Animal Waste, Water Quality and Human Health

Transport of microbial pollution in catchment systems 175were attenuated rapidly which was attributed to early exclusion from macroporeflow but significant groundwater contamination was considered possible fromanimal waste applied to land.Byappanahalli et al. (2003) suggested that significant environmental reservoirsof faecal coliform organisms in stream bed sediments in the mid-west of the USAexisted, indicating protracted survival and potential re-growth outside thealimentary canal within freshwater stream environments. Jamieson et al. (2003)noted the probable existence of a stream bed store of FIOs in a 1000 hawatershed in Ontario and noted a general pattern of of Canadian recreationalwater quality criteria being exceeded in catchment streams draining the livestockfarming and residential area studied.McDonald et al. (1982), Wilkinson et al. (1995, 2006), Wilkinson (1995),Wilkinson (2006) and Muirhead et al. (2004) have used artificial releases of waterto assess the in-channel sedimentary contribution to stream water concentrations.Bai and Lung (2005) adopted a similar approach but then applied a model basedon Environmental Fluid Dynamics Code (EFDC) to model sedimentary and FIOtransport. They suggest the approach facilitates quantification of the channelsedimentary store contribution and the catchment derived inputs at specificpoints, although the published model validation data appear much better for thesediment than for the FIO parameters measured. Collins and Rutherford (2004)developed a catchment simulation model to predict coliform concentration inNew Zealand streams draining livestock grazing areas. The model used dailylivestock data and simulated surface and sub-surface FIO fluxes as well as directdeposition at locations where livestock could access the stream channel. Theynote the uncertainty regarding a number of these processes and the modelsensitivity to the distance between the location of faecal inputs and the catchment.A scenario analysis suggested that riparian buffer areas could be effective boththrough livestock exclusion and microbial attenuation.Land use impacts on catchment-scale FIO budgets have been reported by Fraseret al. (1998) who used a GIS-based sediment delivery model (SEDMOD)calibrated for 12 sub-watersheds of the Hudson River, New York. The sedimentdelivery model together with a GIS layer describing livestock density explained50% of the variance in “average” faecal coliform output. Crowther et al. (2002,2003) reported a multiple regression approach relating percentage land use in aseries of sub-catchments with both high- and low flow geometric mean faecalindicator concentrations observed at sub-catchment outlets during the summerbathing season. The models produced higher explained variance (commonly 60–70%) when predicting high-flow concentration which is the key period ofmaximum delivery (Wilkinson et al. 1995, 2006, Wilkinson 2001). Storm flowwater quality, discharged to coastal waters was examined by Lewis et al. (2005)