Publications: Research reports and publications

National rapid habitat assessment protocol development for streams and rivers

30 January, 2015
Cawthron Report 2649. Prepared for Northland Regional Council.


This report aims to advance the development of a standardised national rapid habitat assessment protocol for rivers and streams. It summarises the analysis of trial data collected by Regional Council and Department of Conservation staff during routine monitoring in 2013 / 2014. The report was funded by Envirolink medium advice grant 1519.

A draft protocol containing nine parameters was trialled at 560 sites throughout New Zealand. Correlation analysis showed strong relationships between some parameter scores and measured visual or modelled estimates of stream habitat characteristics. For example, shade scores were validated by site measures of shade (rs = 0.84, n = 64), and invertebrate habitat scores were validated by invertebrate community metrics (MCI: rs = 0.52, n = 494). There was insufficient data to validate some parameters. Total scores were strongly related to catchment-scale measures of percent native vegetation cover (rs = 0.46, n =553), percent impervious cover (rs = -0.35, n = 553) and land-use intensification (rs = -0.52, n = 553). Total scores were also related to descriptors of environmental variability supporting the use of a comparison to reference approach for reporting final scores as a percentage of reference scores. Overall, correlation results suggest that the draft protocol likely includes a good range of parameters that together provide a representative assessment of stream habitat quality. Inter-user variability was investigated at 17 sites and results showed general consistency among users but highlighted that some parameter assessments were subject to high variability. Total scores were on average within 15% of each other.

User feedback was incorporated into an amendment of the draft protocol. Specifically, feedback on the scoring and wording of habitat parameters was considered and these changes to the protocol resulted.

1. Focus on numerical assessments of parameters to inform parameter scores, which minimises subjectivity.

2. Change the range of scores from 1–20 to 1–10. User feedback and literature review suggests that seven (or more) categories are sufficient.

3. Reinstate invertebrate and fish abundance and diversity as distinct parameters to allow the separate assessment of components. This approach was suggested by the initial working group and supported by data analysis.

4. Exclude human effects, such as channel alteration or the presence of fencing to ensure the stream habitat quality score can be robustly related to causes, rather than incorporating them.

The resulting protocol provides a 'habitat quality score' (HQS) and the future development of a separate habitat modification assessment is recommended, as is the case with comparable river assessment protocols overseas.

The recommended HQS is informed by the following 10 parameters scored 1–10. The total maximum score is 100. However, the total score could be scaled to a reference score to provide a % HQS for reporting.

  •  Deposited sediment
  •  Invertebrate habitat diversity
  •  Invertebrate habitat abundance
  •  Fish cover diversity
  •  Fish cover abundance
  •  Hydraulic heterogeneity
  •  Bank erosion
  •  Bank vegetation
  •  Riparian width
  •  Riparian shade