Spatial and Temporal Patterns in the Condition of Waikato Streams Based on the Regional Ecological Monitoring of Streams (REMS) Programme

Report: TR 2010/04

Author: Kevin Collier, Mark Hamer

Abstract

Waikato Regional Council has been carrying out annual assessments of invertebrate community composition in streams and rivers since 1994 as part of the Regional Ecological Monitoring of Streams (REMS) programme. These sites include wadeable high-gradient streams with stony beds, low-gradient wadeable streams dominated by soft sediments, and some larger non-wadeable streams with long term records that have been retained while appropriate sampling protocols are developed. From 2005, sampling has included a network of 23 wadeable ‘reference sites’ in undeveloped catchments to provide a baseline against which to measure change, and a range of sites representing low, moderate and high levels of pastoral land-cover (‘land-cover sites’). Sampling at 46 ‘long-term sites’, including three reference sites, has been conducted for more than 10 years using consistent protocols that have enabled assessment of temporal trends in ecological condition at these sites. Condition is assessed using four stream macroinvertebrate-based measures derived from 200+ counts of individuals: number of different types of mayflies, stoneflies and caddisflies (excluding algal-piercing Hydroptilidae)—EPT* richness; the percent abundance of these sensitive insects—%EPT*; a measure of tolerance to organic pollution—the Macroinvertebrate Community Index or MCI; and an integrative score of all three metrics—Average Score Per Metric or ASPM. Metrics were also calculated reflecting (i) habitat quality based on qualitative assessments of nine riparian, bank and channel attributes, and (ii) instream plant cover and proliferation.

Of the 43 non-reference ‘long-term sites’ monitored for over 10 years, 15 (35%) showed trends over time using two methods of analysis. No trends were detected at the remainder of sites which can be considered ‘stable’ in terms of macroinvertebrate community indicators. The different metrics showed variable responses over time, presumably reflecting differential sensitivity to various stressors or enhancements. Three sites showed increasing relative abundance of sensitive species (%EPT*) at an average rate of +7% per year (Relative Kendall Sen Estimator—RKSE), while this metric declined at one site by -9% per year. All trends in MCI were negative (nine sites; mean RKSE -2% per year), whereas the ASPM trends indicated three sites declining in condition (mean RKSE -3% per year) and five sites improving in condition (mean RKSE +5% per year).

There was a clear and consistent pattern over multiple years (2005–08) showing that macroinvertebrate metrics were lower where there were higher levels of catchment development (primarily agricultural). Thus, ‘reference sites’ in undeveloped catchments were in better ecological condition than those in developed catchments, where condition declined between levels of moderate (10–49% of upstream catchment modified), high (50-89%) and very high (>90%) development. Declines in condition occurred rapidly with the onset of catchment development, and were most pronounced where more than 40% of the catchment was developed.

A separate analysis of data from 2005/06, when streams in Hamilton City were also sampled, indicated that on average urban streams had lower ecological condition than streams in other developed catchments, although a few urban streams did support invertebrate communities with high numbers of sensitive species. Towns and cities represent only a small area of land-cover in the Waikato region (around 1%), compared to pasture (58%) for example, so the net effect of urbanisation on regional stream health is much less than for agricultural development.

Metrics reflecting instream plant cover and proliferation were highly variable over time, but also indicated a significant response to increasing levels of development upstream, most noticeably for macrophytes. Habitat quality scores declined significantly across land-cover classes, and this was most evident between undeveloped to moderately developed classes, and highly developed to very highly developed classes. Habitat quality scores were significantly related to all macroinvertebrate metrics and accounted for 49–57% of their variation, suggesting that ecological condition as reflected by macroinvertebrate communities is a function of habitat quality as well as water quality.

Spatial and Temporal Patterns in the Condition of Waikato Streams Based on the Regional Ecological Monitoring of Streams (REMS) Programme [PDF, 3.6 MB]