Environmental context. Until now there was no user-friendly code for metal flux computations in natural mixtures of aquatic complexants, which are however essential for prediction of metal bioavailability. The present paper describes the capabilities and limitations of one of the only two such codes presently available, called FLUXY. The results of FLUXY are compared with those of another code, and it is shown that it enables quick computation and is applicable to natural ligands under many environmental conditions.

Abstract. The computation of metal fluxes at consuming interfaces like microorganisms or bioanalogical sensors is of great importance in ecotoxicology. The present paper describes the application of a simple code, FLUXY, for the computation of steady-state metal fluxes in the presence of a very large number of complexes, with broadly varying values of equilibrium constants, rate constants and diffusion coefficients. This code includes two major limiting assumptions, namely, (i) the existence of excess of ligand (L) compared with metal (M), and (ii) the fact that in a series of successive ML_n complexes, the reaction is the rate-limiting step in flux computation. The domains of rate constants for which these assumptions are valid are tested systematically, and the corresponding errors are evaluated by comparison with the exact results given by another code: MHEDYN. FLUXY is then applied and compared with MHEDYN for case studies typical of aquatic systems, namely (i) a culture medium containing simple ligands; (ii) solutions of fulvic compounds including a broad distribution of complex stability and rate constants; and (iii) suspensions of aggregates with a broad size distribution. It is shown that FLUXY gives good results for cases (i) and (iii). Application to case (ii) (fulvic compounds) is also feasible under conditions that are clearly described. Altogether, FLUXY and MHEDYN are complementary. In particular, FLUXY only computes steady-state fluxes and requires the fulfilment of a few conditions, but when these are met, computations require much less computer time than MHEDYN.