Nutrient dynamics in river bed sediments: effects of hydrological disturbances using experimental flow manipulations

River sediments play a crucial role in the storage and transformation of organic matter (OM). Nutrient dynamics are controlled by the interaction of several key parameters, i.e. river discharge, channel geometry and vertical exchanges of water (upwelling vs. downwelling zones). The main aim of this study was to evaluate the effect of channel forms and discharge variation on nutrient spiralling in the hyporheic zone (HZ) of streams. Four experimental flow manipulations (EFM) were carried out at two reaches with different channel forms (straight vs. sinuous) in an oligotrophic subtropical river in Australia. Flow manipulation consisted of reducing the river width with a temporary dam, diverting and concentrating the main water flux on two different geomorphological units (riffle vs. gravel bar), in order to simulate flooding conditions. Hyporheic waters were analysed for their physicochemical characteristics and nutrient (nitrates + nitrites = NO (x) and soluble reactive phosphorus [SRP]) and OM contents at two depths (10 and 50 cm) within the bed sediments, both upstream and downstream of the geomorphological units. The physicochemical parameters clearly demonstrated the existence of hyporheic flow paths, characterized by the alternation of downwelling and upwelling areas, with more consistent gradients in gravel bars than in riffles. The HZ acted as source for NO (x) and SRP, but this role varied between geomorphological units and reaches. The effect of EFM differed between sampling points, irrespective of the type of geomorphological unit. In gravel bars, a flush out during high discharge was observed for NO (x) , SRP and particulate organic matter (POM) at the sinuous channel, whereas storage and removal were recorded at the straight channel for SRP and NO (x) , respectively. At the riffle of the sinuous channel, very fine POM accumulated, while removal was noticed for POM. In contrast, at the riffle of the straight channel, SRP accumulated in the HZ and NO (x) was removed out of the HZ. Nutrient dynamics in the HZ and the response to flow increases were not governed by the geomorphological unit type. Other parameters that determine water residence time in the sediments, such as local heterogeneity in sediment characteristics (grain size, porosity and hydraulic conductivity), channel sinuosity, reach slope and the size and form of the gravel bar, may be more significant explanatory variables for understanding OM and nutrient dynamics in the HZ. This study emphasizes the need for caution in making generalisations about the role of river sediment in nutrient storage and the impact of floods on nutrient dynamics.