Broad scale change in land use from annual plant species to perennials has been proposed as the solution to the increasing trends in land and stream salinity. A number of studies have demonstrated that increasing the area of perennial plants will have substantial negative effects on stream flow. Some studies have suggested that the cost of reducing water for irrigation may outweigh the benefits of ameliorating dryland salinity. There is some evidence that targeting management to areas of high groundwater salinity and soils where recharge is highest may have economic merit. The aim of this study was to estimate the benefits to targeting salinity management in the Boorowa River catchment in NSW, assuming no changes in farm profit from changing land use. The results showed that the net external benefits of salinity management are negative except where management is targeted to areas with the highest groundwater salinity, mainly due to the cost of reduced stream flow. Also for many management scenarios, reducing recharge increased stream salinity. These costs are outweighed however, by only a small increase in profit from that might result from the adoption of a new technology such as new perennial plant species. It is concluded that the economic welfare of the catchment will increase only where research is able to develop or identify more profitable perennial plants. In the Boorowa River catchment the provision of subsidies to increase adoption of salinity management practices will only benefit landholders, as there is insufficient understanding of catchment characteristics to ensure management is targeted to ensure benefits accrue downstream. The study also highlights the need to consider the impact of salinity management on a range of natural resource objectives.

Dryland salinity is perceived as a major natural resource problem confronting rural Australia. Increasing trends of land and stream salinity are a threat to major environmental assets and could lead to substantial costs to urban water users and farmers. The Australian Dryland Salinity Assessment (National Land and Water Resources Audit, 2001) estimates that just over 5.5 million hectares of land throughout Australia have a high potential to become saline and this is expected to increase to 17 million hectares by 2050.

The natural landscape was in a hydrological equilibrium with the prevailing climatic cycles prior to European settlement. Essentially, recharge and discharge were balanced so that groundwater discharged into streams at the same rate that rainfall drained into the water table. Clearing of native perennial vegetation and replacing it with annual species has resulted in lower rates of evapotranspiration, leaving more rainfall to enter streams as surface run-off or via increased recharge through the water table discharging as base flow (Resource Analysis Unit Central West Region, 2001).

Increased deep drainage (water that passes below the root zone) has caused disequilibrium in the water balance in the short term where recharge of groundwater occurs at a greater rate than discharge, leading to increases in aquifer storage and hence rising water tables. Salt stored in the soil profile over the millennia is mobilised by the additional water and enters streams directly, as base flow. Saline water is discharged to the land surface where the capacity of a catchment to discharge water as base flow is exceeded, leading to land salinisation (Resource Analysis Unit Central West Region, 2001).

Changes in evapotranspiration will affect the volume of surface run-off immediately. There may be long delays however, before changes in deep drainage are reflected in a change in the rate of base flow and land discharge. This is due to the time taken for the groundwater system to come to equilibrium with the new water fluxes. While some catchments may respond to changes in the water balance within a year, others may take millennia. The responsiveness of a catchment depends critically on their geological and hydrological characteristics (Coram et al., 2000).

The focus of salinity management throughout Australia has been to reduce deep drainage, lowering the water table and hence the volume of land discharge, ameliorating land salinity. Salt loads in streams will also decrease as the rate of base flow from saline groundwater declines. However, the salt concentration of streams may rise initially due to the lags in the response of base flow to changes in deep drainage and the immediate effect of a change on the less saline surface run-off.

Changes in salt loads and stream flow affect the environment and water users downstream that are reliant on rivers for domestic and industrial consumption and for primary production. Urban users are most affected by changes in the concentration of salt in the river while the environment and irrigators are affected by changes in both the availability of water and salt concentration.

Economic impacts of in-stream changes can be substantial. One study estimates the cost of increasing stream salinity in the Murray Darling Basin to $142,000 per annum for each one unit increase in electrical conductivity (μs/cm) at Morgan (Gutteridge Haskins and Davey, 1999). In addition, sufficient concern has been raised about the environmental impacts of low stream flow that water management plans in NSW have recently decreased the allocation of water to irrigators in certain circumstances for environmental purposes. The implication of this is that decreasing deep drainage will have environmental and economic costs.

A number of economic studies have been undertaken to determine the benefits of decreasing deep drainage in NSW. Read (2000) and Wilson (1994) estimated downstream benefits of reducing recharge in two different catchments in NSW. Read (2000) showed the benefits were small in relation to the cost of changing land use in the Billabong Creek catchment in the Murrumbidgee region. Wilson (1994) found that there were off-site benefits of reducing deep drainage but these were not substantial. Greiner (1999) found that the off-site benefits of reducing recharge in the Namoi River catchment were sufficient to cover costs of amelioration of land salinity. This study did not examine the affect of reducing recharge on river salinity or stream yield.

Heaney et al (2000) showed that the placement of perennials in the landscape had a major influence on the magnitude of benefits to down stream water users of decreasing deep drainage. They also found that targeting reforestation to soils with certain hydrological characteristics could result in net economic benefits to the Macquarie River catchment. If perennials were inappropriately targeted the costs to irrigators of reduced stream flow could be greater than the benefits of a reduction in the area of land salinity.

The affect of perennial plants on land and river salinity and stream flow depend critically on landscape characteristics. These characteristics vary considerably across the landscape and there is a paucity of locally relevant data that describes this spatial variability. There is, therefore, a high risk that management aimed at decreasing deep drainage will be targeted inappropriately, with current levels of knowledge. Alternatively, delaying the adoption of salinity management practices could lead to higher costs through rising stream salinity and the spread of land salinity.

The lack of data is of particular relevance to the catchment authorities that have been established in most states to set priorities and administer funds provided by state and federal governments under the National Action Plan for Salinity and Water Quality (NAPSWQ) and Natural Heritage Trust (NHT). The majority of these funds will be directed to on-ground management through the provision of direct payments to land holders.

The aim of this paper is to estimate the economic benefits (or costs) of reducing deep drainage in the Boorowa Catchment of NSW in different parts of the landscape that differ in the mix of soil types and the salinity of groundwater, to determine the value of targeted management compared to a more indiscriminate approach that is not based on sound data.

The paper has four sections. The first of these describes the Boorowa Catchment and the second outlines the economic model and the underlying hydrological model. The third details the results of the analysis and the implications for salinity policy and research and development is discussed. Conclusions are presented in the final section.

Study Area

The Boorowa River catchment is situated in the Central West of NSW. It is a sub catchment in the eastern headwaters of the Lachlan River. The Boorowa River runs from south to north and joins the Lachlan River just below the main water storage facility, Wyangala Dam (Fig. 1).

Figure 1: Lachlan River catchment and Boorowa River sub-catchment.

The Boorowa sub-catchment covers an area of approximately 182,000 hectares and has an average annual rainfall of 600-700 mm. Grazing is the dominant activity in the region, with some cropping on the more productive soil types. Over the last 10 years, landcare groups in the catchment have observed an increasing incidence of dryland salinity, rising water tables and soil acidification becoming significant problems (Ivey ATP and Wilson Land Management Services, 2000).

The geological formations throughout the Boorowa catchment are complex and have a significant impact on salinity and salinity management options available to land managers. Soil types throughout the catchment vary considerably ranging from robust, durable soils to very fragile soils, naturally acidic and sodic soils (DLWC 2003).

The Boorowa River contributes a substantial volume of water and salt load to the Lachlan River and is a source of water to communities and irrigators downstream of the confluence with the Lachlan River.  It is also an important source of water for some environmental assets. Changes in land use in the Boorowa River catchment that impinge on water quality and stream flow will affect downstream users and environmental assets on the Lachlan River.

The Lachlan catchment covers an area of 84,700 square kilometres and produces around 14 per cent of the agricultural production in NSW, from a land area of approximately 10 per cent of NSW.  Over 400 irrigators use over 90 per cent of the total water entitlement with the remainder allocated to urban use (Department of Land and Water Conservation, 2003).

Irrigated land is used mainly for perennial, summer forage and winter pastures. In 1997 lucerne was grown on about 22 per cent of irrigated land and winter cereals (wheat, oats, barley) about 15 per cent (Environment Protection Authority, 1997). Cotton production has grown rapidly in recent years in the lower valley. There has also been a shift toward higher value industries such as wine grapes and dairy production, but these are still small relative to other enterprises.

The major towns dependant on the Lachlan River for water supplies are Cowra, Forbes, Condoblin, and Lake Cargelligo.  There are also a number of smaller towns and villages dependent on the river (Beumer, 2001). Increasing salinity (salt concentration) of the river could have a major economic impact on domestic and industrial users of water mainly through accelerating the depreciation of capital items.

The dependence of rural communities and farm businesses on the Lachlan River has markedly decreased flows. The average annual flow at Oxley at the western end of the catchment has been reduced from 234,000 ML/year, which is an estimate of flow under natural conditions, to 120,000 ML/year (Environment Protection Authority, 1997).  This has given rise to concerns about the environmental impact of reduced flows.

Natural features of the Lachlan River catchment that have been identified as being of national importance include Lake Cowal, the Booligal wetlands and the Great Cumbung Swamp. There are a number of other reserves and high quality crown land that also contribute significantly to the resources of the catchment that could also be adversely affected by increasing river salinity (Department of Land and Water Conservation, 2003).

Overview of the model

A spreadsheet model was developed that evaluates the impact of changes in excess water (rainfall minus evapotranspiration) on the average area of salinised land and the average salt load and stream flow at critical points in the Lachlan River. At each of these critical points the costs and/or benefits of changes in stream flow and river salinity to urban, industrial and irrigation users are calculated using damage functions estimated by Wilson (2003). The benefit of decreasing land salinity is estimated by the increase in gross margin resulting from reclaiming previously saline land. The present value of the average annual costs and the benefits are summed over 50 years.

Excess water is assumed to flow directly into streams as surface run-off or indirectly via the watertable, as recharge. The lag in the response of base flow to a change in recharge of the groundwater is represented by a cumulative logistic function after Dawes et al (2004). This approach was used in an economic model of the Murray Darling Basin (Heaney et al 2000).

The catchment was divided into four sub-regions (groundwater flow zones) according to the response times of groundwater flow to a change in deep drainage and the salinity of the groundwater. Groundwater flow zones (GFZ) are an aggregation of a number of similar watersheds or groundwater flow systems. Each GFZ was further divided into soil classes according to soil texture, which is the main factor that affects the proportion of excess water that drains past the root zone (the recharge fraction).

Table 1: Area of soil classes and groundwater salinity (GWS) of ground water flow zones (GFZ) in the Boorowa River catchment.

Excess water can be altered on each land management unit in each GFZ. Stream flow and salt load are affected by the groundwater salinity of the GFZ being targeted and the recharge fraction of the land management unit. In this analysis it is assumed that the volume of excess water can be altered at no cost to the landholder. This implies that salinity amelioration strategies are equally as profitable as current farming practices.

Watersheds

The Boorowa catchment is made up of local scale groundwater flow systems (GFS) generally less than five kilometres in length. It was assumed that local GFS are coincident with surface topography, and hence a Digital Elevation Model (DEM) could be used for the disaggregation. A stream-network was constructed from the DEM by extending the stream-network upstream until a certain stream contributing area (1000 ha) was reached. Sub-catchment areas were then defined using this stream-network. These sub-catchment areas are taken to be representative of individual groundwater flow systems, and form the fundamental unit for which a groundwater time response can be calculated. A more detailed description of the method is provided by Evans et al., (2003).

Hydrological model

The hydrological model describes an equilibrium response of the landscape to hydrologic change and is based on the assumptions that:
• The shape of the water table can be approximated as a function of the topography (consistent with local groundwater flow systems);
• A defined sub-catchment is a discrete GFS that contains all recharge and discharge relevant to that GFS; and
• The equilibrium response time can be derived according to Dawes et al (2004) and Gilfedder et al (2003).

The amount of recharge that occurs within a GFS depends on rainfall, the evapotranspiration (Et) and soil texture. The difference between rainfall and Et is excess water. Excess water is assumed to flow into streams, as either surface run-off or base flow. The partitioning of excess water into these two components depends primarily on the soil texture. On heavy soils a higher proportion of excess water flows into stream as run-off, whereas relatively more flows into streams via base flow on lighter soils.

The change in flow of groundwater directly into streams in response to a change in excess water is modelled using a cumulative logistic function as described in Dawes et al (2004) (Equation 1). This relationship describes the time lag between changes in recharge and changes in base flow;

(1)

Where:     R0 is the equilibrium rate of recharge,
                Ri is the recharge in period i, and
                D(t) is the discharge of groundwater into streams described by Equation )

Where:     t_half is the time taken for 50% of the recharge to enter the stream, and
                t_slope is the steepness of the central portion of the curve.

The underlying assumptions of this model are that:
(i) each annual recharge pulse is independent,
(ii) that the response is linear and additive,
(iii) that recharge from year to year is not correlated and
(iv) discharge response is not hysteretic

For a discussion on the validity of these assumptions see Dawes et al (2001).

Given the paucity of data on groundwater response for the study catchment and the local nature of the groundwater systems, the assumptions implied in the model are considered reasonable. The main strength of the approach is that the estimated responses do not need to be underpinned by more complicated process based models, for which data does not exist.

Groundwater response times (thalf) were determined using the method outlined in Evans et al., (2003) is based on ongoing work to develop a dimensional analysis of rates of groundwater filling and draining (Gilfedder et al., 2002).

The frequency distribution of response times (thalf) are shown in Fig. 2. The interesting feature of these estimates is the times are much shorter than expected, the maximum response time being just over four years. The effect of discounting future costs and benefits to reflect the change in value of money over time will, therefore, be small. For a discussion on discounting see Pearce and Turner (1990).

Figure 2: Frequency distribution of hydrological response times (thalf) for the groundwater flow systems in the Boorowa River catchment. The response time is the period required for 50% of the recharge to discharge into streams

Area of salinity

The area of saline soils was assumed to be a linear function of base flow based on current trends. This assumption is likely to result in upward bias of the estimate of the saline area. In reality land salinity occurs only where the volume of recharge exceeds the discharge capacity into stream. Water in excess of discharge capacity is discharged to the land surface. In many watersheds this is unlikely to be exceeded. While this is overly simplistic, a more complex description is unlikely to change the results of the analysis sufficiently to change the conclusions.

Stream salinity and yield

The salinity of streams is simply the ratio of total stream flow to the mass of salt that occurs in rainfall and base flow. The mass of discharged salt is dependant on groundwater salinity and the volume of water discharged as base flow. Stream salinity can therefore be lowered by reducing discharge from more saline groundwater systems, increasing discharge from groundwater systems that have lower than average salinity or by increasing run-off.

The change in mass of salt and the volume of water flowing from the Boowora River catchment as a result of a change in excess water is assumed to have a direct impact on the average salt load and stream yield of the Lachlan River. The effect of changes in excess water on stream yield and river salinity is calculated at critical points of the Lachlan River. The location of these is determined by the presence of economic agents that are affected by changes in either of these variables. In this model the critical points are the Cowra, Forbes, Lachlan and Parkes local government areas (LGA).

Excess Water

The annual volume of excess water for the total catchment used in the analysis was taken from Vaze et al. (2004). They used CATSALT to estimate a mean annual rate of recharge under current land use for the Boorowa catchment of 23 mm per year. The total annual runoff from the catchment was estimated to be 58 mm. The total excess water for the catchment is therefore around 80 mm annually, based on the modelled estimates. In the absence of current spatial information on land use, it was assumed that 80 mm of excess water occurred on each hectare of agricultural land in each GFZ. This excluded the area of trees as it was considered unlikely that excess water could be decreased in these areas.

Derivation of soil classes

The proportion of excess water that contributes to deep drainage depends primarily on the soil texture and the location in the landscape. As a general rule heavier textured soils have a lower drainage fraction compared to lighter soils, however soil profile can influence this. Duplex soils with clay subsoil are less permeable so permit less deep drainage than soils with coarser textured subsoil.

No reliable spatial data for soils was available for the catchment to estimate recharge fraction. The Normalised Difference Vegetation Index (NDVI) was therefore used as a surrogate estimate of soil texture, as calculated from Landsat Thematic Mapper (TM) imagery. The imagery was captured on 22 October 1997 and was chosen on the basis that near-optimum spring growth for the area occurred around this time.

Vegetation indices provide a general measure of vegetation vigour or health, and the NDVI is one of the most common. Such indices can be used to separate vegetated from non-vegetated areas, as a surrogate for leaf area index or biomass, and to distinguish between different types and densities of vegetation (Campbell, 1987). The index is defined as:

(3)

where NIR = Near infrared reflectance, in this case Landsat TM Band 4 and R = red reflectance, in this case Landsat TM Band 3 (Rouse et al., 1973).

Actively growing vegetation strongly absorbs red light for photosynthesis, and strongly reflects near infrared light. The NDVI has a range of negative one to positive one, with vigorous herbaceous vegetation giving an index of close to one, and vegetation of poor vigour providing a low index value, although greater than zero. Areas that are covered by water, are bare or poorly vegetated provide very low NDVI values. Tree cover was masked out.

In the Boorowa catchment the more productive soils are typically heavier in texture and would therefore have a low recharge fraction compared to the less productive soils that are lighter in texture. Based on local knowledge of the geology and soil types of the catchment, the NDVI data was divided into three categories; low (<0.45), moderate (0.46 – 0.60) and high (>0.61) where higher values indicates more plant vigour. Recharge fractions were estimated for each soils class (Table 2). These data were merged with the watersheds in a Geographic Information System (GIS) to provide data on the area of each class within each watershed.

Using these estimates of soil area and recharge fraction the total run-off and recharge were calculated. These estimates compare favourably to measured stream flow and salinity data for the catchment and given the uncertainty in many of the parameter estimates the differences in salinity are deemed acceptable by the authors.

Table 2: Characteristics of soil classes in the Boorowa River catchment

Economic Impacts

Land salinisation

The economic impact of increasing the area of salinity on agricultural land was estimated by Wilson (2002). The known areas of dryland salinity in each LGA were identified from GIS data provided by various state agencies. The relevant dryland salinity costs from that study were then extrapolated to each LGA, taking into account the relative differences in salinity area. Changes in the area of salinity in the catchment are assumed to be a linear function of base flow, while the cost to farmers is dependant on the area of saline land and the loss in gross margin from saline land. This could over estimate the loss in income to land holders, as it assumes there is no opportunity cost of salt affected land. A number of farmers in the region have successfully introduced pastures species into affected parts of the landscape and these are considered to be a valuable source of feed at different times of the year.

Urban users

Urban users of water will be affected primarily by changes in the salinity of the Lachlan River. The cost (or benefits) to commerce and industry of changing river salinity is assumed to be a linear function of stream EC, whereas the costs to urban households is a power function of stream EC. Parameter estimates for these cost functions were made based on data from Wilson (2003).

It is assumed that there are no economic effects of reduced stream flow for urban households or industrial users. The economic impact of water restrictions from reduced flows on households would be difficult to estimate and industrial users are high security users and therefore much less likely to be adversely affected by water restrictions.

Salinity

Stream salinity can adversely affect agricultural production in two ways. Firstly, crop yield will be adversely affected by salinity above a species dependant threshold. However, salinity levels in the Lachlan River are unlikely to reach critical levels. Secondly, frequent application of water for irrigation will lead to a gradual accumulation of salt in the soil profile. Yields will gradually decline as the salt stores reach critical levels and eventually the irrigated land will become unproductive. This has occurred in many irrigated regions of the world after decades or centuries of production. At present there is no information that might be used to predict when this might occur in the irrigated areas of the Lachlan River catchment. Therefore this aspect of the problem has not been considered in the analysis. However, the model does provide the means to estimate the economic value of delaying the accumulation of salt and the loss in crop yields.

Value of water from the Boorowa River

The value of water flowing from the Boorowa River catchment to the Lachlan River depends on the effect these flows have on the availability of water; (i) to irrigators and (ii) for environmental purposes. Estimating the value of water to the environment is problematic. Economic analytical methods have been developed that can be used to quantify environmental values in economic terms but they are relatively expensive and there is much debate over whether estimated values reflect actual values. The development of water management plans in the Lachlan River catchment does enable a value of water to be implied. Irrigators were required under the plan to relinquish three percent of their allocation for environmental flows. By implication a judgement has been made by the catchment community that the redirected water is at least as valuable to the environment as it is to irrigators.

A linear programming (LP) model of the Lachlan River irrigation zones was used to determine the cost to irrigators of a small reduction in the availability of water over a 100 year period. Flow data for the Lachlan River was estimated using IQQM. The LP model is solved to select the mix of crop and livestock enterprises which maximise regional gross margin subject to constraints on land, water and labour availability. For a description of the model see Jayasuriya et al (2001). Estimates of shadow price were between $3140 and $0/ML, with a mean of $151 and a median of $37/ML. For the initial analysis $20/ML was used. At this price 322 of the 500 shadow prices from the model results were higher than this. Trades of temporary water in the Lachlan River catchment during the late 1990’s were between $10 and $30/ML (Panta et al, 1999), although these have been as high as $350/ML during the past 18 months where drought conditions have prevailed.

Method of Analysis

A number of simulations were completed to estimate the benefits of reducing excess water on different soil classes and groundwater flow zones. Benefits (and costs) occur due to changes in the (i) area of land salinity, (ii) salt concentration (salinity) of streams and (iii) stream yield.

Each simulation constituted a catchment management strategy. Excess water on three soils classes in each of the four groundwater flow zones were reduced by 22%. This included a simulation that reduced excess water over the whole catchment, apart from areas of bushland. This is a total of 13 simulations. These runs were repeated at different water prices to assess the robustness of the conclusions.

Results and Discussion

The results show that broad scale reduction of excess water in the Boorowa River catchment will lead to a decline in economic welfare of water users downstream. Economic welfare declined by $6.5 million in present dollars terms over 50 years when excess water was reduced by 22 per cent of total excess water. This is the estimated change is excess water that would occur if lucerne were to be established over the whole treatment area. For each hectare on which excess water was reduced the net cost to the catchment was $2.20 (Fig. 3).

Figure 3: Total benefit of reducing excess water by 22% in four GFZ’s and three soil classes (per hectare of land on which excess is reduced). Soils differ in the fraction of water that contributes to recharge and GFZ’s differ in the level groundwater salinity (EC values are 0.9, 1.6, 2.7 and 4.1 ds/m respectively).

Welfare also declined with all but one of the management strategies where excess water was reduced on individual land management units (Fig. 3). Only where excess water was reduced on Soil Class 3 in GFZ 4 did economic welfare of the Lachlan River catchment increase. The net benefits to the catchment were $1.10 for each hectare on which excess water was reduced.

The greatest benefit (and lowest cost where benefits were negative) of reducing excess water occurred where the leakier soils are targeted (Fig. 3). This is because a reduction in excess water on leakier soils will lower base flow (which is more saline than run-off) by a greater amount than a reduction in excess water on soils with a lower recharge fraction. Therefore there is a larger reduction in saline flows. The interesting feature of these results is that they show that targeting more saline groundwater systems will not necessarily result in lower stream salinity. When the soils with the lowest recharge fraction were targeted in GFZ 4 stream salinity increased, resulting in a net cost to the Lachlan River catchment (Fig. 3).

Figure 4 shows the factors contributing to the changes in net welfare for all management strategies. Reducing the area of saltland increases the welfare of the Lachlan River catchment irrespective of the nature of the soil or the zones targeted. Limiting the leakage of excess water lowers the water table, thereby reducing discharge and hence the accumulation of salt on the soil surface. Limiting recharge on the soils where the recharge fraction is highest will reduce the area of saltland by the greatest amount and therefore lead to greater catchment benefits.

Figure 4: Benefits of a change in salt land area, water quantity and stream EC resulting from a 22% reduction in excess water on different soil classes and groundwater flow zones (per hectare of land on which excess is reduced). Soils differ in the fraction of water that contributes to recharge and GFZ’s differ by groundwater salinity (EC values are 0.9, 1.6, 2.7 and 4.1 ds/m respectively).

Reducing excess water will inevitably lead to lower stream yields resulting in a loss in catchment welfare (Fig. 4). The magnitude of the cost depends most importantly on the value ascribed to the lost water. Small differences between soils in the cost of lower yields occurred due to the lags in groundwater response. Stream flow will decline more slowly where leakier soils are targeted as a greater proportion of excess water flows to streams via the groundwater table.

Stream salinity is only reduced and catchment welfare increased when excess water is reduced in GFZ 3 and 4 where groundwater salinity is highest (Fig. 4). Reducing excess water in GFZ 1 and 2 where groundwater salinity is low, caused an increase in stream salinity and a decline in catchment welfare. Targeting zones with low groundwater salinity will lead to a reduction in the flow of relatively fresh water. These flows dilute the more saline flows from zones with higher groundwater salinity.

The results of this analysis were sensitive to small changes in estimates of soil parameters or groundwater salinity. Data used in this analysis is based on limited scientific measurement and some catchment properties, such as soil characteristics, are inferred. It is therefore not possible to make confident recommendations for specific locations within the catchment, on the basis of these results. Figs. 3 and 4 show that the benefits of reducing excess water is dependent on both groundwater salinity and the recharge fraction, so improved information on these factors is essential to improve the confidence of spatially specific recommendations.

The value of water also has an important influence on the catchment benefits of salinity management. The value of water assumed in this analysis is considered to be conservative but it is consistent with the price paid in temporary trades of water allocation. During the late 1990’s the trade price was between $10 and $30/ML (Panta et al, 1999), with the average price being $25/ML. A small increase in the ‘average year’ price (<$10/ML) would result in the benefits of salinity management being much less than the costs of reduced stream yield for all management strategies evaluated.

These results have a number of implications for the management of dryland salinity. As previously stated, the cost of reducing stream flow outweighed the benefit of reducing stream and land salinity for all but one of the management strategies examined. This implies that the benefits of increasing stream flow will be greater than the cost of increasing salinity for those strategies. Indeed increased stream flow through appropriately targeted management has the potential to lower stream salinity further benefiting downstream users. It can be inferred from Fig. 3 that the benefits of higher stream yield and lower salinity will be greater than the cost of increasing the area of land salinity in the Boorowa River catchment.

The large benefits associated with increasing excess water in some watersheds raises a question about the focus of policy on recharge reduction. Reducing stream salinity and limiting reductions in stream yield are best achieved by increasing excess water in some areas while decreasing excess water in others. The effect of salinity management on stream yield and stream water quality should be considered explicitly in order to achieve a more socially desirable balance between economic and environmental objectives.

The trade-offs outlined above are not an exhaustive list. Neither are trade-offs unique to the problem of dryland salinity. The discussion above highlights the need to consider the impact of resource management strategies on a range of resource objectives, including those not considered in this analysis.

Targeting management within a catchment is problematic. There is presently inadequate data available to have confidence in the effectiveness of amelioration strategies and the costs of data collection are substantial. The loss in catchment welfare of untargeted management however, is less than $4/ha of land on which excess is reduced. Fig. 3 shows that the great loss in welfare occurs when excess water is reduced on Soil Class 3 in GFZ 1. Amelioration strategies such as the adoption of a new perennial plants species need only increase farm profit by a small amount to outweigh the economic costs. They would need to be substantially more profitable than current species to ensure adoption by landholders on a broad scale.

Currently salinity policy aims to encourage the adoption of unprofitable perennials using direct subsidies to farmers. From an economic perspective the use of subsidies are justified only where the downstream (off-site) benefits of reducing excess water outweigh the cost of the public expenditure on the subsidies and administration of the subsidies. In this study offsite benefits were small (per hectare managed) and only occurred where management was targeted to specific parts of the catchment. In every case the on-site or private benefits of reducing excess water were positive and substantially higher than the net benefits downstream.

The economic rationale for subsidies in the context of salinity management is to ensure that adoption of management practices occurs at a level that is socially desirable. Landholders will under invest (from societies perspective) in management of salinity where they are unable to capture all of the benefits of their actions. Provision of subsidies would increase the level of investment in salinity amelioration such that it is closer to the socially optimal level. Where the off-site benefits of management are small however, the optimal level of investment from societies perspective will be close to that of landholders. Indeed, this analysis showed that there is a cost associated with reducing excess water so landholders’ optimal level of investment will be greater than that of society.

The results of this study that indicate farmers are the major beneficiaries of salinity amelioration do not appear to be consistent with the observed low rates of adoption of salinity management practices. This apparent inconsistency may be explained by comparing the benefits of management estimated in this analysis with the costs of amelioration. A number of studies (eg Kingwell et al. 2003) have demonstrated that almost all proposed salinity management options have an opportunity cost that is much greater than the benefits of reducing land salinity estimated in this analysis. That is, farm profit would decline significantly, despite reducing the area of land salinity.

This has implications for the role of other market-based instruments in reducing the impacts of dryland salinity. The aim of such instruments is to improve the economic welfare of the catchment by internalising the off-site costs. That is, landholders are forced via a market instrument, to consider the costs of salinity in their production decisions. In principle, the additional cost (or subsidy) to the farmer after the market-based instrument has been imposed will be equal to the off-site costs of production. The results of this study indicate that in the Boorowa River catchment there is a net off-site benefit to current production practices (compared to current species that would reduce excess water) in most of the catchment. Where current practices do lead to off-site costs these costs are small. The optimal cost (or subsidy) to landholders that would accrue from the imposition of a market based instrument will be therefore be small and therefore likely to be insufficient to alter their adoption of salinity management practices.

Further analyses are required to identify catchments where the off-site costs are sufficient to warrant government intervention through the imposition of market-based instruments. Should such catchments be identified the successful application of market-based instruments is not assured. These instruments would need to be carefully targeted to areas of the catchment where a reduction in excess water will lead to desirable outcomes, in terms of land and river salinity and stream yield. In the absence of appropriate targeting market-based instruments will be ineffective at decreasing stream salinity or maximising the social welfare of the catchment community. As previously stated, such targeting is presently problematic due to the lack of detailed spatial data required to provide the necessary characterisation of catchments in NSW.

The uncertainty regarding the off-site effects of reducing excess water in the Boorowa River catchment and more particularly the uncertainty regarding the magnitude of benefits and costs of providing subsidies to farmers warrants a cautious approach to implementation of amelioration strategies. This analysis showed that misdirected effort will result in substantial costs to downstream water users and benefits will only accrue downstream where specific parts of the catchment are targeted. On the other hand the costs of delaying adoption of amelioration strategies is likely to result in only small costs to landholders. Delaying adoption will ensure costs are not imposed through inappropriate targeting of management practices. There is some concern about the additional costs that will be incurred due to an increase in land and river salinity but this appears to be unfounded, particularly in catchments similar to the Boorowa River catchment. As stated in the methods section of the paper the response times estimated for the Boorowa River catchment are very short. Given these short response times and the number of years the catchment has been used for agricultural production, land and river salinity has most likely reached a new equilibrium. That is, it is reasonable to expect that the trend in land and river salinity in the Boorowa River catchment will be flat.

Pannell (2001) has argued a major focus of salinity policy should be research and development, particularly to make ‘improvements in the range and scope of profitable perennials’. Economic welfare in the Boorowa River catchment would be increased by the identification or development of new perennial species that improve farm profitability by greater than $3.50/ha. The downstream benefits of establishing species that reduce excess water will be greatest where they are most suited to lighter soils that have a high recharge fraction.

Conclusions

Broad scale planting of currently available perennial species in the Boorowa River catchment is likely to reduce economic welfare of downstream users in the Lachlan Catchment. While there will be some benefits to landholders of reducing recharge, downstream users will have less water available and stream salinity will increase.

Downstream costs of reducing stream yield will be significant if excess water is reduced on a broad scale in a non-targeted way. This is likely to occur in the absence of sound scientific data to characterise catchments. Downstream benefits of salinity management will be positive only where management is targeted in watersheds that have high groundwater salinity and on soils with a relatively high recharge fraction.

The trade-off between stream salinity, stream flow and land salinity highlights the need to manage natural resource issues in a systems context. The effect of salinity management on other natural resource objectives needs to be considered to achieve the most appropriate balance of natural resource outcomes. Adopting salinity management strategies without consideration of these influences on could lead to adverse impacts that outweigh the benefits of reducing recharge.

The provision of subsidies to increase the area of perennial plants is unlikely to improve the economic welfare of the catchment. Where subsidies lead to untargeted reductions in excess water the value of lost water is very likely to be greater than the benefits of reducing land and river salinity. It is also unclear whether subsidies would lead to improvements in the quality of environmental assets, particularly wetlands connected to the Lachlan River.

Acknowledgements

We acknowledge the support and assistance of Susanne Wilson for contributing to the development of the catchment model David Buckland for contributing to the development of the model and providing feedback on a draft of the paper, Udai Pradhan for his patience and persistent in completing the GIS analysis, Rohan Jayasuriya for providing data on water prices and Tahir Hameed for completing analyses with IQQM and providing the output. Acknowledgement does not imply endorsement of the paper. The views expressed in this paper are the views of the authors and not necessarily those of the Department of Primary Industries.

References

Beumer R (2001) ‘Water Sharing Plan for the Lachlan Regulated River’. Department of Land and Water Conservation, Orange, NSW.

Campbell JB (1987) Introduction to Remote Sensing. Guilford Press: New York.

Coram JE, Dyson PA, Houlder PA, Evans WR (2000) ‘Australian Groundwater Flow Systems Contributing to Dryland Salinity.’ National Land and Water Resources Audit, Canberra, ACT.

Dawes WR, Gilfidder M, Walker RR, Evans WR (2004) Biophysical Modelling of Catchment-Scale Surface Water and Groundwater Response to Land Use Change. Mathematics and Computers in Simulation 64, 3-12.

Department of Land and Water Conservation (1998) ‘Integrated Quantity-Quality Model (IQQM): User Manual.’ (Paramatta, NSW)

Department of Land and Water Conservation (2003) ‘Lachlan Catchment Blueprint.’ (Orange, NSW)

Environment Protection Authority (1997) ‘Proposed Interim Environmental Objectives for NSW waters’. (Sydney)

Evans WR, Gilfidder M, Austin J (2003) ‘Biophysical concepts to underpin socioeconomic modelling of the Little River.’ NSW Agriculture, Orange.

Gilfedder M, Smitt C, Dawes WR, Petheram C, Stauffacher M, Walker GR, (2003) ‘Impact of changes in recharge on groundwater discharge: Development of a simplified function using catchment parameters.’ Land and Water Technical Report 19/03, CSIRO, Canberra.

Gilfedder M, Stauffacher M, Walker, GR and Coram J, (2002) Characterising groundwater flow systems using a dimensionless similarity parameter. In: ‘Proceedings of International Association of Hydrogeologists, International Groundwater Conference.’ Darwin, NT

Greiner R (1999) An integrated modelling system for investigating the benefits of catchment management. Environment International 25, 725-734.

Gutteridge, Haskins and Davey PTY LTD (1999) ‘Salinity Impact Study’. Murray Darling Basin Commission. (Canberra)

Heaney A, Beare S, Bell R (2000) Targeting reforestation for salinity management. Australian Commodities 7, 511-518.

Ivey ATP, Wilson Land Management Services (2000) ‘Costs of land degradation in the BRC and downstream.’ Boorowa Regional Catchment Committee.

Jayasuriya R, Crean J, Hannah R (2001) ‘Economic assessment of water charges in the Lachlan Valley.’ NSW Agriculture, Orange.

Jayasuriya R, Crean J, Hannah R (2001a) ‘The on-farm impacts of environmental flow rules in the Lachlan Valley.’ NSW Agriculture, Orange.

Kingwell R, Hajkowicz S, Young J, Patton D, Trapnell L, Edward A, Krause M, Bathgate A (2003) Economic enalusation of salinity management options in cropping regions of Australia. Grains Research and Development Coroporation and the National Dryland Salinity Program. (Canberra)

National Land and Water Resources Audit (2001) ‘Australian Dryland Salinity Assessment 2000.’ (Canberra)

Pannell, D.J. (2001) Dryland salinity: economic, scientific, social and policy dimensions. The Australian Journal of Agricultural and Resource Economics 45 517-546

Panta K, McMahon TA, Malano HM, Turral HN (1998) ‘Water balance of the Lachlan River system’. Centre for Environmental Applied Hydrology, The University of Melbourne (Parkville, Victoria)

Pearce DW, Turner RK (1990) ‘Economics of natural resources and the environment.’ (The John Hopkins University Press, Baltimore, Maryland)

Read, M., Watson, A., Sturgess, N., Pannell, D., (2001), ‘Capacity to Change – case studies of dryland salinity and watertable control (constraints on adoption of salinity control).’ National Land and Water Resources Audit (Canberra)

Resource Analysis Unit Central West Region (2001) ‘Assessment of Riverine Salinity in the Macquarie and Cudgegong River Catchments 1999/2000.’ Department of Land and Water Conservation, Orange, NSW

Rouse JW, Haas RH, Schell JA, Deering DW (1973) Monitoring vegetation systems in the great plains with ERTS. In: ‘Proceedings of the 3rd ERTS Symposium.’ 1, 301-317.

Vaze J, Barnett P, Beale G, Dawes W, Evans R, Tutega N, Murphy B, Greeves G, Miller M (2004) (in press). Modelling the effects of landuse change on water and salt delivery from a catchment affected by dryland salinity. Hydrological Processes.

Wilson SM (1994) ‘Procedures for economic assessment of management option for dryland salinity.’ The Land and Water Resources Research and Development Corporation. Occassional Paper No 06/94 (Canberra)

Wilson SM, Laurie I (2002) ‘Costs functions to assess the cost of saline town water supplies to households, commerce and industry.’ Murray-Darling Basin Commission (Canberra)

Wilson SM (2002) ‘Dryland Salinity: What are the impacts and how do you value them?’ Natural Dryland Salinity Program.

Wilson SM (2003) ‘Economic model of salinity and its management in the Lachlan River Catchment: User Manual, Version 2.0.’ NSW Agriculture, Orange.

***

Citation: Bathgate, A., Woolley, J., Evans, R. and McGowen, I. (2004) Farm and catchment benefits of reducing recharge to ameliorate dryland salinity: an economic study of the Boorowa River catchment in NSW. SEA Working paper 1704. CRC for Plant-based Management of Dryland Salinity, University of Western Australia.