Exoenzyme activities as indicators of dissolved organic matter composition in the hyporheic zone of a floodplain river

1. We measured the hyporheic microbial exoenzyme activities in a floodplain river to determine whether dissolved organic matter (DOM) bioavailability varied with overlying riparian vegetation patch structure or position along flowpaths. 2. Particulate organic matter (POM), dissolved organic carbon (DOC), dissolved oxygen (DO), electrical conductivity and temperature were sampled from wells in a riparian terrace on the Queets River, Washington, U.S.A. on 25 March, 15 May, 20 July and 09 October 1999. Dissolved nitrate, ammonium and soluble reactive phosphorus were also collected on 20 July and 09 October 1999. Wells were characterised by their associated overlying vegetation: bare cobble/young alder, mid‐aged alder (8–20 years) and old alder/old‐growth conifer (25 to >100 years). POM was analysed for the ash‐free dry mass and the activities of eight exoenzymes (α‐glucosidase, β‐glucosidase, β ‐N‐acetylglucosaminidase, xylosidase, phosphatase, leucine aminopeptidase, esterase and endopeptidase) using fluorogenic substrates. 3. Exoenzyme activities in the Queets River hyporheic zone indicated the presence of an active microbial community metabolising a diverse array of organic molecules. Individual exoenzyme activity (mean ± standard error) ranged from 0.507 ± 0.1547 to 22.8 ± 5.69 μmol MUF (g AFDM)^?1 h^?1, was highly variable among wells and varied seasonally, with the lowest rates occurring in March. Exoenzyme activities were weakly correlated with DO, DOC and inorganic nutrient concentrations. 4. Ratios of leucine aminopeptidase : β‐glucosidase were low in March, May and October and high in July, potentially indicating a switch from polysaccharides to proteins as the dominant component of microbial metabolism. 5. Principal components analysis indicated that there were patch effects and that these effects were strongest in the summer. 6. DOM degradation patterns did not change systematically along hyporheic flowpaths but varied with overlying forest patch type in the Queets River hyporheic zone, suggesting that additional carbon inputs exist. We hypothesise that the most likely input is the downward movement of DOM from overlying riparian soils. Understanding this movement of DOM from soils to subsurface water is essential for understanding both the hyporheic metabolism and the carbon budget of streams and rivers.