Quick Dips

Control those Weeds, Generate that Power

Rohan Wells, NIWA

The oxygen weed, Elodea canadensis is widely distributed throughout New Zealand. Introduced to enhance habitat for trout, it has displaced native aquatic plants that also provided good habitat. It has now reached nuisance proportions in many water bodies and adversely affects many users, including electricity generators.

In 1994, E. canadensis clogged the Rangitaiki canal to such an extent that only 6.5 cubic metres of water per second could pass through it, compared to a design capacity of 20 m³/s. This represented a major loss in generating capacity for the then Rotorua Electricity (now Trustpower) which used the canal to convey water from the Rangitaiki River to the Whaeo Hydroelectric Power Station. Numerous weed clearing options had been tried, including flushing (ineffective), mechanical cutting (too expensive and operationally difficult), and draglining (ecologically damaging).

NIWA aquatic weeds specialists recommended a controlled trial using the herbicide diquat. Apart from being the only herbicide with New Zealand registration for use in waters, diquat is highly effective against oxygen weeds while having little effect on a range of native plant species. From an environmental point of view, it has the advantage of being short-lived and non-toxic to fish and most other aquatic life if used correctly.

Despite these advantages, the Bay of Plenty Regional Council was reluctant to approve the use of diquat in flowing water because its effectiveness and the predicted lack of adverse effects were much less certain. In addition, the Eastern Fish and Game Council asked that a three-metre wide margin of weed on each bank be left untreated to provide cover for trout.

A resource consent required the use of diquat under strictly controlled conditions. NIWA scientists found that 20 litres of aqueous diquat applied to two 100-metre sections of the canal was sufficient to control E. canadensis throughout the 4.7 km long canal. The margin of weed was achieved by lowering the level of the canal by about one metre during diquat application and then reflooding upon completion. Insufficient herbicide remained in the water to affect the untreated E. canadensis margins once water levels were raised.

After the diquat application, the flow rates in the canal could be increased to 17 m³/s. Subsequent monitoring showed that flushing the canal for two three-hour periods per day at maximum flow was sufficient to erode the devegetated sediment from the central portion of the canal (about 10 metres in width) which would otherwise have been available for rapid recolonisation. Subsequent trials have shown that annual applications of diquat are sufficient to control E. canadensis biomass to levels where it does not reduce electricity generation.

Monitoring of trout stocks by Eastern Fish and Game showed that numbers of all size classes of trout were considerably greater than when the weeds were controlled by draglining. No other impacts on aquatic life were noted other than those associated with the weed removal.

The trial demonstrated that well controlled diquat application is a cost-effective means of controlling oxygen weeds in canals. Trustpower recovered the costs of the diquat application, including the monitoring, in just eight hours of increased generation! This economic benefit, combined with the lack of any adverse environmental effects, makes this truly a win-win situation.

Rohan Wells works for NIWA in Hamilton