solid fractions with different physicochemical properties (Darmend-rail, 1987). Five fractions are usually considered:
(1) The soluble exchangeable fraction, of ten low (Emmerichetal., 1982; Shuman, 1986; Sahaetal., 1991; Keller and Vedy, 1994);
(2) The acid-soluble fraction, usually high in carbonated soils (Misra and Tiwari, 1966; Cavallaro and McBride, 1984);
(3) The reducible fraction, associated with iron and manganese oxides and oxyhydroxides in the organic matter content is low (Kuo et al., 1983; Singh et al., 1988);
(4) The oxidizable raction, preponderant in soils with a high organic matter content (Emmerich et al., 1982; Saha et al., 1991); and
(5) The residual fraction, which can be higher than the sum of the other four fractions (Emmerich et al., 1982; Kuoetal., 1983).

The five fractions can be determined analytically using chemical or physical methods (X-ay difractometry, electron microscopy, NMR, etc.). Physical methods give very good results when applied to reference materials, but were shown to be much less precise when applied to raw heterogeneous soils (Flores Velez, 1996). Chemical methods are used more frequently and make it possible to isolate and extract the metals of the different soil constituents by means of extractants. Most extractants, however, are non-specific and also extract elements from the untargeted soil constituents (Nirel and Morel, 1990; Ure, 1996). As a result, the solid phase speciation obtained differs rom speciation in the strict
sense of the term, since it is rather an ‘‘operational’’ speciation dependent on the extraction method used. This type of approach can nevertheless be very useful in
environmental studies (Dawson and Macklin, 1998). It can be used to estimate the risk of trace elements being released when exposed to changes in physicochemical conditions.

In this study, the chemical extractions proposed by Tessieretal. (1979) were applied to suspended matter samples collected during a heavy autumn storm