Archived publication
This publication is no longer current or has been superseded.
6 Discussion
The picture of water quality in New Zealand rivers that has emerged from the analyses and summaries presented here is consistent with a shift in relative importance of point source vs. non-point source pollution as key anthropogenic pressures on surface waters.
The summary of 2005 data (Section 3) indicates strong associations between nutrient concentrations and percent pastoral land cover at the national scale. Median concentrations of all nutrient species and levels of the faecal indicator bacteria E. coli were positively correlated with extent of pastoral land cover. 2005 is the first year of E. coli analyses in the NRWQN and it is encouraging to see its usefulness as a national-scale indicator of water quality correlated with pastoral land use. It should be noted again that percent pastoral land cover in the catchment does not allow a clear cause-effect relationship to be inferred between water quality patterns and non-point source pollution because several of the NRWQN sites are downstream of significant point sources (eg, Waingongoro River), and this has not be factored into the analysis.
In Section 4, summaries of annual state over time (1989-2005) highlighted a number of patterns of significance to resource managers. For example, concentrations of NOx-N in rivers with naturally low concentrations (ie, 5th percentile rivers) showed a significant decreasing trend with time, whereas rivers with very high levels of NOx-N showed a significant increasing trend over time. Given the strong association of NOx-N concentrations with pastoral land cover it is reasonable to conclude that rivers draining large areas of pastoral land have deteriorated significantly over the last 17 years with respect to NOx-N concentrations. Levels of DRP showed a somewhat different pattern. While there was a significant increasing trend for DRP concentrations at the 80th percentile, there was not a similar trend in the 95th percentile. DRP concentrations at the 95th percentile actually show a non-linear response over time with concentrations in our most enriched rivers peaking in the late 1990s and showing a decreasing trend since.
Other patterns worthy of note are the increasing temperature trend in our coolest rivers, and the reducing pH trend at the 95th percentile. Both of these trends could be associated with climate change but the level of analysis carried out so far is insufficient to allow more than speculation at this stage. Increasing levels of atmospheric CO2 could lead to reductions in pH in waters as a result of increasing dissolution of CO2 to form carbonic acid.
Detailed trend analysis for the 1989-2003 period confirmed some of the patterns identified from the higher-level analysis in Section 4 (eg, increases in TN) but also produced some inconsistencies. For example, the moving median trends in Section 4 showed relatively weak increasing trends for DRP over the 1989-2005 period whereas the formal trend analysis in Section 5 identified a national-scale trend of increasing DRP concentrations over the 1989-2003 period. This inconsistency probably reflects the differing time periods used for analysis and highlights a major issue with interpreting trend analysis results. Interpretations should always be restricted to the period of record as it is dangerous and inappropriate to extrapolate trend results beyond the period of record.
The strong decreasing trends in concentrations of ammoniacal nitrogen and biochemical oxygen demand observed over the 1989-2003 period are consistent with water quality improvements that would be expected from reductions in point source pollution. Increases in visual clarity may also reflect improvements in point source management, although improvements in farming and forestry practices could be drivers of improving water clarity. However, median clarity was negatively correlated with percent pastoral land cover, so it is seems more likely that clarity improvements are associated with improved point source management. In contrast, increasing trends in NOx-N, total nitrogen and dissolved phosphorus are consistent with increases in non-point source delivery of contaminants to waterways.
Overall, the results derived from state and trend analyses of data from the NRWQN over the last 17 years paint a picture highlighting the good, the bad and the ugly of river water quality in New Zealand.
Improvements in levels of BOD, ammoniacal nitrogen and visual clarity are very positive signs and should be highlighted as success stories where clear cause-effect relationships between removal or improved management of point source pollution and improved water quality can be established.
The decreasing trends in pH and increases in conductivity are difficult to interpret and are of relatively low magnitude. However, the pH result in particular highlights the potential for climate change to have significant consequences for water quality and aquatic ecosystems. Further research is needed to identify what effects future changes in pH and temperature might have on our rivers.
The least positive results from these analyses are the increasing trends in total nitrogen and phosphorus and the significant increases in NOx-N observed in our most enriched rivers. The challenge for resource managers and industry over the next decade or more is to control non-point source contamination of our waterways.
