6. Management Implications
Results from the EW in-stream monitoring programme indicate that E. coli concentrations generally increased with flow by roughly an order of magnitude over the flow range encountered. Results elsewhere from dedicated storm monitoring indicate, however, that a 2 or 3 order of magnitude increase in microbial concentration may occur over flood events (Muirhead 2001, Wilkinson et al. 1995). These findings suggest that low frequency sampling regimes, since they are unlikely to capture large storm events, preclude an accurate estimation of microbial flux. Where estimates of microbial loads are required (for example, in catchments supporting estuarine shellfish farming), high frequency sampling during storm events is necessary (Davies-Colley et al. 2001).
Bivariate relationships derived from the EW water quality monitoring program exhibit a correlation, at individual sites, between E. coli and turbidity. Within a site, therefore, turbidity data may be of use as a surrogate variable for E. coli offering a cheaper alternative to the direct monitoring of faecal contamination, especially if high frequency sampling is desirable. Across a region, however, turbidity may only provide a broad indication of median faecal contamination since, for example, highly erodible soils (that contribute to high turbidity) may not be subject to livestock grazing. Conversely, non-erosive soils may be subject to intensive grazing and point sources of faecal contamination.
The pattern of microbial contamination across the Waikato region is strongly influenced by the presence of grazing livestock. This finding supports the assertion of Vant (2001) that non-point agricultural sources now provide the dominant contribution to faecal contamination in the Waikato River. Strategies to reduce faecal contamination of streams and rivers must, therefore, address this primary, diffuse source. It is likely, although unproven within this study owing to data limitations, that the degree of cattle access to streams is important in determining the level of faecal contamination of waterways. This is because cattle deposit faecal material directly to streams, and onto stream banks where it is readily washed into the channel by overland flow or entrained by rising streamwater. Wash-in by overland flow may be accentuated by cattle treading which reduces the trapping efficiency of riparian soils (Nguyen et al. 1998). Permanent fencing to exclude livestock from stream channels and a proportion of riparian land is likely to be an effective measure to reduce faecal contamination by grazing cattle. Quantifying the effectiveness of this management intervention is difficult and, at present, the relative importance to faecal contamination of direct and near-channel deposition compared with overland flow from elsewhere in the catchment is not clear (Davies-Colley and Parkyn 2001). There are a number of riparian management alternatives to permanent fencing that may not be as effective, but could help reduce faecal contamination. These are summarised from Davies-Colley and Parkyn (2001), and illustrated in Table 6.
Table 6. Options for livestock management in riparian zones to reduce faecal contamination. Summarised from Davies-Colley and Parkyn (2001).
|
Management approach |
Benefits |
Notes |
|
Permanent fencing, and therefore, growth of riparian vegetation |
Removal of direct and near-channel deposition of faecal material. Increased trapping efficiency of microbes washed downslope in overland flow |
Fencing costs, planting costs, weed and pest management required. Planting needed for best outcome ? |
|
Temporary fencing |
Temporary benefits as above. Can also be used to selectively control animal access, for example prevent access when soils are wet |
Considerable management required for weed control and maintenance of grass sward |
|
Rest-rotation grazing |
Permits soil and grass recovery between grazing episodes aiding trapping efficiency of microbes in overland flow |
Requires considerable fencing and stock management. |
|
Off-stream watering |
Removes one incentive for livestock to access streams. Reduced direct and near-channel deposition of faecal material |
Water may not be the only or main reason for stock access to streams |
|
Off-stream shade & shelter |
Removes one incentive for livestock to access streams. Reduced direct and near-channel deposition of faecal material |
Shade and shelter may not be the only or main reason for stock access to streams |
|
Livestock bridges on farm races |
Removes livestock access where raceway intercepts stream channel |
Costly? Main application on dairy farms ? |
Despite its likely importance, stock access to streams is difficult to quantify, particularly at a regional scale. Although riparian planting can be identified from land cover maps or aerial photos, its presence does not necessarily indicate stock exclusion. Field survey within this study indicated that cattle access to streams is highly variable within a catchment. Methods are required for characterising riparian zones across a region, with respect to livestock access and the entrapment of faecal material entrained within overland flow.
Soil drainage properties explain much of the variation in streamwater faecal contamination and the percentage of poorly drained soil within a watershed is a key factor within the statistical model that predicts median E. coli across the region. Two possible mechanisms giving rise to this correlation are: enhanced surface flow, and artificial drainage. It is not possible to determine which mechanism predominates in a statistical analysis of this kind and further experimental studies are desirable. Nevertheless, the strength of poorly drained land as an explanatory variable has important implications for land management, suggesting that bacterial water quality on poorly drained land would benefit from (1) better riparian protection to maximise filtering of faecal material within overland flow and, (2) better management of subsurface drainage systems, for example, through wetland treatment of drainage flows.
The importance of soil drainage properties is further illustrated through reference to the relationship between median E. coli and the discharge of dairy effluent to land (Figure 12). Here, the sites with the greatest discharge of dairy effluent (sites 8, 31 and 54) do not have particularly high median E.coli concentrations. Since these three sites are all characterised by well-drained soils (>90%) an inference may be drawn that the good soil drainage properties act to attenuate the transport of faecal material from land to the channel network. In contrast to poorly drained land where riparian planting may be particularly effective, on well-drained soils, faecal material is more likely to be attenuated by infiltration into the soil matrix.
Given the weak inverse relationship between the presence of wetlands within pastoral (> 40%) catchments, and median E. coli (section 4.2.4), it can be tentatively concluded that wetlands generally act to attenuate faecal contamination. Destruction of existing wetlands is, therefore, likely to have a detrimental impact on bacterial water quality. Such an inference may only be applicable to the wetlands large enough (> 1 ha) to be included in the land cover data used in this study, as they are unlikely to be substantially grazed by livestock. Studies of small (<10 m) hill-country wetlands (Collins 2002) indicate, however, that cattle are attracted to them to graze, and that they can be a source of faecal contamination. Management of these wetlands to exclude livestock from them may therefore be necessary to reduce faecal contamination of hill-country streams.
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
