Skip to main content.

Summary and Recommendations

This section provides a summary of baseline land use and water quality in the 14 monitored catchments, drawing on information in the previous two sections. It then offers some considerations and recommendations about the ongoing use of regional data sets, and the requirements for analysis to enable future national-scale trend assessments.

Summary of catchment characteristics and water quality

The monitored catchments discussed in this report differ from each other in a number of respects that have implications for measuring national water-quality outcomes related to dairy land management (including initiatives such as the Dairying and Clean Streams Accord).

Catchment features and land use

The catchments range in size from about 6 to 211 km2. The smallest catchments (the Rhodes, Petrie and Powell) have the most detailed land-use information available as a consequence of the relative ease with which these catchments can be surveyed. For example, in these catchments there is accurate council survey information, which augments the Fonterra figures on the extent and condition of stream fencing and crossings.

Although most of the Accord targets are reported to have been met (or to be close to being met) in the majority of the monitored catchments, there are two catchments (Enaki and Puwera) where significant progress has yet to be made in fencing and culverting.

For most catchments there is no information on the extent to which nutrients are being actively managed (as opposed to nutrient budgets being in place). It is probably reasonable to expect, on the basis of statements in the 2006/2007 Snapshot of Progress report (Ministry for the Environment, 2008), that many of the monitored catchment nutrient budgets are not being fully implemented.

The drainage regime, geology and upstream land use of each catchment differs. Of particular note are:

  • the Puwera catchment, where the variability in flow means that much of the main-stem stream is not perceived as an Accord waterway, and therefore has not been subject to Accord actions

  • the Washpool catchment, where the prevalence of mole-and-tile drains appears to be having a dominant influence on water quality

  • the Rhodes and Petrie catchments, where free-draining substrates and nitrate–nitrite nitrogen-enriched groundwater recharge of the streams means that water quality in these catchments is driven to some extent by land-use practices external to their area of surface water capture.

Water quality

Concentrations of nitrogen (total, nitrate–nitrite nitrogen and soluble inorganic nitrogen) are close to, or in excess of, guideline thresholds in almost all of the monitored catchments. However, the range of medians across catchments is considerable. For example, nitrate–nitrite nitrogen concentrations in the Rhodes catchment in Canterbury are about 40 times those in the Puwera catchment in Northland. This variation possibly reflects differences in the relative importance of nitrogen as a focal point for management response, although this will depend on observed effects.

Concentrations of phosphorus are close to, or in excess of, guideline thresholds in all of the
Tier 1 catchments but are notably lower in most of the Tier 2 catchments (more than half of the Tier 2 catchments have concentrations of total and dissolved phosphorus that are not in excess of guideline thresholds). Although most catchments are tending towards phosphorus-limiting conditions, four catchments (the Rhodes and Petrie, Powell and Taharua) are strongly phosphorus-limited, while two (Inchbonnie and Puwera) are tending towards nitrogen limitation.

Although the majority (10) of the monitored catchments had uniformly high median E. coli concentrations, the Taharua catchment in Hawke’s Bay had much lower concentrations – similar to those found in catchments with predominantly natural land use (ie, relatively undisturbed).

Macroinvertebrate scores varied widely. Five catchments had average metric scores indicating relatively clean water or mild degradation, while another five had average scores indicating moderate to severe pollution (no data was available for four catchments). Of note, one of those catchments in the highest quality class for macroinvertebrate community health, the Waiokura, could be considered one of the poorest in terms of physical and chemical water quality.

Considerations for ongoing monitoring

As set out at the beginning of this report, there are a number of challenges to successfully assessing the water-quality outcomes of best practice management in dairying catchments. Some of these challenges have been highlighted in the description of baseline water-quality results and require further consideration in the context of any ongoing monitoring programme.

Land-use information

Improvements in the coverage and resolution of land-use information are required to identify the relationships between changing farm actions and water-quality trends in the monitored catchments. In addition to the land-use activities quantified and described in this report, there are many localised activities (eg, the change in rate of take from a river water abstraction point) that may also have relatively large impacts on water quality. However, there are considerable practical and financial constraints to capturing and quantifying these types of activities in a consistent way as part of the monitoring programmes under discussion.

Potential opportunities to draw on other studies to provide additional context for the land-use information presented in this report, and any subsequent related reports, should be considered. One example is a three-year research project, started in 2007 and led by AgResearch, titled The State and Change of Industry Practice and N and P Losses from Pastoral Farms. One of the objectives of this project is to benchmark, and show change in, land-use practices (N. Botha, AgResearch, pers. comm., July 2008). One thousand farmers have been independently surveyed as part of this study and there are plans to follow up with the same farmers in 2009.

Water-quality information

The existing sampling regimes (see table A4-A in appendix 4) allow some aspects of water quality to be quantified at the catchment scale, and it is likely that net changes in measurements over time will generally be detected at this scale in the data being collected. However, it is unclear, particularly with respect to the Tier 2 catchments, whether the existing monitoring is sufficient to detect more subtle long-term changes in water quality (ie, those that are seasonally driven or of a magnitude that requires intensive sampling to detect), and/or those that can be attributed to particular dairy management actions (including those taken under the Accord).

For example, it has generally not been possible in this report to characterise water quality during particular periods of the year when dairying impacts are likely to be accentuated (eg, channel disturbance during low flows in summer and higher nutrient delivery to waterways in winter). This is either because an intensive seasonal monitoring programme has not been undertaken, or because more data analysis is required at both the regional and national scale.

One exception is the Powell Stream, where continuous data logging of temperature and dissolved oxygen was carried out during the summer months of 2006/07; dissolved oxygen results for this period are considerably lower (worse) than the annual average results presented for the other monitored catchments. Focusing future monitoring on the time of the year when effects are most pronounced, such as has been done in the Powell catchment (and more recently in the Taharua catchment), is more likely to capture the most significant water-quality improvements resulting from changes in land-use practice.

To resolve uncertainties about whether present monitoring programmes meet national reporting needs, this report recommends that each catchment monitoring programme be reviewed, while recognising that contributing data for national reporting is only one of many outputs for these monitoring programmes. Some preliminary considerations are offered below.

Merits of the existing monitored catchments for national reporting

The five Tier 1 Best Management Practice catchments

These catchments are valuable components of a national long-term monitoring programme. Monitoring regimes are well established, with a detailed and consistent approach to water-quality measurements being taken. Routine stream monitoring is augmented by two-yearly surveys of farm management practices as well as discrete research projects,18 both of which add to the overall understanding of land-use effects on water quality. There is a wealth of water-quality data for these catchments, much of which has not been reviewed here because it is the subject of other study outputs, which should be examined for future reports.

Tier 2 catchments

  • Puwera catchment in Northland. It appears that few actions have been taken in this catchment to minimise the losses of pollutants to the stream. Northland Regional Council estimates that stock is excluded from less than 10 per cent of waterways in the catchment. The stream flow can cease altogether in summer months, which means the stream does not meet the Accord definition of “permanently flowing”. This leads to some uncertainty about the extent to which landowners will adopt Accord and other best practice management actions over time. However, high flow variability is a feature of many Northland catchments. The Puwera catchment therefore (arguably) provides a typical and realistic picture of the relationship between water quality and dairying land use in the region. If Puwera is retained as a long-term monitoring catchment, more detailed information on land-use change will be required.

  • Taharua catchment in Hawke’s Bay. Like the Tier 1 catchments, the Taharua catchment has a history of routine water quality monitoring dating back to 2000, which provides a good basis for assessing any change in the future. Studies currently underway will establish the extent of interactions between surface and ground waters (G Sevicke-Jones, Hawke’s Bay Regional Council, pers. comm., July 2008), and this may help clarify the impacts of dairying activities in the catchment. However, relatively recent (late 1990s) and continued expansion of dairying in the catchment means that any water-quality improvements resulting from the adoption of best practice management may be offset to some extent by the increase in overall dairy activity – a complicating factor that needs to be considered as part of any ongoing monitoring and reporting programme.

  • Mangapapa in Manawatu. The Mangapapa catchment is the only one of the Tier 1 and 2 catchments in which dairying occupies significantly less land area (27 per cent) than sheep and beef farming (39 per cent). Also, much of the dairying catchment area does not meet the Accord requirements for stream size and depth (K McArthur, Horizons Regional Council, pers. comm., 2009). Both of these factors mean that, for this catchment in particular, action related to the Accord and other initiatives on the dairy land may make relatively little appreciable difference to water quality over time. However, Horizons Regional Council are committed to long-term water quality monitoring at the lower end of the catchment, so it may be worthwhile exploring whether additional focus on one of the sub-catchments that is dominated by dairy land use is possible.

  • Enaki in Wairarapa. There is a history of routine water-quality monitoring in this catchment dating back to 2002, which provides a good basis for assessing any change in the future. Detailed analyses have been reported recently by Greater Wellington Regional Council (2008). Much of the monitoring work has focused on the water-quality changes resulting from riparian rehabilitation work on a small stretch of the Enaki Stream. The understanding of stream responses to very localised actions gained so far may be particularly useful for future interpretation of water-quality results relating to wider catchment land-use changes. Both Fonterra survey figures in this report and regional council comments (Greater Wellington Regional Council, 2008) suggest substantial progress has yet to be made in the Enaki catchment towards good practice. Assuming such progress will be made, the water-quality responses in the catchment will be of particular interest.

  • Powell, Rhodes and Petrie catchments in Tasman and Canterbury. These South Island catchments have quite different characteristics within the overall group of monitored catchments. Their relatively small size has enabled the monitoring councils to compile quite detailed information on land use and farm practice. Updates to this information through repeated surveys and ‘ground-truthing’ of management actions (such as the length of effective fencing in place) may enable associations between these actions and any changes in water quality to be made with greater confidence than may be the case for other catchments.

However, there are complicating factors in all three catchments: the Rhodes and Petrie catchments potentially receive nitrate–nitrite nitrogen-rich inflows of groundwater from outside the stream catchment, while relatively high nitrate–nitrite nitrogen and E. coli loadings in the upper Powell catchment are thought to originate from sheep and beef farmland.

  • Washpool catchment in Otago. Water quality appears to be influenced to a large extent by the mole-and-tile drainage system in the catchment. There is an Accord-related agreement between Otago Regional Council and Fonterra to specifically address mole-and-tile drains in the Washpool catchment and some progress has been reported. The drainage system in Washpool is typical of that in many intensive dairy farming areas in Otago and New Zealand more generally, and deserves continued investigation. Like Puwera, Washpool offers good insight into a catchment where minimising the impacts of land use may be particularly challenging.

  • Rai catchment in Marlborough. The focus of attention in the Rai catchment has been on eliminating stream crossings, and good progress has been made. Marlborough District Council has recently reconfigured the monitoring programme so that temporal trends in water quality can be measured with more certainty (Marlborough District Council, 2008). Given the commitment to ongoing monitoring and detailed information available on the status and number of stream crossings, it is appropriate to retain this catchment in any future national monitoring programme.

Recommended next steps

As noted earlier, the aim of this report is to establish a broadly representative nationwide picture of baseline water quality in dairy environments. This baseline may help, when used as a reference point over the long term, to assess how the adoption of best practice management on dairying land is affecting water quality (along with other local and regional studies). Ultimately, however, the utility of the report depends on the extent to which the monitoring challenges that have been identified or restated in this report can be addressed. It is therefore recommended that the next steps should include – but not be limited to – the following.

Review of Tier 1 and 2 monitoring programmes

It is recommended that each of the Tier 1 and Tier 2 monitoring programmes be reviewed by relevant agencies to assess their ‘fit’ with the objectives of the national programme. Such a review should be coordinated by the Ministry for the Environment to help promote a consistent process, and should begin by clearly re-confirming, and agreeing between parties, the national monitoring and reporting objectives.

Subsequent to this, other review considerations should include the following.

  • New/additional land-use data requirements – determine what improvements in land-use information might be possible. Examples of useful additional information include:

    • council consent records of stream allocation (eg, as a percentage of mean annual low flow) and the number of discharges to land and water
    • detailed independent surveys of channel length, the extent and effectiveness of channel and bank protection (including riparian buffers), and nutrient and effluent management (including fertiliser application).

  • New/additional water-quality data requirements and reporting formats – this report has identified a number of areas where it would be informative to modify the type and format of data reported.

    • Presentation of 5th, 25th, 75th and 95th percentiles in addition to data ranges would better capture the distribution for each water-quality variable. This is particularly important for the E. coli data to make comparisons with appropriate contact recreation guidelines. In some cases it is also more appropriate to use mean values than medians (eg, when comparing nitrogen and phosphorus with periphyton guidelines), and a consistent approach should be taken here.
    • Information on periphyton and macrophytes is currently patchy across the monitored catchments. Nuisance growth is a useful – if somewhat subjective – indicator and should be a routine measurement in all catchments. In addition, consideration could be given to introducing visual clarity and fine sediment deposit measurements in those catchments where it is not currently routinely measured.
    • In-depth analysis of the upstream to downstream gradient in water quality would be useful. This might include statistical comparisons of the site data, along with assessments of the frequency and strength of positive and negative gradients across sampling events and seasons.

  • Data requirements for formal trend analyses – it is unlikely that the comparison of annual (or multi-year) medians will be sufficiently statistically robust to analyses for trends until long data sets have been assembled. For shorter time periods (eg, five years), conventional wisdom suggests, as a minimum, monthly data should be gathered (eg Scarsbrook and McBride, 2007). While this is already happening in most catchments, quarterly data is being gathered in some.

  • Data requirements to detect extremes, seasonal and/or flow-related differences in water quality – in addition to the year-round monthly data being collected in the target catchments, continuous monitoring could usefully be focused on periods when land use is likely to have the most detrimental effect (and where land-use changes may produce the largest gains); for example, determining changes over consecutive summers (during low flow conditions) in the minimum oxygen and maximum water temperature profiles. Flow-related analyses of several water quality variables could also be very informative
    (eg, determining how the faecal load in the monitored catchments is changing over time during base flow conditions, compared with the overall flow profile).

Agreement between parties on funding and resourcing arrangements

The cost of catchment water quality monitoring programmes is not insignificant. For example, the cost of monitoring and analyses of data for one year (2006/07) in four of the Tier 2 catchments was $100,000, while around $160-180,000 per year has been spent on the water quality component of the monitoring and reporting programmes in the five Tier 1 catchments. Similar ongoing levels of investment will be required to maintain the existing programmes.

However, the costs of any additional monitoring and analyses deemed necessary for national reporting will need to be considered and agreed between relevant parties. This should include consideration of the interaction between the monitoring objectives of local / central government, science and research providers and industry for each of the catchments.

Future reporting

Notwithstanding the outcome of the review of the existing monitoring programme, the monitoring and reporting strategy for the Clean Streams Accord (Ministry for the Environment, 2006) recommends a five-yearly reporting schedule starting from 2012. Under this strategy, reporting is to comprise regional reports on individual catchments (produced by regional councils) and a national summary report (produced by the Ministry for the Environment), both of which would be referenced against the baseline monitoring that has been the subject of
this report.

Although the overall life span of the monitoring and reporting programme is also likely to be the subject of further discussion, the 2006 strategy recommended a minimum life span of
10 years (from 2006/07), and preferably 15 years, to account for time lags between land-use changes and water-quality effects. Assuming these recommendations stand, the Ministry for the Environment expects to produce further reports on water-quality outcomes in the monitored catchments using 2012 and 2017 data.

 

18 For example, the results of a two-year study on nitrate–nitrite nitrogen levels in the shallow groundwater of the Toenepi catchment have recently been published by Stenger et al (2008).

 

[ |