Soil water in the riparian zone as a source of carbon for a headwater stream

Abstract

Conflicting views on the potential importance of soil water as a source of dissolved organic carbon (DOC) in headwater streams are addressed by comparing organic chemistries of soil waters in the riparian zone with an adjacent stream at an upland site in mid-Wales during one year. DOC, which as a whole is relatively refractory, was contrasted with the labile dissolved free amino acids (DFAAs). DOC distributions in riparian soil waters showed little temporal or spatial consistency, but concentrations almost always exceeded those in the stream (mean = 9.4 and 2.1 mg l −1, respectively). This represents an average carbon input of some 2500 g C m −2 year −1 to the stream, indicating the quantitative importance of DOC in riparian soil waters to stream productivity. Total DFAA concentrations were <0.3% of the DOC measured. Although their mean concentrations were also higher in soil waters (168.2 nmol l −1) than in the stream (113.9 nmol l −1), this was not a consistent trend. Most probably the DFAAs' lability resulted in their almost complete immobilization within the streambed during microporous discharge of catchment runoff through the streambed. Three major hydrological flow paths in the riparian zone were inferred by correlating dynamics of the organic chemistry, as well as inorganic nitrogen (ammonium and nitrate), between sampling points. A lack of any dynamic relationship between organic concentrations in saturated through-flow of the stream bank and the stream suggested an efficient immobilization of this soil water DOC within the streambed, prior to its discharge into the stream; an immobilization particularly significant at baseflow. Correlations for both DOC and inorganic N between the stream bank unsaturated zone and the stream suggested that, at least at higher flows, this soil water input could directly influence the stream. A high degree of correlation between soil waters of the forest edge (6m from the stream) and the stream suggested that some form of macroporous transfer of soil water from further back in the riparian zone could also directly influence the stream. These results reveal the complexity of the riparian environment. It is concluded that the riparian zone can contribute substantial amounts of DOC to a stream ecosystem, and that the streambed must be a key area of chemical reactivity where much of this material is initially processed. As the delivery of DOC to the streambed will depend on the hydrology of the riparian zone, a knowledge of this is an essential prerequisite to modelling the transfer of organic material from a catchment to a stream.