Toxicité de l ion fluorure envers les organismes d eau douce et effets de la dureté revue et nouvelle analyse de données existantes
Fluoride toxicity towards freshwater organisms and hardness effects review and reanalysis of existing data
Les plus importantes sources anthropiques de fluorures dans les systèmes d'eau douce comprennent les eaux usées municipales, les industries productrices de fertilisants et les alumineries. Plusieurs études montrent que la toxicité des fluorures est réduite lorsque le test toxicologique est réalisé en eau dure plutôt qu'en eau douce. Trois mécanismes peuvent être invoqués pour expliquer une telle tendance : (I) influence des ions de dureté (Ca2+; Mg2+) sur les organismes tests (soit au niveau de la barrière biologique séparant l'organisme de son milieu, soit au niveau de leur métabolisme interne); (II) complexation entre le fluorure et les ions de dureté dans le milieu d'exposition, menant à une réduction de la concentration en fluorure libre (F); (III) précipitation de fluorite (CaF2) dans les milieux d'exposition, menant à une réduction de la concentration effective en fluorures. Pour identifier le ou les mécanisme(s) responsables de l'effet protecteur de la dureté, nous avons réalisé une revue de la littérature existante sur les poissons, les invertébrés et les insectes aquatiques d'eau douce. Parmi ces études, les plus complètes ont été sélectionnées et la spéciation des fluorures modélisée pour chaque cas. Les modélisations réalisées indiquent que la spéciation physique du fluorure (distinction entre les espèces dissoutes et particulaires) a beaucoup plus d'importance que sa spéciation chimique en solution dans les systèmes étudiés.
Important anthropogenic sources of fluoride to the aquatic environment include municipal waste waters and effluents from fertilizer producing plants and aluminum refineries. Many studies have demonstrated that fluoride toxicity to aquatic organisms is reduced when the toxicological test is performed in hard water rather than soft water. In principle at least three mechanisms could explain this trend: (I) a direct beneficial influence of the hardness cations (Ca2+; Mg2+) on the test organism (either externally, at an epithelial membrane, or internally); (II) complexation between fluoride and the hardness cations, leading to a reduction in free the fluoride concentration (F-) in the exposure media; (III) fluorite precipitation (CaF2) in the exposure media, leading to a reduction in the effective fluoride concentration. The present literature review was designed to identify which of these mechanisms might be responsible for the apparent protective effect of hardness on fluoride toxicity. An inventory of the existing literature on the toxicity of fluoride to freshwater fish, invertebrates and aquatic insects was prepared. The most complete studies were selected and the chemical data needed to model cation and anion speciation in the exposure media were extracted from the papers. Speciation at equilibrium was then modelled using as input data the total concentrations of the key constituents (calcium, magnesium, fluoride and chloride), together with the temperature and the pH. The initial speciation calculations revealed a particularity of the chemical systems studied: frequently precipitation of fluorite (CaF2) was predicted by the speciation model (MINEQL+), but the article from which the data had been extracted did not mention the appearance of a precipitate. Fluorite solubility, at a pH of 7 and at an ionic strength of 2.7 mM, is approximately 17 mg CaF2/L (0.22 mM). When high fluoride concentrations are used in h a r d water, both fluoride and calcium concentrations are predicted to decrease markedly as a result of fluorite precipitation. This analysis of the published results thus suggests that the reported lower toxicity of fluoride in hard waters is likely due to the chemical precipitation of CaF2 and MgF2, resulting in a decrease in the effective fluoride concentration to which the test organisms are exposed. In other words, changes in the physical speciation of fluoride (i.e., its distribution among dissolved and particulate species) are much more important than changes in its chemical speciation in solution. Given the low solubility of fluoride in hard waters, it would seem difficult if not impossible to carry out fluoride toxicity tests in hard water. However, in a few fluoride toxicity studies the researchers checked for precipitation by monitoring fluoride and calcium concentrations throughout the toxicity test. In some of these cases, even though speciation calculations predicted fluorite formation at the exposure concentrations used, the authors did not detect any precipitation; these systems were thus apparently in a metastable, over-saturated state, where the kinetics of precipitation were slow relative to the duration of the toxicity test. The chemical equilibrium software was used to simulate fluoride speciation in these systems, by allowing the over-saturated solid phases to remain in solution. In particular, we looked for evidence that under such circumstances the hardness cations exerted a beneficial effect. However, no clear picture emerged from this second analysis: two of the studies designed particularly to test the effect of calcium on fluoride toxicity showed a protective effect, whereas one study indicated the opposite effect, i.e. an increase in fluoride toxicity as the calcium concentration was raised. All fish studies for which calcium concentrations were available (N=20 studies; 58 toxicity tests) were pooled and tested for a possible calcium effect on fluoride toxicity. No relationship was observed between fluoride ion toxicity (LC50, expressed as calculated free [F]) and calculated dissolved calcium concentrations for these pooled data (figure 2). Fluorite solubility was the most important factor influencing the data point distribution in the relationship. The same exercise was performed for all the invertebrate studies (N=11 studies; 22 toxicity tests) but again no relationship was found (figure 3). Several factors other than hardness affect fluoride toxicity to aquatic organisms. Fluoride toxicity to fish increased with exposure duration up to 200 h, where it reached a threshold LC50 level around 100 mg/L (5.3 mM) expressed as free fluoride (figure 4). Fish life stage (figure 5), the temperature of the exposure media (figure 6) and the chloride concentration also affected fluoride toxicity in fish. For invertebrates, fluoride toxicity increased with exposure duration but to a lesser extent than for fish. In summary, water hardness clearly reduces fluoride toxicity to aquatic organisms by limiting the equilibrium solubility of the fluoride ion. However, in those cases where the precipitation of CaF2(s) and MgF2(s) is slow, and where the hardness cations and fluoride co-exist in the dissolved state in the exposure medium, the experimental evidence for a protective effect of hardness on fluoride toxicity is equivocal. To answer the question, new experiments should be performed under conditions that take into account the behaviour of calcium and fluoride in the natural environment. Metastable environments where fluoride concentrations exceed the solubility limit imposed by CaF2 or MgF2 could be reproduced in laboratory toxicity tests by using continuous flow systems. For tests below the solubility limit, toxicity tests with varying levels of Ca or Mg could be designed to stay within the solubility range of CaF2 or MgF2. In both cases, dissolved calcium, magnesium and fluoride concentrations should be monitored throughout the toxicity tests.
Reçu le 18 octobre 2002.
Accepté le 4 novembre 2003.
poisson, invertébrés, dureté, fluorure, spéciation, toxicité.
fish, invertebrates, hardness, fluoride, speciation, toxicity.