Assessment of ecological risk associated with irrigation in the Fitzroy basin: Phase 2

This project in the Fitzroy catchment in Queensland was one of three case studies used to illustrate how ERA may be employed in the management of irrigation areas in Australia. In the problem formulation stage of the ERA process community stakeholders identified issues and helped to prioritise research needs for the second phase. These were that the research should focus on the irrigation area scale and the impact of the decline of water quality on macroinvertebrate populations with emphasis on determining the relative contribution of various environmental factors to changes in these populations.

As part of the ERA analysis phase, an initial pilot study was successfully undertaken in 2001/2002 and based on this some changes to the original design were required. The study addressed four key questions in relation to a drain that delivers irrigation runoff to the Dawson river via Gap Creek. Findings of the risk assessment phase were that the magnitude of the effect of water from the drain on macroinvertebrates in the river was too small to be detected, based on the end points used. (Further study at the species level of identification may be warranted.) However, effects (loss of sensitive taxa) were recorded at a site on Gap Creek, where risks from endosulfan exposure were determined to be high from December to February. Statistical analyses suggested that the most important factors explaining differences in the macroinvertebrate structure were discharge variables and levels of the pesticide endosulfan sulphate. Discharge variables that were identified as being of significance are discharge divided by time since that discharge, maximum discharge and discharge > 200 ML/day. Based on a lack of significance of effects on macroinvertebrates at a site 3 km downstream from the impacted site on Gap Creek, it was determined that the risks of runoff from irrigated land on macroinvertebrates were highly localised. Rate of recovery from effects was difficult to assess given the single impacted site where measurable effects were recorded. It is postulated that measured pesticide contamination of stream sediments may hinder such recovery and this may be a useful avenue of future research.