Pathogen Pathways – Best management practices
4. Mitigation
4.1 Direct pathways
The study of Davies-Colley et al. (2004) shows that herd crossings are a key faecal microbe transmission route. Clearly, therefore, stream crossings of dairy herds should be avoided by use of bridges or culverts. Unpublished work in the Sherry River, site of the crossing study by Davies-Colley et al. (2004) shows that bridging of crossings along that waterway has appreciably improved water quality. However, guidelines for contact recreation are still often exceeded there, presumably because of pollution via indirect pathways and direct deposition where dairy cattle have access to unfenced tributary streams and drains (Dr Rob Davies-Colley, NIWA-Hamilton, pers. Comm.).
Permanent fencing is the most stringent and ‘absolute’ measure with which to prevent sporadic incursions by cattle into waterways. This has three effects: (1) The source of faecal deposition is removed from water and channels, and from riparian areas proximal to the stream from which surface runoff can deliver pathogens, (2) a riparian buffer can be created (provided the fencing is set back from the bank) that can entrap microbes washed in from upslope, preventing transport to the water (Section 4.2.6), and (3) the source of soil and vegetation damage (devegetation, soil compaction, creation of cattle ‘ramps’) is removed, so that riparian functioning is restored over time, particularly a high infiltration capacity that reduces wash-in (to waterways) of faecal microbes by surface runoff. Askey-Doran (1999) gives a number of suggestions for permanent and electric fencing near and across streams, including methods to avoid flood damage to fencing infrastructure. Some guidelines on fencing are also provided by Collier et al. (1995).
Alternative options to fencing exist, that can potentially reduce the input of faecal material directly to waterways. These involve the encouragement of livestock away from waterways through providing alternative resources such as water, shade and shelter. The provision of an alternative water source, in particular, has been shown experimentally to have significant potential in the United States (e.g., Miner et al. 1992; Sheffield et al. 1997). However, the impact of providing alternative resources remains untested in New Zealand.
4.2 Indirect pathways
4.2.1 Soil properties, vadose zone and groundwater characteristics
Soil properties
The identification of appropriate (and, conversely, inappropriate) soils for grazing and the irrigation of effluent and water, is a key step towards mitigating microbial pollution.
The ability of a given soil to attenuate microbes is strongly dependent upon the degree to which water from rainfall or irrigation can infiltrate into the soil, rather than contributing to the generation of surface runoff, a process that can rapidly deliver microbes to waterways. Poorly drained soils (low infiltration rate) have been shown to be a relatively strong predictor of streamwater faecal contamination (Collins 2003).
Soil microbial attenuation is also strongly dependent upon the degree to which infiltrating water movement passes through the fine pores of the soil matrix, coming in contact with reactive internal surfaces that aid attenuation. The more water movement that bypasses these fine pores, flowing instead through macropores, the less opportunity there is for microbial attenuation (Section 3.2).
Whilst water infiltrates slowly through the matrix in poorly drained soils (encouraging the generation of surface runoff) these soils can often also be characterised by high levels of bypass flow, whereby water and pollutants flow rapidly to depth. Consequently, microbial contamination of both surface and subsurface flows is commonly associated with poorly drained soils. In contrast, soils with high infiltration rates are often well drained with less potential to generate surface runoff. In addition, macropores are typically absent and hence infiltrating water passes through the soil matrix with an associated efficient filtration and adsorption of microbes.
A series of laboratory experiments (Aislabie et al. 2001; McLeod et al. 2001, 2003, 2004) has been conducted with which to determine the potential for microbial bypass flow on a number of New Zealand soil types found on flat to rolling land (< 15° slope). The experiments involved the irrigation of undisturbed soil cores with 25 mm of dairy shed effluent, followed by the continuous addition of artificial rainfall. Leachate discharging at the lower end of each core was analysed to determine concentrations of a host-specific bacteriophage (virus) and faecal coliforms, over time. The resulting microbial breakthrough curve for each soil, i.e., the plot of microbial concentration against drainage volume, provides a measure of the potential for microbial bypass flow. Each soil is classified by one of three categories of bypass potential: high, medium and low. Typically, the soil properties that lead to high bypass flow also contribute to a high risk of surface runoff. Conversely, low bypass flow soils are also often characterised by a low risk of surface runoff.
High bypass flow soils
High bypass flow soils are characterised by the rapid vertical movement of water and microbes, with minimal microbial attenuation. Many of these soils are clayey or can rapidly become waterlogged. Topsoils may pug easily as a result of stock trampling under wet conditions. As a consequence, they may generate significant surface runoff under heavy or prolonged rainfall. Larger surface runoff volumes will be generated upon sloping land and undrained sites. The avoidance of grazing when soils are wet enough to become pugged by stock, and rate and volume controlled irrigation upon these soils, are appropriate management practices, likely to lead to improvements in bacterial water quality.
In the field, soils with a high potential for microbial bypass flow can be recognized by the following features:
- well developed structure where cracks or large worm holes are easy to see;
- grey colours dominant immediately below the topsoil;
- rust coloured mottles immediately below the topsoil;
- the presence of artificial drainage;
- slow drainage through the soil matrix;
- waterlogged conditions.
In the North Island, approximately 50% of the 40,500 km2 of flat to rolling land is underlain by soils with a high potential for microbial bypass flow. Large areas occur, for example, in Northland on old strongly weathered, strongly structured, clayey soils, and in the Manawatu, Wairarapa and Hawke’s Bay regions where soils have a drainage or structural impediment. Low-lying soils on the Hauraki Plains and in the Waikato are also ranked as having a high potential for microbial bypass flow, as is the large area of Organic Soils found in the Waikato.
In the South Island, approximately 45% of the 51,800 km2 of flat to rolling land is underlain by soils with a high potential for microbial bypass flow. Large areas occur, for example, on the west coast where Podzols and Gley Soils have developed under high rainfall. A drainage impediment indicates that many of the soils in the south and east of the South island are also classified as having a high potential for microbial bypass flow.
Medium bypass flow soils
With careful grazing and irrigation management, medium bypass flow soils are capable of retaining microbes within their horizons. Brown and Pallic Soils are classified with medium potential and are widely distributed throughout both islands with Pallic Soils predominantly found in the drier eastern regions.
Low bypass flow soils
In low bypass flow soils, the microbial contamination of leachate is at low concentrations, or effluent irrigation can be readily managed so as to minimise the microbial load in the leachate. These soils are the most appropriate for grazing and irrigation, and are able to produce leachate with minimal microbial contamination when irrigated correctly. The weakly developed soil structure, high porosity and fine peds associated with loamy Allophanic Soils, and the apedal or single grain structure of Pumice Soils encourages the movement of water through their matrices. As a consequence, the microbial bypass flow potential of these soils is low and their filtering of microbes is high. Typically, these soil do not require the implementation of mitigation strategies.
In the field, soils with a low potential for microbial bypass flow can be recognized by the soil having any of the following features:
- well drained soils developed in loamy or silty volcanic tephra;
- well drained soils developed in pumiceous soil material possibly excluding those developed on coarse rubbly pumice;
- young alluvial soils developed in loamy soil material with weakly developed soil structure;
- soils with low bulk density that have many fine pores.
Approximately 40% of North Island soils on flat and rolling land have a low potential for microbial bypass flow. These are predominantly volcanic soils (Allophanic and Pumice Soils) around Taranaki, the central North Island, and the Waikato. In contrast to the North Island, only 25% of South Island soils (on flat or rolling land) have low microbial bypass flow potential. These are associated predominantly with Allophanic Brown Soils, developed from wind blown material or loess.
The vadose zone
The vadose zone occurs from immediately beneath the soil, down to the top of the groundwater table. Spatially, it is of variable depth and not permanently saturated. The thickness and nature of the vadose zone will affect microbial transport to the regional groundwater table. For example, naturally porous vadose zone material such as pumice will physically trap a higher proportion of microbes than fractured rock. Sandy or gravelly sand alluvium may also be an effective trap for microbes. Grazing and irrigation restrictions are appropriate mitigation measures on land overlying a shallow vadose zone with poor microbial attenuation properties.
Groundwater
Grazing and irrigation restrictions are likely to be appropriate mitigation measures on land overlying large-pore heterogeneous aquifers that are vulnerable to microbial contamination e.g. aquifers of alluvial gravels, fractured volcanic rocks and karst lithology (Davies-Colley et al. 2003).
4.2.2 Irrigation management
Effluent irrigation
In addition to the identification of appropriate soil properties (section 4.2.1) timing, volume, location and technique are also key factors in the optimal irrigation of effluent, with respect to minimising pollutant loss to waterways. Robb and Barkle (2000) provide guidelines for the application of effluent to land.
Irrigation when soils are at or near saturation, can promote the generation of surface runoff, and bypass flow down through the soil horizons to either groundwater or subsurface drains; both processes are rapid transmission pathways for faecal microbes. Ideally, irrigation should occur when the volume to be applied does not exceed the water storage capacity of the soil, with effluent being stored until such soil moisture conditions arise. This ‘deferred irrigation’ has led to marked decreases in nutrient loss to waterways (Houlbrooke et al. 2004a) and is likely to be similarly successful with respect to faecal microbes. Experimental and modelling studies by Monagahan and Smith (2004) also support the merits of deferred irrigation. Measurement or prediction of soil moisture on a daily basis is a central requirement of the deferred irrigation approach. In addition, sufficient effluent storage capacity is a key requirement, particularly during winter and spring when soil moisture deficits are small or non-existent (Houlbrooke et al. 2004a). At a West Otago study site, Monaghan and Smith (2004) estimated that between 44 and 109 days of effluent storage would be required per year depending upon rainfall and the groundspeed of the irrigator. Houlbrooke et al. (2004a) point out that if insufficient storage is available with which to practice deferred irrigation, then effluent applications should be applied at the lowest rates possible during the critical wettest times of the year.
Where possible, the application of effluent to land should be made to those soils on a dairy farm that have low risk of surface runoff and bypass flow. However, on soils with high and medium bypass flow risk, where drainage is extensive, application should ideally be made to those paddocks furthest from waterways. This practice maximises the opportunity for microbial die-off to occur within the network of open ditches that link subsurface drains with waterways (Nguyen et al. 2002).
Irrigator type and operating practice can influence microbial loss: The problems associated with a non-uniform pattern of application by a standard rotating irrigator can be reduced by using the highest irrigator groundspeed, thereby applying less effluent more often to any given ground area (Houlbrooke et al. 2004b; Monaghan and Smith 2004). Furthermore, an assessment of travelling effluent irrigators found an oscillating irrigator to have a more uniform pattern of application (and hence less likely to generate bypass flow) compared with a rotating irrigator (Houlbrooke et al. 2004b).
Water irrigation
As with the irrigation of effluent, soil properties (section 4.2.1) timing, volume, location and technique are key factors in determining optimal water irrigation practices. Close et al. (2005) identify the need to avoid border strip irrigation of pasture immediately following the grazing of stock. Hedley et al. (2005) and Connolly et al. (2004) report that drainage water Campylobacter concentrations fall at least an order of magnitude if the interval between grazing and irrigation increases from 1 to 7-10 days. A delay between grazing and irrigation permits change in the physical and chemical properties of faecal material deposited by livestock and (usually) some net microbial die-off to occur, reducing the leaching of microbes to groundwater.
Accounting for the soil moisture deficit when determining the volume of water to be applied can lead to reduced flows to, and hence microbial contamination of, groundwater. However, it should be noted that the practice of border-strip irrigation involves flooding of the soil surface and hence, even if reduced water volumes are applied, the potential for preferential or bypass flow to depth remains. Results from the study by Close et al. (2005) indicate that under normal irrigation practice, spray irrigation results in much less bypass flow and a much lower microbial contamination of groundwater than the border strip technique. An effective mitigation measure would be to convert from border strip irrigation to spray irrigation, or to limit border strip irrigation to areas where there is less potential for microbial leaching, e.g., upon soils with few macropores, and regions with a deep vadose zone and groundwater table.
4.2.3 Advance pond systems treatment of dairy shed effluent
Advanced Pond Systems (APS) are a pond-based upgrade option for conventional 2-stage ‘oxidation’ ponds, and have particular application where soil and climatic conditions are unfavourable for land application of effluent. APS consist of four types of ponds arranged in series (an anaerobic - for dairy applications - pond, a high rate pond, algal settling ponds, and a maturation pond) that result in effluent of a considerably higher quality (notably, far higher microbial quality) than the traditional two-stage oxidation ponds, with opportunities for resource recovery (energy, nutrients, water) (Craggs et al. 2004). The effluent from the maturation pond can be irrigated onto land, to provide even greater overall treatment.
4.2.4 Treatment of drain flows within constructed wetlands
Recent studies using constructed wetlands have shown potential in the treatment of drain flows under grazed and irrigated dairy pasture, particularly with respect to nutrients (Tanner et al. 2005). This mitigation measure also has the potential to attenuate faecal microbes within drain flows, an aspect that is the subject of ongoing research.
Grazing location
The identification of appropriate (and conversely, inappropriate) locations for livestock grazing can lead to a reduction in faecal contamination of waterways. Aside from excluding or encouraging stock away from riparian areas (sections 4.1 and 4.2.6), water quality improvements can be realised from fencing stock out of wetlands and seepages on pastoral land. For example, studies of hill-country wetlands (Collins 2004) have shown that cattle are strongly attracted to the smaller shallower wetlands for grazing, though not the larger deeper ones, presumably for fear of entrapment. Consequently, these wetlands are critical source areas with respect to faecal microbes, sediment and nutrients. Such wetlands have been shown to attenuate nitrate through the process of denitrification, provided that water moves through the wetland slowly enough (Burns and Nguyen, 2002; Rutherford and Nguyen, 2004). Modification of wetland drainage through cattle trampling, installation of subsurface drains or artificial channels is, therefore, likely to diminish their nitrate attenuating properties.
The practice of stock exclusion can be extended to those paddocks located adjacent to waterways that are characteristically prone to saturation, during wet weather. Such paddocks are vulnerable to pugging damage in the very type of weather (rainstorms when antecedent soil moisture is already high) that are most likely to generate surface runoff that can wash faecal matter directly to water bodies. Grazing rotations on dairy farms could be arranged such that when a heavy storm event is predicted to occur, cows can be grazed on paddocks away from permanent waterways. This process would involve managing pre-grazing pasture covers that allowed for alternate 12-24 hr grazing of riparian and non-riparian paddocks.
Close et al. (2005) suggest that cows should be grazed down (groundwater) gradient, or as far away as possible, of wells for at least a week prior to and during border strip irrigation events. Permanent fencing to exclude stock around wells is also expected to be beneficial (Close et al. 2005). Further protection of wells can be achieved by sealing (concreting) an area of at least 1 m diameter around the wellhead. This prevents infiltration of microbes through the permeable disturbed material adjacent to the well.
During prolonged spells of wet weather, improved water quality may result through the relocation of stock from paddocks to feed or wintering pads, and herd homes. Appropriate disposal of effluent is, however, required to ensure that benefits to water quality are realised.
4.2.6 Riparian buffer strips
Fencing to exclude livestock from stream channels and a proportion of riparian land has the potential to be a particularly effective measure in reducing the faecal contamination of pastoral streams. Not only does this prevent the deposition of faecal material directly into streams and near channel contributing areas (section 4.1) the dense vegetation associated with riparian buffer strips (RBS) reduces the momentum and magnitude of surface runoff, thereby aiding infiltration and promoting the entrapment of faecal material and other agricultural pollutants (Parkyn 2004). Furthermore, riparian buffers also benefit stream habitat, notably by the shading provided by shrubs and trees (Parkyn et al. 2003). It is the attenuation of pollutants washed in by surface runoff that is the focus of this section, with the development of guidelines for optimal RBS design with respect to faecal bacteria (Table 1). It should be noted that these guidelines do not encompass the intermittent grazing of livestock within riparian areas, but such an approach may have still benefits for water quality compared to continuous grazing of the riparian zone. For example, the removal of stock during periods of wet weather is likely to improve water quality, although clearly the improvement would be less marked than that realised through permanent stock exclusion.
Studies under the PTRRP and elsewhere (Collins et al. 2005b) have shown that the effectiveness of RBS in attenuating faecal microbes washed in by surface runoff is influenced by: slope angle, soil type, buffer width, the type of faecal material and the rate of surface runoff. However, it is not possible to derive quantitative RBS design guidelines, from these relatively few experimental studies, that are widely applicable across pastoral land within New Zealand. To do so would require experimental work to be undertaken across the whole range of soils, slope angles, buffer types, and magnitude of rainfall events found within New Zealand. Instead, quantitative guidelines for RBS with respect to faecal bacteria (Collins et al. 2005b) have been derived from those reported for sediment attenuation. These, in turn, were derived from a detailed sediment modelling study (Collier et al. 1995) that captured the variability of the New Zealand pastoral landscape, and simulated the effects of a permanent buffer characterised by dense vegetation.
A current lack of understanding of the form in which faecal microbes are transported in surface runoff limits our ability to predict buffer effectiveness for bacteria with confidence; microbes transported as single un-attached and dispersed particles will be less readily trapped in a buffer than those transported in aggregates of faecal material or soil. The RBS bacteria guidelines attempt to account for this uncertainty by including estimated efficiencies for varying degrees of bacterial attachment (Collins et al. 2005b). This was achieved by assuming that the % clay content of each soil in the original sediment modelling could be used as a surrogate for the % of dispersed (non-attached) bacteria. In addition, it was assumed that the % of soil particles greater than clay-sized represented the % of bacteria transported in an attached form. Whilst this approach is a crude approximation, it has the advantage of being based on the similarity in size of clay particles and single faecal bacteria: clay particles are < 2 μm, whilst bacteria typically range between 0.3 and 2 μm, with E. coli 1-2 μm. It is appreciated that, typically, the degree of bacteria attachment will be unknown at any given location, and is likely to vary with a number of factors, including the type of faecal material, the amount and type of eroded soil and, rainfall characteristics. However, the inclusion of varying levels of attachment in these bacterial RBS guidelines provides an indication of the sensitivity of optimal buffer width to this factor.
The quantitative guidelines for buffer design with respect to faecal bacteria are presented in Table 1. These illustrate the optimal buffer width for each combination of slope, soil drainage and bacterial attachment, where the optimal is defined as giving the best return in terms of efficiency for the amount of land given over to the buffer (Collier et al. 1995). Important caveats are associated with the derivation of these guidelines, however, and these require that the reported efficiencies should be treated as a ‘best-case’. These caveats are as follows:
- The proportion of bacteria transported as unattached in the derivation of Table 1 was assumed to range between 10% and 60%. In reality, however, this figure may, at least on occasion, be higher than 60%, thereby reducing buffer efficiency. For example, Muirhead et al. (2005) report attachment fractions of E. coli to sediment of ≤25%, i.e., at least 75% of bacteria in this study were transported unattached.
- Whilst most bacteria may approximate a clay particle in size, other bacteria and all viruses (25-350 nm) are considerably smaller. Unless a large proportion of these smaller microbes attach to other particles their entrapment efficiencies will be lower than those reported in Table 1. Furthermore, bacteria are less dense than mineral clays and hence, even if they are of comparable size, are less likely to deposit within a buffer, assuming all other factors are equal.
- Results from the PTRRP (Collins et al. 2004) have shown that microbes trapped in a RBS can be remobilised in a subsequent rainfall event some days later. These guidelines do not account for survival and re-mobilisation.
- The guidelines report long-term average efficiencies over numerous individual rainfall events of different magnitude and frequency. Efficiency will, therefore, be lower in large rainfall events. Conversely, efficiency will be higher for low magnitude rainfall events.
Table 1: Estimated optimal width and efficiency for RBS with respect to faecal bacteria. Buffer width is given as a percentage of hillslope length. Buffer efficiency is expressed as a percentage reduction, and represents a ‘best-case’ estimate of average efficiency.
| Slope | Soil Drainage Rate |
Bacterial Attachment |
Buffer Width | Buffer Efficiency | Notes |
|---|---|---|---|---|---|
| Flat to Undulating | Low | High: ≈ 90% | 1 | 95 | On well-drained soils, RBS may not be warranted since vertical movement of water and microbes dominates. Poorly draining flat to undulating land often has artificial subsurface drains. High intensity rainfall, however, can generate significant surface runoff on poor or imperfectly drained soil, even with artificial drainage. |
| <1-4 mm/h | Medium: ≈ 70% | 5 | 90 | ||
| Low: ≈ 40% | 9 | 80 | |||
| Medium | High | 1 | 95 | ||
| 5-64 mm/h | Medium | 2 | 90 | ||
| Low | 4 | 80 | |||
| High | High | 1 | 95 | ||
| 65‑250 mm/h | Medium | 1 | 95 | ||
| Low | 3 | 85 | |||
Rolling to Moderately Steep |
Low | High | 2 | 90 | Generally, these are the most appropriate slope angles for RBS since sufficient surface runoff is generated, and as spatially diffuse sheet flow rather than concentrated in rivulets or channels. |
| Medium | 7 | 70 | |||
| Low | 15 | 50 | |||
| Medium | High | 1 | 95 | ||
| Medium | 4 | 80 | |||
| Low | 11 | 55 | |||
| High | High | 1 | 95 | ||
| Medium | 2 | 85 | |||
| Low | 4 | 60 | |||
| Moderately steep to Very Steep | Low | High | 5 | 45 | RBS efficiency can be limited by topographical convergence of surface runoff, causing channelised flow. Buffers may need to extend some distance upslope, following flow pathways. Exclusion of stock from critical source areas (e.g., wetlands, flow pathways) is an important mitigation measure. |
| Medium | 15 | 30 | |||
| Low | 30 | 20 | |||
| Medium | High | 3 | 60 | ||
| Medium | 7 | 50 | |||
| Low | 13 | 35 | |||
| High | High | 3 | 75 | ||
| Medium | 4 | 70 | |||
| Low | 11 | 50 | |||
Contact for Enquiries
Phil Journeaux
Manager
North Island Regions
Sector Performance Policy
MAF Policy
Private Bag 3123
Hamilton
NEW ZEALAND
Phone: +64 7 957 8313
Fax: +64 7 957 8315
Contact this person
