Overall the results show that many of Northland's rivers have poor water quality on occasion and some consistently have poor water quality. Water quality is related to surrounding land use. Sites in catchments dominated by native forest have the best water quality including the Waipoua, Waipapa and Victoria River sites, followed closely by those with catchments dominated by exotic forestry such as Mangahahuru Stream at Main Road. The worst water quality is at sites in intensive pastoral farming and urban environments. The worst being Mangere Stream, followed by Awanui upstream of Waihue Channel, Wairua at Purua and Mangahahuru at Apotu Road. These latter sites typically have low clarity and high nutrient and bacterial levels.

This relationship between water quality and land use is consistent with findings from other regions (Quinn et al. 1997, Quinn and Stroud 2002) and national results (Larned et al. 2004, Larned at al. 2005 and Scarsbrook 2006).

Many sites have dissolved oxygen levels outside the ideal range. This appears to be limited to a few areas of New Zealand. Scarsbrook (2006) in an analysis of data from 77 river sites throughout New Zealand found that sites in Northland, Auckland and on the lower Tarawera River had the lowest dissolved oxygen levels.

As it was noted in previous Annual Environmental Monitoring Reports (NRC 2005 and 2006) that the extreme high and low dissolved oxygen levels were of concern, further investigation was carried out in 2007. This showed large daily fluctuations in dissolved oxygen at unshaded sites with high plant biomass, with significant drops in DO at night while the plants respire. Large diurnal patterns for DO in unshaded streams with high plant biomass is consistent with the findings of Wilcock et al. (2006a and 2006b). This suggests that in some situations riparian fencing and planting will help with dissolved oxygen levels as well as reducing the inputs of faecal pathogens, where riparian vegetation creates enough shade over the stream to reduce photosynthetic respiration.

Overall the majority of sites occasionally had water quality outside the trigger values for the protection of aquatic ecosystems (ANZECC 2000) and exceeding the recreational bathing guidelines (MFE 2002). The exceptions are the relatively pristine sites in native or exotic forest catchments, which are often in compliance, including Waipoua River, Waipapa River in Puketi Forest and Mangahahuru Stream at Main Road. The following sites consistently had water quality outside the guidelines: Mangere Stream, Wairua River at Purua and Mangahahuru at Apotu Road.

Again this is consistent with the rest of New Zealand. Larned et al. (2005) in an analysis of all national and regional river water quality data for New Zealand found that most sites in natural land cover (e.g. native forest or tussock land) were within guideline values except for ammoniacal nitrogen. They found very few urban sites met the water quality guidelines, while over 50% of the pastoral sites failed to meet all the guidelines.

Overall there were several positive changes in water quality detected in the trend analysis for 1996 to 2006. There were decreasing trends for dissolved reactive and total phosphorus, oxides of nitrogen (NNN) and ammoniacal nitrogen at the Mangere Stream site. Several other sites exhibited decreasing trends in several nutrient parameters, including Victoria River, Awanui River above Waihue channel outflow, Waipapa River in Puketi Forest, Whakapara River near SH1, Mangahahuru Stream at Apotu Road, Mangakahia at Titoki and Manganui River.

The decreasing trends in ammoniacal nitrogen at several sites is a good indication of improvements in point source discharges, which is consistent with the national picture (Scarsbrook 2006) and research into the impacts of dairy farming (Wilcock et al. 2006).

There were positive trends in water clarity at three sites that had a corresponding positive trend for turbidity: Waipapa River, Mangahahuru at Apotu Road and Mangere Stream. Similarly to the positive trends for ammoniacal nitrogen, this suggests improvements in point source discharges upstream of these sites.

It is encouraging that overall many of these positive trends were found in the rivers ranked as having the worst water quality (table 9 in appendix D). However some of these rivers also showed negative trends in other parameters.

Overall there were much fewer negative trends detected in water quality. The most noteworthy are the increasing trends in: dissolved oxygen detected at five sites, conductivity at three sites, pH at nine sites, E. coli levels in Mangere Stream and dissolved reactive and total phosphorus in Waipoua River. The large magnitude of the increasing trend in E. coli, DRP and TP is of particular concern. All of these negative trends will be investigated further.

In conclusion, many of Northland's rivers have severely deteriorated water quality, particularly in pastoral farming catchments, while in contrast water quality is best in native forest catchments. These findings are consistent with elsewhere in New Zealand. However many of these degraded Northland Rivers have shown some improvements in water quality in the last 10 years. These improvements are most likely as a result of improvements in point source discharges. As a region the focus now needs to shift to reducing the impacts of diffuse surface runoff on water quality.