Executive summary

There is a widespread understanding that water chemistry alone cannot fully capture river ecosystem health, highlighting the need for indicators such as dissolved oxygen (DO) and ecosystem metabolism (EM) – the combination of gross primary production (GPP) and ecosystem respiration (ER). As one of the key attributes in the National Policy Statement for Freshwater Management 2020 (NPS-FM), DO provides essential insights into river health. Auckland Council has continuous high-frequency DO monitoring at 27 rivers, with records extending up to 22 years for some sites.

This study represents the latest analysis of DO and EM data across the Auckland Region. Using R-based code, DO minima and EM rates were assessed across the full dataset, moving beyond prior limitations of short, snapshot assessments. The report also contextualises DO minima within NPS-FM standards and provides a comparable evaluation of EM, even though the indicator currently lacks a formal NPS-FM banding system. The temporal and spatial variability of these metrics is also explored.

The findings highlight substantial variability in DO minima across Auckland Region sites, with reference sites often maintaining higher DO minima levels (above 8 mg/L), and more intensively modified catchments such as urban and high-intensity rural sites experiencing near-anoxic conditions (0 mg/L). Reference sites, such as West Hoe Stream and Cascades (Waitākere) Stream, maintained high attribute bands over the 2020–24 period. In contrast, Rural – High, Urban and Rural – High catchments in the Pukekohe area mostly fell within C and D bands, signalling lower water quality in terms of DO and consequent risks to aquatic life.

Most sites showed relatively stable summer 5-year rolling medians for the 1-day minimum and 7-day mean minimum DO metrics, indicating consistent DO levels over time. There was no strong evidence for trends over time, but this would need to be confirmed with formal trend analyses. 

A close examination of the DO data indicated that tidal patterns may impact DO variability in several sites near the coast, with substantial drops in DO coinciding with high tides. This may be a common phenomenon in the lower reaches of rivers throughout Aotearoa New Zealand. It may also represent a potential stress on aquatic organisms living in these tidal lowland river areas, as well as migratory fish that must pass through them as they swim to and from the ocean to complete their life cycle.

These results highlight stable long-term DO conditions for most sites and emphasise the importance of ongoing monitoring of DO to capture shifts in water quality and changes in ecosystem health.

This report also identified substantial spatial and temporal variability in GPP and ER across the monitoring sites. The reference site (West Hoe), achieved an A-band grade for both GPP and ER, as did Opanuku and Te Muri. Several Urban and Pukekohe sites (Lower Vaughan, Puhinui and Waitangi) were graded as being in the D band for EM.

Between the baseline period (2013–17) and the current period (2020–24), nearly all sites showed statistically significant changes in metabolism rates. However, these differences did not always translate to shifts in band classifications. For those sites that did shift bands, a mix of improvements and degradations was recorded. 

Variability in GPP and ER across sites further indicates ecosystem health. Stable sites such as West Hoe and Rangitopuni indicate healthier conditions, while sites with high metabolic variability (e.g. Waitangi, Kaipara and Puhinui) were often associated with lower attribute bands.

In conclusion, this report highlights the substantial variability in ecosystem health across Auckland Region rivers, driven by differences in land use and specific environmental factors such as tidal patterns. DO minima, GPP and ER collectively reveal that reference sites are in good health. In contrast, urban and high-intensity rural sites frequently showed lower DO minima and higher variability in metabolic rates. Changes in metabolic rates between the baseline and current periods, while often statistically significant, did not always correspond to shifts in attribute bands, highlighting the need for nuanced long-term analysis to truly understand ecological trends. These findings reinforce the importance of tailored monitoring efforts, especially for vulnerable and degraded sites, to further understand changes in DO, EM and ecological health over time.

Cawthron for Auckland Council, January 2025