Publications: Research reports and publications

Cobb power scheme 2001 resource consent renewal: Takaka River drying

1 May, 2001
Cawthron Report 635. Prepared for Natural Gas Corporation.


Due to the unique geology of the area, some parts of the Takaka River regularly dry up. This is a natural phenomenon and has probably occurred for thousands of years. This report provides information on the section of the Takaka River that dries up. In particular, we describe changes in water level and temperature in 2 pools through the drying zone over the summer of 2000/2001. We also summarise the results of fish surveys through the drying zone and develop a predictive model relating river flows, groundwater levels and power station discharges to river drying. This enables us to determine the potential effects of the Cobb Power Scheme on the ecology of the river.

Over the summer of 2000/2001 the entire drying zone dried out. In the upper reach of the drying zone drying is only dependent on low river flows. In the lower reach drying is dependent on both low flow and groundwater level. The threshold for drying in the upper reach is about 3 500 l/s, whereas in the lower reach it depends on groundwater level. When groundwater levels are high river drying will occur when flows drop below about 7 000 l/s. However, if groundwater levels are particularly low then flows as high as 15 000 – 20 000 l/s are required to maintain surface flow. Given these thresholds, the Cobb Power Scheme is physically unable to stop parts of the Takaka River drying up in dry years.

Once the river dries up remnant pools are left behind in some reaches and these may provide refuges for fish. In the upper reaches of the drying zone, these remnant pools drain completely within just a few hours making them poor refuges. However, in the lower reaches of the drying zone, some deep pools are connected with groundwater and may support fish and aquatic invertebrates for weeks to months after the river has stopped flowing. Connection with groundwater maintains relatively cool water temperatures, and diffusion of oxygen through the surface of the pool and production of oxygen by algae appears to provide sufficient oxygen for fish survival.

Brown trout, longfin eels, torrentfish, upland bullies, inanga and koura were observed during drift dive surveys of the drying zone over the summer of 2000/2001. We estimate that between 7 – 105 adult trout were present in the drying zone prior to drying. It was not possible to estimate the number of native fish or juvenile trout present in the reach prior to drying. The fate of these fish once the river dried out is unclear. Some fish were stranded in remnant pools and were either rescued by locals or died. Others probably moved out of the drying zone themselves in response to decreasing flows and/or temporary periods of no flow.

Overall, storage of water in the Cobb Reservoir during periods of high flow and release of water during periods of low flow will reduce the total amount of time that the river will be dry. For example, over the period from 1 January – 5 May 2001 outflows from the power station were just over twice the inflows to the Cobb Reservoir. The Cobb Power Scheme also influences river drying when flow fluctuations from the power station are close to the thresholds for river drying mentioned above. When this is the case the river will stop flowing and then resume flow in response to changes in power generation. If the Power Scheme were not present we would expect just one 'drying event' between natural freshes in the river. From a fisheries perspective the river only has to dry up once for fish kills to occur. Multiple 'drying events' caused by the power station may act as strong cues for fish to move out of the drying zone if they can use remnant pools as refuges during the actual drying events. However, multiple drying may also result in multiple fish kills if fish are being repeatedly drawn into the drying zone during periods when the river is flowing and then stranded when the river stops flowing. This is not considered to be a major problem since periods of power scheme induced flow generally last for only a few hours and thus may not be long enough to encourage fish to move down into the drying zone.

The summer of 2000/2001 was a particularly dry one, so the patterns of river drying observed this year were particularly severe. Flow duration analyses indicate that the thresholds for drying will normally be exceeded for 75% and 85-90% of the time for the lower and upper reaches of the drying zone, respectively. Therefore, during wet years drying is likely to be restricted to the lower reaches of the drying zone. During wet years, groundwater levels are likely to remain high enough to maintain remnant pools in the lower reaches of the drying zone through periods when the river stops flowing and these pools will be able to act as more permanent refuges for fish and invertebrates.