In-stream habitat assessment
How much water do trout need in small streams?
Roger Young of the Cawthron Institute and Lawson Davey, Fish and Game Officer for the Nelson/Marlborough Region
(Note: this article first appeared in Fish & Game, 2002)
In recent years Fish and Game has increasingly become concerned with the number of applications to take water from small streams, and their effects on trout habitat and the numbers of trout small streams can support. Unfortunately instream values are often perceived to be relatively low in small streams, despite the fact that these streams provide important spawning and rearing habitat for sports fish as well as a variety of native fish species. Present methods used to guide decisions on flow management in large rivers, such as the Instream Flow Incremental Methodology (IFIM) are time consuming and expensive and often cannot be justified on a small stream due to the cost. As a result, decisions in the past have often been based on "educated guesses" as to what the effects of the abstraction will be on the stream and its aquatic life.
Cawthron Institute scientists have been commissioned by Fish and Game NZ, Tasman District Council and the Motueka River Integrated Catchment Management Programme to test some quick and cheaper methods of assessing flow requirements of fish in small streams to guide decisions on flow management.
Two reaches on the Rainy River (an important brown trout spawning and rearing stream of the Motueka River) were chosen and used to compare results from a quick hydraulic method, developed for larger rivers by NIWA in conjunction with the Auckland Regional Council, against a more sophisticated habitat-based method for relating flows with habitat availability. The main focus of the study was on how ecological values change with water quantity, however water quality and flow variability are also recognised as important components needed to maintain the ecological values in small streams.
Given trout are an "indicator species" and have higher minimum flow requirements than any of the native species present, it was decided that in this case the primary instream management objective was to maintain sufficient habitat for yearling brown trout, which by default would also protect native fish and invertebrate habitat. Yearling trout are usually found in run habitats, and therefore the depth and velocity of runs were considered to be critical parameters controlling the quality of yearling trout habitat. These areas are often the most affected during low flows.
Fieldwork was carried out in two reaches of the Rainy River in April and May 2002, with depths, widths and velocities taken at three contrasting flows. Electric fishing confirmed juvenile brown trout habitat use in the Rainy River was comparable to that predicted by habitat suitability curves developed overseas.
Preliminary results reveal the two techniques predicted similar mean depths (although the quick hydraulic method prediction diverged at flows approaching zero), mean widths and mean velocities over a range of flows. Since the quick hydraulic method predicts only the response of average depth, velocity and width with flow, rather than the distribution of depths and velocities, it is not possible to directly relate the results with habitat suitability for particular species present. Generally juvenile trout favoured the deeper areas within the runs, but were found over a range of velocities in roughly the same proportion as they occurred, indicating that for this river, water depth is likely to be the limiting factor during low flows.
Cawthron Scientist Dr Roger Young says, "The research indicates that in the Rainy River juvenile habitat is well below optimal levels at the natural mean annual low flow (MALF) and it declines almost linearly at flows below the MALF. Although the quick hydraulic method is not quite as accurate as the more expensive and time-consuming IFIM habitat-based method, it is in the "ballpark". It has potential to enable a better understanding of the effect a water take (or for that matter a drought) will have on habitat availability, and therefore better decisions to be made in relation to water management in small streams."
In the past, due to an absence of better information, minimum flows have often been based on historic flow statistics, for example either a 1-in-5 year low flow or a 1-in-10 year low flow. This assumes the natural flows are suitable for the species present. In the Rainy, the IFIM analysis shows that a 1-in-5 year low flow (92 litres/second) and a 1-in-10 year low flow (56 litres/second) would equate to a 15% and 37% reduction respectively in yearling trout habitat availability compared with that at the MALF of 123 litres/second. Taking, say, 50 litres/second for irrigation, would seriously reduce the habitat available and turn a one in five year drought into a worse than one in 10 year drought.
The technical methods tested in the study provide information on changes in habitat availability at different flows, but they do not define a minimum flow, or the amount of habitat loss that is acceptable. The critical part of setting minimum flows is balancing the instream and out-of-stream water demands and deciding what reductions in habitat availability compared to the MALF, if any, are acceptable. For example, if a 10% reduction in yearling trout habitat availability compared to the MALF was taken as an acceptable threshold, the IFIM method recommends a minimum flow of 105 litres/second, compared to 97 litres/second recommended by the quick hydraulic method. The similarity between these recommendations, being less than the margin for error in estimating flows, gives some confidence that either method could be used for the Rainy River.
Ultimately, the question of minimum flow setting is a matter for decision makers and, where necessary, the Courts. Fish and Game New Zealand can use this information to advocate that enough water will be left in the stream to provide for recreational interests. Techniques such as this will be another tool to assist.
Report can be downloaded. 
536 KB