Aquatic macrophytes as bioindicators of carbon dioxide in groundwater fed rivers.

Aquatic plants have been used as hydrological tracers in groundwater fed river systems. In nature, patterns in plant distribution have been attributed to ammonium (NH(4)) toxicity and phosphate (PO(4)) limitation, while some laboratory studies have focused on the role of the partial pressure of CO(2) (pCO(2)). The aims of this study were (i) to test whether plant distribution was more related to pCO(2) than NH(4) and PO(4) in nature, (ii) to develop and test the predictive power of new plant indices for pCO(2), NH(4) and PO(4), and (iii) to test the potential causality of the relationships using species eco-physiological traits. These tests were carried out with field data from the Rhine, Rhône and Danube river basins. Species composition was best related to the effect of pCO(2). The pCO(2) plant index was well calibrated (r(2)=0.73) and had the best predictive power (r(2)=0.47) of the three indices tested on independent datasets. The plant-pCO(2) relationship was supported by a biological mechanism: the ability of strictly submerged species of aquatic vascular plants to use HCO(3) under low pCO(2). This was not the whole story: the effects of pCO(2), NH(4) and PO(4) on plant distribution were partially confounded and interacted all together with temperature. However, neither NH(4) toxicity nor P limitation could be asserted using species eco-physiological traits. Moreover, the predictive power of the NH(4) and PO(4) plant indices was not as strong as pCO(2), at r(2)=0.24 and r(2)=0.27, respectively. Other potentially confounding variables such as spatial structure, biotic and physical factors were unlikely to confound the findings of this study.