Nouvelles connaissances des formes du phosphate : conséquences sur le cycle du phosphate dans les sédiments des eaux douces peu profondes

New knowledge of phosphate fractions : consequences for the phosphate cycle in shallow freshwater sediments

¹ Station Biologique de la Tour du Valat, le Sambuc, 13200 Arles, France
² « Avec tristesse je dois annoncer la mort soudaine de M. Kees de Groot, survenue le 21 septembre 1994 à Barcelone. Que ce travail soit un monument pour ce jeune chercheur prometteur ; il nous manquera fortement. »

Résumé

Abréviations utilisées : o-P : ortho-phosphate dissous Fe(OOH) ≈ P : phosphate adsorbe au Fe(OOH) CaCO₃ ≈ P : phosphate relié au CaCO₃, sous forme d'un mélange de calcaire et de micro-cristaux d'apatite POAS : phosphate organique particulaire, soluble en milieu acide POR : phosphate organique particulaire résiduel P_tot : phosphate total d'un sédiment (mg.g^-1) La composition chimique des phosphates liés aux sédiments d'eau douce a été analysée et l'adsorption de phosphate sur ces sédiments a été étudiée. Le rôle du Fe(OOH) dans cette adsorption a été confirmé et peut être décrit par l'isotherme d'adsorption de Freundlich : P_ads = A.(o-P)^B où P_ads = concentration du phosphate adsorbé (par les sédiments ou Fe(OOH)) o-P = concentration du phosphate dissous A et B = constantes Nous avons calculé les constantes A et B par ajustement des courbes par la méthode des moindres carrés ; les deux sont des fonctions du pH. On perd très peu de précision si B est ajusté à 0.33 et « A » à A = 23626* 10^{(-0.42 pH)}. L'adsorption peut donc être quantifiée en connaissant la concentration du Fe(OOH) dans les sédiments et le pH. Puisque le rôle de Fe(OOH) est important, la conversion du Fe(OOH) en FeS par exemple, peut avoir une forte influence sur l'adsorption de l'o-P sur les sédiments. Dans les eaux « dures », la solubilité de l'o-P est en plus limitée par la solubilité de l'apatite, Ca₅(PO₄)₃. OH, Nous avons trouvé que le produit de solubilité est de 10_-50. En combinaison avec l'adsorption quantifiée sur le Fe(OOH), on peut établir un diagramme de solubilité d'o-P dans l'eau en fonction de la concentration de Fe(OOH) dans les sédiments, de la concentration de Ca²⁺ dans l'eau et du pH. Il est apparu qu'il existait une forte corrélation entre la somme (Fe(OOH) ≈ P + CaCO₃ ≈ P) et le P disponible pour certaines espèces d'algues dans 11 lacs néerlandais. En outre, on trouve dans les sédiments deux « pools » de phosphate organique, l'un soluble dans les acides (HC1 ou H₂SO₄) et l'autre dans la soude (NaOH). Nous pensons, que pour le premier, il s'agit d'un complexe avec la matière humique, tandis qu'il a été démontré que le deuxième est formé en majorité d'un complexe entre l'inositol hexa phosphate (phytate) et probablement, le Fe(OOH). Ce complexe peut expliquer pourquoi ce phosphate organique est protégé vis-à-vis des bactéries. En utilisant le diagramme de solubilité on peut maintenant calculer ce qui se produit si un lac peu profond est chargé en phosphate et pourquoi, si la charge est arrêtée, la situation ne peut s'améliorer que très lentement.

Abstract

The purpose of this paper is to analyse the different phosphate compounds in sediments of freshwater systems and to quantify the processes by which they are formed and which may lead to the equilibrium between uptake and release. For the analysis a sequential extraction scheme is therefore developed, using chelators (such as NTA and EDTA) to extract first the 2 inorganic phosphates, i.e. the iron and the calcium bound phosphate. Fe(OOH) ≈ P is extracted with a Ca-NTA + dithionite solution, (resp. 0.02 M 1^-1 and 0.05 M 1^-1) and CaCO₃ ≈ P with Na-EDTA (0.05 M 1^-1). Both extractants are adjusted to the pH of the sediment to be analysed. Following these extractions, two pools of org-P can be identified, using subsequent extractions with H₂SO₄ (0.5 M 1^-1 during 30 min) followed finally by one with NaOH (2 M 1^-1, at 90° C during 5 min). The advantage of this scheme lies in the fact that the inorganic phosphates are extracted before the organic phosphates, together with the Fe(OOH) and CaCO₃, so that no interactions between these adsorbents and phosphate can disturb the analysis of the organic phosphates after hydrolysis. The adsorption of ortho-phosphate onto freshwater sediments was studied. The influence of Fe(OOH) on this adsorption process was confirmed in the laboratory. It was found that this adsorption could be described satisfactory by the Freundlich adsorption isotherm : P_ads where = A.(o-P)^B P_ads = o-P adsorbed onto the sediments or onto Fe(OOH) and A and B are constants We have quantified « A » and « B » by least squares fitting ; they are both functions of pH, although for B only moderately. If, however, B is set at 0.33 little precision is lost, and A can be approximated by : A = 23626* 10^{(-0.42 PH)} . In waters with a low Ca²⁺ concentration, the adsorption onto the sediments can therefore now be quantified as a function of the annual loading, if the Fe(OOH) concentration in the sediments and the pH of the water are known. As Fe(OOH) plays such an important role in the adsorption process, the conversion, e.g. into FeS as usually occurs in anoxic sediments of shallow water bodies, will influence the adsorption strongly. The constant « A » does not only depend on the pH, but on the Ca²⁺ concentration as well. As yet no formula is available to quantify this influence. In hard water, the solubility of o-P will also be limited by the solubility product of apatite, Ca₅(PO₄)₃OH. Using published data from the two hard water rivers Rhine and Rhone, we have found an 'apparent' solubility product of 10^-50 , not taking into account the influence of the activity coefficient due to ionic strength. With this solubility product the maximal o-P concentration can be calculated as a function of the Ca₂₊ concentration in the water and the pH. Together with the equilibrium constants of the Fe(OOH) ≈ P adsorption complex, we have also calculated the maximal o-P concentration as it depends on the Fe(OOH) concentration in the sediments and the pH of the system. Combining these two functions, we have calculated a solubility diagram of the maximal o-P concentration in water in equilibrium with the Ca²⁺ concentration in the water, the Fe(OOH) concentration in the sediments and the pH. The solubility of dissolved ortho-phosphate appears to depend only on the Fe(OOH) concentration in the sediments, the Ca²⁺ concentration in the water (not the CaCO₃ concentration in the sediments) and the pH. In 11 dutch lakes it appeared that the sum of (Fe(OOH) ≈ P + CaCO₃ ≈ P) is available for phytoplankton growth. Furthermore we have demonstrated the existence of two organic phosphate pools in the sediments, the first acid soluble (ASOP) and the second alkaline soluble (ROP). We think that the first pool is a humic acid-phosphate complex although we can only present circumstantial experimental evidence. During the extraction, the phosphate of this pool is hydrolysed and goes into solution as o-P, which makes further identification difficult. The concentration of this pool in the sediments varied between 60 and 225 µg.g^-1. During desiccation of the sediments of a marsh in the Camargue (Les Garcines), this pool disappeared and became inorganic phosphate. By using an enzymatic hydrolysis with phytase we could demonstrate that most of the second pool (ROP) was phytate, i.e. inositol hexaphosphate. Its concentration varied between 100 and 300 µg.g^-1. Experimental evidence suggests that this phosphate is probably complexed with Fe(OOH). The existence of this complex can explain why this organic phosphate is not available for bacterial mineralisation, is therefore not biodegradable and can consequently accumulate in sediments. In shallow lakes and marshes receiving an important P loading, we can now quantify with the solubility diagram what will happen with the phosphate concentrations in water and sediments. It appears that during a first phase, nearly all phosphate will enter the sediments, while the concentration in the water increases only very slowly. Furthermore we can predict what will happen later'when the P-loading is stopped : depending on the water retention time, the situation will only improve very slowly and it may take even longer than the period of loading to return to the low, initial concentrations. In marshes without water renewal, the concentration will remain constant. The available phosphate may, however, still decline because of an accumulation of non-bioavailable phosphate, e.g. phytate.