Ecological evaluation of proposed flow control structure at Chowilla significant ecological asset

"June 2007".

Project Number: Ecological Evaluation of proposed flow control structure at Chowilla significant ecological asset - AW227 M/BUS/227.

MDFRC item.

Associated reports see (Risk Assessment of Cyanobacteria and Blackwater events in Chowilla Floodplain) and (Comments on the Ecological case for a flow regulator on Chowilla Creek, SA).

Construction and operation of a flow regulator at the lower end of Chowilla creek has been proposed as a large scale flow management and water delivery option to the Chowilla floodplain. It is anticipated that this option will allow extension of the current benefits from maintaining the condition of fragmented landscape patches to maintenance of ecological function across the floodplain ecosystem. The degraded state of Chowilla and the prospect of continued low flows demand a bold approach but one that is planned, implemented, managed and reviewed. However, a number of factors, including cyanobacterial blooms, blackwater events (deoxygenation caused by microbial degradation of natural organic matter), invasion by weeds, reduced lotic or flowing water habitats, interrupted fish passage, decrease in large-bodied native fish populations and increases in carp populations have been identified as possible risks associated with the operation of such a regulator. This project examines two of the risks, cyanobacteria and the occurrence of blackwater, or more specifically de-oxygenation resulting from a blackwater event, more thoroughly in order for the environmental benefits and possible dis-benefits of regulated flooding to be properly considered in order to inform the debate on whether a regulator should be constructed and how it should subsequently be operated. Return of nutrients and carbon from the floodplain to the main river channel is an important part of the function of a healthy low-land river. It is clear from experiments undertaken as part of the current project that the Chowilla floodplain would produce a pulse of nutrients and carbon on flooding. Consequently there is a risk of cyanobacteria occurring in Chowilla floodplain with the implementation of artificial flooding using a flow regulator. This is due to shallow water in lagoons, nutrient release from sediment and decaying vegetation and low water velocity in some locations. There are two scenarios where the development of cyanobacterial blooms in the Chowilla system could represent a potential hazard to public health and/or water supply: 1. cyanobacterial populations that are restricted to the wetlands that become isolated during drawdown 2. populations that remain connected and may contaminate the main river channel with return flows from the Chowilla anabranch These two scenarios represent different risks. Isolated wetlands with high cyanobacterial abundance have little impact on the main river channel, although these blooms may be locally significant. On the other hand if the wetlands hosting cyanobacteria drain into the main channel this may be a significant source of toxins or taste and odour compounds. The hazards associated with cyanobacteria range from public health related toxicity issue to aesthetic water taste and odour issues. The toxins produced by cyanobacteria include both hepatotoxins (liver damaging) and neurotoxins (nerve damaging). The taste and odour compounds produced by cyanobacteria are geosmin and MIB. These compounds are difficult to remove with conventional water treatment and require expensive activated carbon for adequate removal. Consequently it is important to minimise cyanobacterial biomass in the River to reduce the risk from toxins and the taste and odour compounds. It will be important that flooding of the floodplain is timed to coincide with periods of above entitlement flow in the main river channel to avoid supplying nutrients to the river at times of thermal stratification. A combination of lower than normal flow in the river and higher than average temperatures would provide the worst case scenario for persistent stratification. Under these low flows it would be unwise to operate the regulator. Modelling indicates that at flows in the River Murray of 4,000MLday-1, the river would stratify at low wind speeds. Consequently although the minimum proposed flow into the anabranch of 4,000 MLday-1 with a flow-through of 2,000 MLday-1 could be achieved at entitlement flow (QSA = 7,000 MLday-1) this would only maintain a flow in the River Murray Channel downstream of the Regulator of 5,000 MLday-1. This flow would prevent minimal capacity for dilution, increase the potential for stratification and subsequently increase the risk of generating a problematic algal bloom in the River Murray. If the flow regulator is only operated at the higher end of the planned flow band than the risk of the cyanobacteria posing a major problem in the main river channel is significantly reduced. With the exception of the areas dominated by redgum, the concentration of leaf litter on the floodplain at Chowilla is quite low and reflects the generally poor state of the vegetation in the region, which inturn reflects on both the lack of flooding to ensure soil-moisture and grazing of the floodplain. Based on the results of mesocosm experiments undertaken as part of this project, there does not appear to be a substantial risk of a significant blackwater (deoxygenation) event in those areas that are either deep (>1m) flooded or have good rates of water exchange. The risk of blackwater increases substantially in shallow flooded areas and redgum areas. The results of hydrodynamic modelling show that 3325ha of floodplain is inundated by less than 1m, with 1844ha inundated by less than 0.5m. Estimates of oxygen demand in the respective WMU’s indicate that the critical minimum depth of inundation to prevent establishment of anoxia in wetland, lignum shrubland and grassland areas is 0.7m, 1.0m in the blackbox woodland areas, and 4.5m in the redgum forest areas. Previous watering projects on the Chowilla Floodplain demonstrate that anoxic conditions develop in shallow flooded areas when the water is ponded on the floodplain. These results highlight the importance of maintaining flow and water exchange in order to maintain water quality. Nutrient and carbon pulses from the floodplain would have occurred under natural conditions and were probably important for supplying energy and nutrients to the river. Furthermore, because of the higher frequency of floods, the vegetation would have been in substantially better condition than it currently is which would have been reflected in higher levels of litter standing stock, and hence a higher likelihood of supplying nutrients to sustain algal blooms and carbon to create blackwater events. Decreasing the frequency of flooding of the Chowilla floodplains has no doubt significantly harmed the ecological condition of the floodplain itself but also the River Murray. The risk of algal blooms and blackwater events appears to be tightly coupled to the ability to maintain flow (both volume and velocity) within both the inundated area and the River Murray channel. Consequently the risk of algal blooms or blackwater events can be managed. However, to minimise the potential for adverse effects it is strongly recommended that in addition to maintenance of flow, the operation of the regulator be done in an adaptive management framework which includes a robust monitoring program, that not just determines the change of condition of the floodplain, particularly the floodplain vegetation, but also assess the impacts and benefits of the manufactured flood to floodplain soil condition, groundwater and receiving water (Murray River).