b'Causes of phosphorus elevations during low-flows across three contrasting watercoursesVero SE1, Daly K2, McDonald NT1, Leach S1, Sherriff SC3 and Mellander P-E1,21Agricultural Catchments Programme, Teagasc, Johnstown Castle Environment Research Centre, Wexford, Co. Wexford, Ireland2Crops, Environment and Land Use Programme, Teagasc, Johnstown Castle Environment Research Centre, Wexford, Co. Wexford, Ireland3Geography and Environmental Science, School of Social Sciences, University of Dundee,Dundee, UKElevated in-stream phosphorus (P) concentrations have been observed during periods of low river flow. These elevations can exceed Water Framework Directive (WFD) threshold (0.035 mg l-1) during ecologically sensitive periods. This research identified source, mobilization andpathway factors influencing in-stream concentrations across the main watercourse andtributaries of three rivers during each of the spring, summer and autumn low-flow periods. Up to 50 water samples were taken in each catchment per season and analysed for total (TP) and dissolved reactive phosphorus (DRP). Bed sediment was sampled from a subset of seven points in each watercourse and P, Al, Ca and Fe concentrations were assessed (Mehlich extract).In each catchment, the greatest number of sampling points exceeding the WFD thresholdoccurred during summer (57% - well-drained grassland, 11% - well-drained arable, 71% -poorly-drained grassland). Concentrations declined during autumn but typically, did not return to spring baseline. Different controls on P were identified in each catchment. The well-drained grassland exhibited diffuse transport through the groundwater pathway, coupled by lack of dilution. This manifested as similar concentration changes across most sampling points. The poorly-drained grassland showed a combination of identifiable point sources (agricultural and domestic) and cumulative loading across the watercourse. Mobilisation from in-stream stores occurred in the arablecatchment, which typically exhibited greater Al:P ratios in bed sediment than the otherwatercourses. These sediments are more likely to attenuate P delivered from external sources (agricultural and wastewater treatment facility), which may potentially be released subject to in-stream conditions. This was the only catchment which returned to base concentrations at most sampling points by autumn. Assessment of stream P must take into account the flow of the watercourse, as measurements taken at low-flow may be elevated above typical concentrations. Longitudinal water sampling allows greater spatial accuracy in identifying potential sources and pathways.9'