7. Conclusions
Analysis of the EW E. coli dataset has confirmed that faecal contamination of streams and rivers occurs throughout the region. Median concentrations range from 1 to 1,300 cfu/100 mL and, at 53 of the 73 sites sampled, they exceed the guideline of 126 cfu/100 mL for contact recreation. Bivariate relationships derived at individual sites were characterised by considerable scatter, but showed moderately high correlation between flow, E. coli, and turbidity. Some of the variance in E. coli concentration at a given site can be attributed to variations in flow. Turbidity may be a useful surrogate variable for E. coli at a given site although sampling a range of flows would be required to establish the correlation between E. coli and turbidity. Such correlations may still have a moderately high unexplained variance and generally cannot be transferred to other sites.
Examination of the microbial dataset in conjunction with a range of environmental factors, such as land use and soil type, has highlighted the key controls upon faecal contamination. The pattern of contamination across the Waikato is strongly influenced by the presence of grazing livestock and the highest median E. coli concentrations are generally associated with the most intensive dairy farming in the centre of the region. Conversely, the lowest median values are found in forested catchments, although concentrations are always detectable, indicating contamination by wild animals.
A relatively strong relationship exists between the percentage of a catchment characterised by poorly drained soil and median E. coli across the region. This is probably attributable to the generation of a relatively large volume of surface runoff on these soils that is able to entrain faecal material and quickly transport it to the stream network. It is also probable that the installation of sub-surface drains and drainage ditches in poorly drained soils accelerates the transport of faecal microbes to streams. The bacterial water quality of streams draining such soils is likely to be particularly sensitive to livestock grazing and the application of effluent to land. Appropriate mitigation measures on land underlain by poorly drained soils may include the adoption of less intensive farming practices, optimising the timing of effluent application to land to avoid periods when the soil is saturated, wetland treatment of wastes, surface runoff and drainage flows, and retirement of riparian areas.
Median turbidity is a relatively strong predictor of median E. coli across the region. This indicates that the processes mobilising fine sediment, both on the hillside and in-stream, apply also to bacteria.
A statistical model, developed to predict median E. coli concentrations at the regional scale, explains almost 70% of the observed variance. The percentage of poorly drained soil, the density of cattle within a catchment, and median turbidity at a catchment outlet, are the 3 key factors incorporated into the model.
With the exception of a few sites, the discharge of point sources direct to waterways appears not to influence median E. coli concentrations in the Waikato region. This is attributed to a relatively low number of consented discharges that can cause faecal contamination, and to recent improvements in the treatment of waste water.
Strategies to reduce faecal contamination of streams and rivers in the Waikato region should focus upon grazing livestock. Cattle access to streams and near-channel areas is likely to be important in determining the level of faecal contamination and mitigation measures may therefore be best directed at riparian zones. Permanent fencing to exclude livestock from stream channels and a proportion of riparian land is likely to be the most effective measure in reducing faecal contamination by grazing cattle. There are also a number of riparian management alternatives to permanent fencing that may not be as effective, but may still result in reduced faecal contamination.
A weak inverse relationship was derived between the presence of a wetland within a catchment and median E. coli. A tentative inference is drawn therefore, that those wetlands within the region that are large enough to be discriminated on the land cover map, act to attenuate bacteria. Protection of these wetlands is therefore likely to be beneficial in reducing faecal contamination. A distinction is made, however, between the larger lowland wetlands found, for example, on the Hauraki plains, and small hill-country wetlands. Studies elsewhere have shown that cattle are attracted to the latter, and that they can sometimes be a source rather than sink for faecal contamination to waterways.
Contact for Enquiries
MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND
Fax: +64 4 894 0721
Contact this person
