Executive Summary
A dataset comprising water quality measurements from 15 regional councils and unitary authorities and NIWA's National River Water Quality Network was used to assess the current state and recent trends in water quality in rivers across New Zealand, and to examine spatial patterns in water quality. The River Environment Classification (REC) was used to provide a spatial framework to partition variability. The REC characterises river segments at six levels that form a spatial hierarchy from large- to small-scales: climate, source-of-flow, geology, land cover, network position and valley landform. In this study, we focused on the climate, source-of-flow, geology, and land-cover levels.
The specific objectives of the study were:
- Identify differences in water quality state among REC classes, and compare water quality state to guidelines recommended for the protection of river ecosystems.
- Identify temporal trends in water quality and compare trends among REC classes.
- Compare water quality state and trends among regions, and between regions and the national average, to determine whether water quality within classes is consistent across different regions.
- Determine whether stream size is a significant source of spatial and temporal variability in water quality.
Five water quality parameters were used in the state analyses, nitrate+nitrite (NOX), ammonium (NH4), dissolved reactive phosphorus (DRP) and Escherichia coli concentrations, and water clarity. The same parameters were used in the trend analyses, plus temperature and flow. Recent state and temporal trends in these parameters were assessed at the climate, source-of-flow and land-cover levels of the REC. Additional assessments were made at the national-scale of all natural (undeveloped) land-cover classes, and of the pastoral land-cover class. Regional assessments were made by comparing the source-of-flow and land-cover classes in each of the regions and districts in the dataset with the national average for the same classes. Inter-regional comparisons focused on Canterbury, Southland and Waikato regions. These regions have some source-of-flow and land-cover classes in common, and are the topics of case studies in upcoming Ministry for the Environment water quality issue-based reports. Comparisons among stream orders within REC classes were made because of concerns that variability caused by pooling sites from various stream orders into single REC classes obscures large-scale water quality patterns. Assessments of water quality degradation were made by comparing median values of each water quality parameter to guidelines recommended by the Australia and New Zealand Environment and Conservation Council, and the Ministry for the Environment.
National water quality state
Median plots and comparisons of water quality state at the climate level indicate that the dry climate classes (Cool Dry and Warm Dry) and the wet classes (Cool Wet, Cool Extremely Wet, Warm Wet and Warm Extremely Wet) form two groups. The dry climate classes had higher median NOX, NH4 and E. coli concentrations than the wet classes, and median nutrient and E. coli concentrations and clarity in the dry climate classes did not meet guideline values. In contrast, the wet classes had at least moderately high water quality, and the climate class with the greatest annual precipitation in the dataset, the Cool Extremely Wet class, had consistently high water quality. These results suggest a general pattern of improved water quality with increasing rainfall. However, predominant land-cover classes vary strongly among climate classes in New Zealand, which raised the possibility that the influences of land cover and land use are on water quality is equal to or greater than the influence of climate type per se.
Comparisons among source-of-flow classes indicated that three low-elevation classes, CD/L, WD/L, and WW/L had low overall water quality. Median nutrient and E. coli concentrations and clarity in each class failed to meet recommended guidelines. Six mountain, hill, and lake source-of-flow classes had consistently high water quality: CW/H, CW/Lk, CW/M, CX/H, CX/Lk, and CX/M. In these classes, median values of all parameters met the recommended guidelines. The low-elevation classes with low water quality are dominated by pastoral land-cover classes (82 - 100% of sites). The low-elevation class with the best water quality, CX/L, is dominated by indigenous forest land-cover classes. The general pattern of poor water quality in low-elevation classes may be related to the fact that most low-elevation rivers are dominated by agricultural land uses
Comparisons among land-cover classes were made using three different frameworks: fully-classified classes, land-cover classes within climate classes, and land-cover classes across climate classes. The results of these comparisons consistently showed that pastoral classes have poorer water quality than natural classes. Overall water quality state in land-cover classes can be ranked as follows: urban streams are in the worst condition, followed by pastoral, exotic forest and natural streams. Few urban land-cover classes in the dataset met any of the water quality guidelines. In contrast, most natural land-cover classes were within guideline values, except the NH4 guideline. Generally poor water quality in pastoral classes was evident within climate classes, nation-wide, and within individual regions.
Of the 265 land-cover classes that occur in New Zealand, 88 were represented in the dataset. In 59 of the 88 classes (and in 58% of the monitoring sites), median NH4 concentrations exceeded the recommended guidelines. Similarly, the E. coli guideline was exceeded in 55 classes (72% of sites), the DRP guideline was exceeded in 54 classes (61% of sites), the NOX guideline was exceeded in 32 classes (42% of sites), and the clarity guideline was not met in 31 classes (40% of sites). Comparisons of the four land-cover categories (natural, pastoral, exotic forest, urban) to the recommended guidelines revealed a distinct pattern. Nearly all sites in the urban land-cover category failed to meet the guideline for each parameter. Over 50% of sites in the pastoral class failed to meet the guideline values for each parameter, while less than 50% of sites in the natural classes exceeded the guidelines for any parameter. In particular, very few sites in the natural classes had excessive median NOX concentrations. Proportions of sites in the exotic forest classes that exceeded each guideline were generally similar to those of natural sites, with the exception of DRP and NOX, which exceeded the guideline value more frequently at exotic forest sites.
National water quality trends
All of the trends summarized below were statistically significant. The magnitude of most significant trends in REC classes was less than 1% of the parameter median per year, (although some individual sites had trends of 5 to 10% of the median per year). It is important to note that statistically significant trends are not always ecologically consequential; small trends may be significant, but the annual or long-term change indicated by the trend may not have substantial effects on river ecosystems.
At the national scale, negative trends were observed for flow and flow-adjusted NH4 concentrations, and positive trends were observed for flow-adjusted temperature and NOX. These trends were generally consistent across source-of-flow classes. Analyses of trends at the land-cover level were made using both fully classified land-cover classes, and natural and pastoral classes pooled over higher REC levels. For the fully-classified classes, the directions of significant trends were consistent across classes, and consistent with trends observed at higher levels
Regional water quality state and trends
Regional and national water quality state differed significantly in about 25% of the comparisons made at the source-of-flow and land-cover levels. In most cases, the regional average condition was better (i.e., lower nutrient or E. coli concentrations, or higher clarity) than the national average. Cases in which regional average conditions were worse than the national average are noteworthy as they may indicate groups of rivers that need close monitoring or remediation. The following cases are those in which regional average conditions were at least 50% worse than national conditions: DRP concentrations in the CW/H class in Waikato (~3 times higher than the national average), and the WW/L class in Northland (~2 times higher); NH4 concentrations in the CW/H/HS/N class in Wellington Region (~50% higher); clarity in the CW/L class in Southland and the WW/L class in Auckland (~50% lower).
Water quality in low-elevation source-of-flow classes in Canterbury, Southland and Waikato Regions generally failed to meet recommended guidelines; median E. coli concentrations in all low-elevation classes in each region exceeded the guideline value, and median DRP, NOX, and NH4 concentrations exceeded the guidelines in all low-elevation classes but one.
Inter-regional comparisons were limited by the number of classes common to two or all three of the regions. Comparisons at the land-cover level were restricted to Canterbury or Southland. However, comparisons could be made among all three regions at the source-of-flow level, and between Canterbury and Southland at the land-cover level. Results of the comparisons varied strongly across parameters and classes. Median NOX concentrations in the CD/L class were higher in Canterbury than in Southland, but higher in Southland than in Canterbury in the CW/L class. Median clarity in the CW/H class was higher in Southland than in Waikato, but the reverse was true in the CW/L class. In the CW/H class, median DRP concentrations were higher in Waikato than in Southland, and median NH4 concentrations were higher in Canterbury than in Waikato.
Due to low site replication, comparisons of temporal trends in the Canterbury, Southland and Waikato Regions could only be made at the source-of-flow level. Comparisons between Waikato and Southland within the CW/H class indicated significant differences in flow (negative trend in Waikato, none in Southland), temperature (greater positive trend in Southland), and in DRP and clarity (positive trend for both in Waikato, none in Southland). Comparisons between Canterbury and Southland within the CD/L class indicated a significant difference in temperature (positive trend in Southland, none in Canterbury).
Stream order effects within REC classes
Significant differences among stream orders within REC classes were rare at all REC levels. The hierarchy of spatial scales that characterises the REC led us to predict that differences among stream orders would be more frequent at the land-cover level than at higher levels, but this prediction was not supported.
Comparisons of temporal trends across stream orders were limited to the source-of-flow level. Trends in DRP concentration varied significantly among stream orders in the CW/H class, and trends in clarity slopes varied significantly among stream orders in the WW/L class. The magnitude of positive trends in clarity in the WW/L class increased with stream order, as indicated by linear regression. This pattern suggests that large rivers in the WW/L class are improving in clarity faster than smaller streams.
