• About this site|
• Site index|
• Contact us|
• Jobs

A Monitoring and Reporting Strategy for the Dairying and Clean Streams Accord

1 Summary

2 Introduction

3 This Project

4 Measuring Progress Towards Meeting the Components of the Accord

5 Measuring the Environmental Outcomes of Implementing the Accord

6 Establishing the Tier 2 Catchments

7 Reporting

8 A Proposed Monitoring and Reporting Strategy

9 Suggestions for Future Best Practice

10 Recommendations

11 Acknowledgements

12 References

Appendix 1: The Project Brief

Appendix 2: Summary of Responses from Regional Authorities

Appendix 3: The Best Practice Dairy Catchments

Appendix 4: Base Information from a Proposed Tier 2 Catchment

5 Measuring the Environmental Outcomes of Implementing the Accord

This chapter presents an outline of how the environmental outcomes of implementing the Accord may best be monitored. Section 5.1 reviews the challenges of trying to quantify any broad scale environmental benefits that result from actions taken under the Accord. Section 5.2 establishes a set of criteria for assessing any monitoring components. Section 5.3 sets out the recommended elements of a strategy to determine the environmental outcomes of implementing the Clean Streams Accord.

5.1 The challenges

There are some significant challenges in preparing a monitoring and reporting strategy to measure the environmental outcomes of implementing the Accord. Perhaps the greatest single challenge is how to separate out the benefits of the often relatively minor on-farm management practice changes resulting from the Accord (in an overall catchment context) from all the other activities occurring in a particular catchment. There are many factors that can affect water quality in a catchment, and entirely separating out the effects of implementing the Accord from this 'background noise' is not possible using the general state of the environment (SoE) monitoring data currently collected by all councils. As a consequence, a more targeted programme is recommended that builds on the existing SoE monitoring where possible, and that will provide sufficient information to enable a broad picture of what environmental benefits the Accord is having.

There are many factors that contribute to the background 'noise' in a catchment including:

(a) naturally occurring events such as floods, which can have major effects on suspended sediment, turbidity and in-stream biota

(b) changes in farm personnel, such as the sharemilker, with subsequent changes in environmental performance

(c) changes in land use or land-use intensity, such as changing from sheep to deer farming, bringing more beef cattle on to a particular property, or increasing dairy herd size

(d) the amount of fertiliser used by all farmers, and the timing of its use and how it is applied. Aerial topdressing of hill country may, for instance, have significant impacts on phosphate levels and algal biomass in streams. Similarly, if there is heavy rainfall within several days of fertiliser application, much of that fertiliser may run off to water

(e) stock access to streams, not just by dairy cattle on smaller streams but also beef cattle and deer

(f) fences that are too close to streams, or which are broken or poorly maintained and allow stock access to streams

(g) discharges to water, including those of dairy effluent

(h) wintering dairy stock off-site - this has potential major benefits for water quality during these months

(i) run-off from land irrigation of dairy effluent entering watercourses from mole or tile drains, or from the mechanical failure of irrigation equipment

(j) factors such as changing a farm dairy discharge from water to land.

There are also many regulatory interventions that may have significant impacts. These could be as simple as requiring a farmer with poor environmental compliance with effluent discharge consent conditions to improve that performance, with consequential benefits for water quality. Most regional authorities also encourage all farmers to fence out streams, retire and plant stream margins, and improve management of discharges of dairy effluent to land. Such factors further confound the separation of the effects of implementing the Accord from all else that goes on in a catchment.

Many streams in dairy catchments have soft bottoms, which typically do not support diverse stream invertebrate communities. This is an issue because measuring the 'health' of stream communities, through indices such as MCI or QMCI, is an excellent way of measuring any improvement attributable to improved management practices such as implementing the Accord (although such soft-bottomed stream indices are being developed).

Other streams in dairy catchments have high levels of silt covering stony-bottomed beds. A very good example is the Waikakahi Stream in South Canterbury, which is one of the four best practice dairy catchments. This silt, which is likely attributable to stock having access to streams, also affects the 'health' of stream communities and can limit their value for trout spawning. Once in a stream, silt is only likely to be removed during large floods.

Another possible confounding factor is the 'lag time' between an action being taken, and the effects of taking this action. The exact nature of this lag time is not well known, but it is accepted that applying nutrients to land, for instance, does not have an immediate effect on water quality (unless the nutrients directly enter water) and that the lag time before changes are seen in surface water quality can span several years. Any monitoring programme needs to be long term, and it may be that clear results are not evident for some time.

5.2 Proposed assessment criteria

Prior to considering and recommending an appropriate national monitoring strategy for the environmental outcomes of the Accord, it is necessary to develop a set of criteria against which potential programme elements could be critically assessed. These criteria form the 'critical success factors' that the programme must achieve in order for it to be justifiable and robust. The assessment criteria are aimed at the catchment level and are as follows.

5.2.1 Nature of the catchment

Ideally any monitoring should be undertaken in catchments which are very largely used for dairying, or where dairying occurs in a largely discrete part of the catchment. In either case, monitoring of water quality and biota at one or more control sites upstream of the influence of dairying, and other sites downstream of a cluster of dairy farms, will allow the effects of dairying to be reasonably well separated from the effects of other farming activities in the catchment.

An ideal catchment would be one where the land use upstream of dairying is either native bush, or extensive traditional dry-land pastoral farming. Streams that drain from native forest (e.g. from the Ruahine or Tararua Ranges or Mt Taranaki), or spring-fed streams, are ideal examples.

There should be no major point source discharges to the catchment upstream of the lowest monitoring point, apart from on-farm dairy effluent discharges. Ideally, there will be little summer water abstraction for consumptive use.

Catchments monitored should ideally be smaller in size, of say 5 to 50 km2, as these are likely to provide situations where the influence of other factors that may confound the interpretation of the monitoring will be fewer and more easily isolated.

The catchments selected should be spread geographically across the country and represent a range of stream types. The River Classification Database (REC) should be used to type any selected catchments.

5.2.2 Implementation of the Accord

Ideally implementation of the three parts of the Accord relevant to water quality (stock exclusion from streams and their banks, bridging or culverting of regular crossings, and nutrient budgeting) would initially be little progressed in the catchments selected for intensive monitoring. This may seem counter-intuitive, but it would allow for baseline or control data to be obtained against which any changes resulting from the Accord's on-farm management practices could be assessed.

Alternatively (and perhaps more likely), a control site should be available upstream of the dairy farm properties to enable 'control' water and habitat quality data to be obtained. In these situations, however, only the impact of on-farm changes undertaken after the initiation of downstream water quality data collection will able to be assessed.

5.2.3 Existing water and habitat quality

The more degraded the existing water and habitat quality is downstream of the dairying properties, the easier it would be to separate the benefits of implementing the Accord through measurable improvements in water quality. As noted above, however, high levels of silt covering stony-bottomed beds of creeks will reduce measures of stream health, and will only be removed during floods.

Alternatively, where the background water quality above the farms is high, one might expect downstream water and habitat quality degradation to be evident. This would provide evidence of the adverse effects of dairying activity and the nature and scale of that degradation. Any improvement resulting from implementation of the Accord could then be assessed.

5.2.4 Knowledge of land use

One of the major difficulties in implementing this strategy is that the findings will only be as good as the information available about all land-use changes in a catchment above the downstream monitoring site. It will be necessary to understand many of the environmental factors listed in section 5.1 above, if there is to be any realistic prospect of linking cause (implementation of the Accord) to effect (any environmental outcome). This reinforces that small-scale catchment monitoring programmes focusing on discrete blocks of dairy farms are likely to be much more useful than general catchment-wide monitoring.

5.2.5 Existing water quality data

Ideally, there would be some information gathered over at least five to 10 years and at least at a bi-monthly frequency to determine existing water quality in the catchment. Ideally also, some biological monitoring of algal and macroinvertebrate populations will have been carried out. This could have been carried out as part of a SoE programme, or undertaken as part of more specific investigations. [In general terms most SoE programmes are either insufficiently intensive (e.g. there might be one site in a medium-sized catchment versus the three to four that may be required to show any benefits of implementing the Accord), and/or have only been initiated in the last few years, and/or sampled too infrequently to show anything other than broad trends.]

Although a lack of existing background information may indicate a particular catchment would not be suitable for inclusion in this strategy, it should not be dismissed for this reason. Given this is a long-term strategy, there is time to collect that background information in emerging dairy areas likely to be subject to future land-use changes (dairy conversions), but ideally there will be good baseline material available to start with.

Similarly, there is also time to collect information in catchments which are presently in intensive dairying but where the implementation of Accord targets has made little progress.

In terms of what needs to be monitored in rivers and streams, any programme should ideally include the following range of parameters:

Table 3: Suggested in-stream monitoring parameters

5.2.6 Intensive or extensive monitoring, or both

One of the matters that needs careful consideration is whether monitoring should be carried out extensively over an entire year, intensively over a shorter timeframe (which would be repeated in selected years), or a combination of these approaches. An associated issue is whether intensive monitoring needs to be carried out every year.

Monitoring of water quality is most likely to show improvements due to implementation of the Accord over the drier summer and autumn months when stream flows are generally low. This is because there will be less direct contamination of small and ephemeral watercourses and less risk of effluent discharge run-off directly to water (as might occur due to entry via mole or tile drains, or by overland flow when soil moisture levels are high). The results will also not be confounded by stock being wintered off-site.

Biological monitoring is best carried out during summer months when larger instars of insects are present, allowing for more straightforward identification.

Intensive monitoring has the advantage that the effects of some of the possible confounding factors listed in section 5.1 above can be more readily identified, and so aberrant results can be more readily attributed to a distinct cause.

The difficulty with intensive monitoring in isolation is that it may be too infrequent to readily show statistically valid, long-term trends. Ideally, a mix of extensive, monthly water quality sampling of basic parameters, along with more intensive monitoring during a relatively short period over summer, is preferred. In saying this, it is recognised many councils could find it difficult to fit such work in over the busy summer period.

The extensive monitoring should be sufficient to show any long-term trends. Biological monitoring need only be undertaken during the summer period.

Furthermore, it is not necessary to carry out intensive monitoring over each and every summer. Once intensive sampling has been carried out to establish 'baseline' condition, further intensive sampling need only be carried out once every three to five years, with more extensive monitoring in between times. This will be quite adequate to show long-term trends.

This is discussed further in section 6.2 below.

5.2.7 Monitoring of groundwater?

At this stage there is no recommended additional systematic monitoring of groundwater quality as part of assessing the benefits of implementing the Accord. This is because the Accord's primary focus is on minimising the impacts on surface waterbodies. The main contaminant in groundwater resulting from intensive land use is nitrate-nitrogen, with some associated faecal coliform contamination of very shallow groundwater. While nutrient management may help reduce nitrate levels in soils, it is phosphate that is 'overfertilised' to the greatest extent on dairy farms.

Groundwater monitoring could be included if the targets of the Accord were to change. Two practices that would significantly reduce nitrate-nitrogen leaching to groundwater are wintering of dairy stock off-site, and the use of nitrification inhibitors. If it is proposed that these, or similar, practices be added to Accord targets sometime in the future, then additional groundwater monitoring should be carried out.

There are some existing studies looking at the impacts of dairying on groundwater, including two very comprehensive compliance programmes in Hawke's Bay. These could be used as case studies that supplement the work proposed in the Tier 2 catchments below.

5.2.8 Timeframe

Any national monitoring strategy must by definition be long term, given the significant lag times that can arise between land-use changes (including the Accord on-farm management practices) and subsequent effects on water quality.

As a minimum, the programme needs to run for at least 10 years, and preferably 15 years. It is also likely that it will take at least five years for any definitive results to be gathered. Accordingly, reporting is recommended at five-yearly intervals.

Within the timeframes recommended it is likely that Accord targets will change, with perhaps additional targets added. This will not affect the robustness of the proposed programme, and, depending on the nature of the targets adopted, may assist to show the environmental benefits of implementing the Accord.

5.2.9 Provision of supporting information

Any monitoring programme will require at least four separate sets of information to determine existing condition for the catchments as follows:

• detailed information about dairying land use in the catchment [It would be desirable to understand the history of dairying in the catchment in at least broad terms. This is because there is often a lag time between land-use changes and the impacts being observed in watercourses.]
• information about other land uses in the catchment
• a robust assessment of progress towards Accord targets
• an assessment of existing water quality information in the catchment.

Responsibilities for how this information will be collected and/or made available are set out in section 6.

5.2.10 Will monitoring change behaviour?

One factor that cannot be removed from any recommended monitoring programme is the extent to which farmers in catchments that are being monitored, or on farms where more intensive monitoring programmes will take place, will change their behaviour and improve their environmental performance such that they behave better than the 'average farmer'. [By, for instance, constructing more secure fences preventing stock entering waterways, or by undertaking more robust nutrient management.] For instance, in the four best practice catchments farmers are being strongly encouraged to undertake on-farm management practice improvements. This may bias the results from the monitored catchments, which could be problematic if those results are scaled up to provide national-level information.

There is no easy way around this potential problem. Only 'samples' can be taken in a programme like this, and there is no way of knowing if those samples are atypical. If, however, the monitored catchments do show improvements, this can be safely extrapolated to say that other catchments where progress has been made towards implementing Accord targets will also show improvements in environmental and habitat quality. This is the stated target of implementing the Accord. [An MSc study just initiated by Gymn Morgan of the ARC looks at the attitudes of farmers to the Accord. The information gathered in that study may help better understand this question. What has been very clear to us in talking to regional authority staff is that the Accord has had a very positive effect on the attitude of dairy farmers to the environment.]

5.3 National and regional data sets

It is clear from the discussion above that neither national nor regional data sets derived from existing national (or aggregated regional) SoE programmes can be used with any reliability to measure the benefits of implementing the Accord. There are too many confounding factors, and the monitoring at such a broad level spatially, to expect any benefits of the relatively minor and highly localised land-use interventions occasioned by the Accord to be detectable in national or regional water quality data sets.

Indeed, it is considered inappropriate to make any such links given the impacts of all possible land-use changes, the potential effects of point source or unauthorised discharges, activities such as water abstraction, and natural perturbations such as floods and droughts, on water quality and biota.

However, such data sets will be useful in providing the overall broad-scale environmental trends against which Accord-specific data can be assessed.

5.4 Catchment data sets

The best chance of monitoring the environmental outcomes of implementing the Accord is using a three-pronged approach in:

• the best practice dairy catchments
• the Tier 2 catchments
• site-specific studies.

These are described in turn.

5.4.1 The best practice dairy catchments

The first element of this approach is continuing work in the four intensively-studied best practice dairy catchments, namely:

(a) the Toenepi catchment in the Waikato

(b) the Waiokura catchment in Taranaki

(c) the Waikakahi catchment in the Waitaki Basin

(d) the Bog Burn catchment in Southland.

Each of these catchments is discussed in Appendix 3. [Another intensive catchment study has started recently at Inchbonnie in the Lake Brunner catchment on the West Coast. We do not consider that catchment here as it is outside of the regions where Fonterra operates.] In doing so, however, it is noted that there are two significant hindrances to using these catchments to monitor the benefits of implementing the Accord:

1. Strong effort has already gone into improving dairy practice in these four catchments. This means, for instance, that farmers have already been encouraged by scientists working there to fence out streams, manage nutrient levels and the like. In other words many of the elements of the Accord have already been undertaken in these catchments. This, along with the other best practice initiatives, will make it difficult to separate out the benefits of implementing the Accord in its own right.

2. Of the four catchments, only the Tonepi and Bog Burn are considered 'typical' dairying catchments by the relevant regional authority. The Waiokura is regarded as quite atypical by the Taranaki Regional Council in that it rises on the upper slopes of the ring plain, rather than on the higher slopes of Mount Taranaki as most streams in that region do. The Waikakahi is not typical of Canterbury streams as it is irrigated, with average summer flows being nearly an order of magnitude higher than winter flows.

5.4.2 The Tier Two catchments

The second element of monitoring the environmental outcomes of implementing the Accord is to initiate (or continue) intensive monitoring in a further eight or so representative small catchments around the country which broadly meet the assessment criteria outlined in section 5.2 above. These are referred to as the 'Tier 2' catchments. Information will also be needed about land use and land-use change in these contributing catchments. Assessing the impacts of land-use interventions at a small scale will allow the effects of those changes to be determined in a semi-quantitative manner. For example, over time it may be possible to quantify the improvements to water quality achieved by fencing out all Accord streams on small groups of dairy farms. [Obtaining similar data sets from several different councils would enable different land forms, climatic conditions and stream types to be accounted for in a very rough order manner.] This level of benefit could then be scaled up to derive regional or national benefits that might be achieved if all farms complied with the Accord targets.

The catchments which we propose be included in this programme based on discussions from regional authorities, are as follows.

5.4.3 Site-specific studies

The third element of the recommended strategy is to carry out site-specific studies on the environmental outcomes of implementing one or more aspects of the Accord on a few properties, or even one property. This could be, for instance, by pre and post monitoring once one or more crossings are put in, and some fencing carried out. Ideally, such farms would be the catchment's most upstream dairy farms with native bush headwaters. [In saying this we are aware that some such studies already exist. We have not reviewed these at this time.] No specific recommendations are presented at this time, but future studies may be needed.