Allocating limited water: linking ecology and economics

Abstract

Available water is limited in quantity with many potentially competing uses, including irrigation, environmental amenity and domestic supply. A question is how to allocate a limited water quantity to different uses for society. Economically we allocate water to equate marginal social benefits across different uses. For environmental amenity this framework relies on ecological response functions and prices. The economic framework uses prices to represent social welfare - representing social willingness to pay for extra ecological amenity. Valuation (pricing) of ecological improvements involves non-market valuation; estimates are available for ecological amenity improvements. An economic decision framework for environmental water requires flow regimes (the decision variable) on the x-axis and ecological responses or endpoints (e.g. for Golden Perch in the Goulburn River) on the y-axis. This analysis characterises the ecological response functions relatively simply with several steps to translate existing ecological response curves into functions suitable for economic analysis. First was harmonisation of x-axes to total volume of water. The overall response of a complex organism like Golden Perch to flow regimes consists of separate sub-responses to different components of the flow regime - spawning from a spring high flow (fresh); survival of larvae from provision of shallow, slow flowing habitat over summer; and adult condition from provision of deep pools through autumn and winter. Each of these sub-responses was expressed graphically with ecological response on the Y axis, and total volume of environmental water used to achieve the response on the X axis. Having common units on the X axes allowed the three sub-responses to be combined for the economic analysis. Second was harmonisation of y-axes to estimate population size. Strictly speaking, the sub-components listed above are realized as number of eggs produced, larval survivorship, and adult survivorship, respectively. For the economic analysis, these endpoints were translated to an estimate of impact on total population size - by translating previously held 'traffic light' (i.e. good, moderate, poor) outcome scores to estimates of the Golden Perch population from previous research carried out in the Goulburn-Broken catchment. Third was translation of piece-wise linear functions to smooth equations. We translated those functions to smooth equations, allowing us to characterise the economic model in terms of marginal responses. Fourth was the issue of combining sub-responses. A compound function for Golden Perch production was derived by assuming that the temporal periods for each of the sub-responses do not overlap. Therefore the overall production function is driven by spawning in spring, slow-flow habitat in summer, and pool habitat in autumn and winter. This assumption is untrue for real systems; however, within the bounds of our other uncertainties and simplifications, this assumption is minor. The economic model maximises total social (defined here as environmental plus irrigation) benefits expressed in dollar equivalent terms for water allocation decisions. The total quantity of water available is limited, hence the economic problem for social welfare. Decisions are made for water allocations to environmental assets or endpoints, and agricultural outcomes. The shadow price of water is derived for limited water shared between an environmental and agricultural use.