THE LABYRINTH OF NUTRIENT CYCLES AND BUFFERS IN WETLANDS - RESULTS BASED ON RESEARCH IN THE CAMARGUE (SOUTHERN FRANCE)

Wetlands, especially in the Mediterranean area, are subject to severe eutrophication. This may upset the equilibrium between phytoplankton p roduction in undesirable quantities and a quantitatively desirable mac rophyte production. In order to manage this equilibrium, a quantitativ e knowledge of nutrient input and fluxes is essential and the role of sediments in these processes must be understood. This knowledge can be useful even for agriculture, e.g. rice cultivation, where optimal uti lization of fertilizers can lead to an economic benefit.In this articl e different aspects of nutrient cycles are discussed in view of approa ching a sufficiently precise quantification. The nutrient input balanc e of the Camargue was therefore measured which showed that the input o f nutrients with the irrigation water, taken from the river Rhone, rou ghly equals the quantity of fertilizers added. Phytoplankton growth ca n be approached reasonably with the Monod model, although there are st ill many practical problems, such as the influence of the pH on P upta ke and the problem of measuring P uptake in the field. The situation i s worse for macrophyte growth; quantitative data are scarce and studie s have often been carried out with unrealistic nutrient concentrations or without addressing the influence of the sediment. This influence c an also include negative factors, such as high concentrations of Fe2+, H2S or FeS, but cannot yet be quantified. The nitrogen cycle in wetla nds is dominated by denitrification. Most wetlands have sediments with high concentrations of organic matter, therefore with a large reducin g capacity. Besides this process, we have shown that denitrification c an also be controlled by FeS. In the Camargue sediments this denitrifi cation is mediated by bacteria from the sulfur cycle; this appeared to be the major pathway. It was shown that a stoicheiometric relation ex ists between nitrate reduced and sulphate produced. The influence of t he temperature was quantified and appeared to be stronger at high orga nic matter concentrations than at lower ones. Denitrification with FeS means that the bacteria use nitrate also for their N demands, while t his is not necessarily the case during denitrification with organic ma tter. Mineralization of macrophytes is a much slower process than that of phytoplankton, probably because of their high C/N ratio. We could, however, not confirm the general assumption that the addition of nitr ogen stimulates this mineralization. On the contrary, we found that tw o amino acids both with a C/N ratio of 6 had different mineralization rates. The amino acid composition of dead macrophytes and the C/N rati o may be of equal importance. Unlike nitrogen, phosphate is always str ongly adsorbed onto sediments. The two mechanisms of the adsorption of inorganic phosphate onto sediments, i.e. the adsorption onto Fe(OOH) and the precipitation of apatite, have been quantified. The adsorption of phosphate onto Fe(OOH) can be satisfactorily described with the Fr eundlich adsorption isotherm: P-ads = A(o-P)(B). The adsorption coeff icient A depends on the pH of the system and the Ca2+ concentration of the overlying water and can be quantified preliminarily by A = a.10(( -0.416pH)).(2.86 - (1.86.e(-Ca2+))). B can be approached by 0.333, wh ich means the cube root of the phosphate concentration. The second mec hanism is the solubility of apatite. We found a solubility product of 10(-50) for hard waters. The two mechanisms are combined in solubility diagrams which describe equilibrium situations for specific lakes. Th e conversion of Fe(OOH) to FeS has a strong influence on phosphate ads orption, although the partial reduction of Fe(OOH) approximate to P by H2S does not release significant quantities of phosphate. Even after complete conversion to FeS only a small part of the bound phosphate wa s released. Besides the two inorganic phosphate compounds, we establis hed the existence of two organic pools, one soluble after extraction w ith strong acid (ASOP), the other one with strong alkali. The first po ol is probably humic bound phosphate, while the larger part of the sec ond pool was phytate. The ASOP was remineralized during the desiccatio n of a Camargue marsh; this drying up oxidized FeS, thus improving the phosphate adsorption and decreasing the denitrification capacity. It can, therefore, be an important tool for management. The phytate was s trongly adsorbed onto Fe(OOH), which explains the non-bioavailability towards bacteria. The fact that the sediment phosphate concentration c an be approached by multiplying the relevant sediment adsorption const ant with 3 root o-P concentration has the consequence that much larger quantities of phosphate accumulate in the sediments than in the overl ying water. This means that even if the phosphate input is stopped, th e eutrophication will only be reversed very slowly, and not at all, if the shallow waters in wetlands have no through flow - as is often the case in many marshes in Mediterranean wetlands.