New Knowledge Means New Approaches to Solving Dryland Salinity

Resource Economist, formerly managing director of Read Sturgess and Associates, and recently retired for health reasons

This paper summarises a recent study of dryland salinity, undertaken for the National Land and Water Resources Audit (Read et. al. (2000)). The findings are important because they overturn some long held fallacies that have shaped governments’ policy responses to dryland salinity in Australia. This mainly reflects a new understanding of the biophysical processes involved with dryland salinity (see Coram et. al. 2000 and Pannell 1999).

The study for the National Land and Water Resource Audit by Read et. al. (2000) involved a substantial amount of fieldwork, along with economic modelling, and was concentrated on four particular catchments as case studies:

This work was commissioned in conjunction with scientific studies, for which Commonwealth Scientific and Industrial Research Organisation (CSIRO) specified catchment water balance models for the same four catchments and formed projections about future extents of salinity, for scenarios with and without salinity control in each catchment (Baker et. al. 2001, Hekmeijer et. al. 2001, Short et. al. 2001 and Stauffacher et. al. 2001).

The work by Read et. al. (2000) benefited greatly from other major economic research projects that were undertaken concurrently, particularly those undertaken in Western Australia by Dr David Pannell, and those undertaken by the Australian Bureau of Agricultural and Resource Economics (ABARE) in the Murray Darling Basin. In each of those research projects, the economic benefits and costs of various salinity management options have been compared for particular catchments. In aggregate, those catchments represent a significant and representative sample of areas affected by dryland salinity across Australia, and all the studies have arrived at very similar conclusions.

A common misconception of dryland salinity in Australia has been that it is typified by actions of particular farmers affecting mainly other parts of the catchment where salinity emerges, often long distances from the particular landholder (see for example, ABARE 1992). Such external effects represent ‘economic externalities’ and could justify government funding. The analyses that concluded that external effects were paramount were based on the view that there was a high degree of hydrological transmissivity such that changes in recharge at one location would benefit areas way beyond the area treated.

To the contrary, recent research has shown that the adoption of practices to reduce recharge mainly leads to benefits only for that land on which the treatment is implemented. For example, evidence of the limited area of benefits beyond the site of implementing works to reduce recharge comes from observations of extensive tree planting in Western Australia. George et. al. (1999) surveyed the effectiveness of tree planting as a salinity management measure at 80 sites in Western Australia and concluded that trees had little effect on the water tables beyond 10 to 30 metres from the planted area.

Important research by Coram (2000) undertaken as part of the National Land and Water Resource Audit’s Dryland Salinity theme, has emphasised that such observations are not limited to Western Australia, and that the type of groundwater flow system for each sub-catchment influences greatly the scope of externalities and the effectiveness of particular options for managing and controlling dryland salinity. The extent of the flow system, or the distance between groundwater recharge and groundwater discharge, provides an indication of how quickly salinisation is likely to manifest at the ground surface in each groundwater flow system, and how long management strategies are likely to take to achieve results.

Coram (2000) considered three main types of groundwater flow systems; local, intermediate and regional:

Because of the high incidence of local groundwater flow systems, and the low transmissivity of intermediate and regional groundwater flow systems, it is not common that actions to control salinity by one landholder in one region can have a substantial impact on neighbouring and downstream regions, with respect to land salinisation. It should be noted that this is very different to the hydro-geological processes associated with salinisation due to irrigation, for which externalities are much more relevant.

Results from the case studies are summarised in Tables 1 and 2. Four is only a small number of case studies from which to seek generalisations, but even so the case studies have produced some very interesting results:

Only Lake Warden involves substantial environmental benefits and that case study has emphasised that a major disadvantage for farm-scale treatment is that it is unlikely to lead to substantial improvements for downstream water quality in streams. Only catchment-scale treatment of salinity, or appropriate engineering approaches, can avoid water quality impacts since the hydrologic balance throughout an entire catchment contributes to water quality at the bottom end of the catchment. Local scale treatments can rarely have a substantial impact on water quality as salt continues to be mobilised from the untreated areas of the catchment.

The results for all catchments show conclusively that large-scale recharge control based on tree planting would represent a very poor investment in most catchments. Balancing that disappointing result, the good news is that a shift towards greater use of perennial pastures in crop rotations has been shown to be profitable in some cases.

Trees simply are not well suited to most salinised areas in Australia. The Bureau Of Rural Sciences (BRS) has estimated that the area of cleared agricultural land potentially suitable for commercial timber plantations and subject to salinity risk is only about 6 per cent of the total area subject to salinity risk (Tickle et. al. 2000). Re-vegetation for commercial timber production, under currently accepted parameters, is therefore likely to have only a small role in the overall control of dryland salinity.

Another reason mitigating against the attractiveness of catchment-wide tree planting is that typically only a relatively small area of any one catchment is salinised. For example, in the Wanilla catchment, (unprofitable) re-vegetation for about 70 percent of the catchment would be required in order to protect the 8 per cent of the catchment which is at risk. Furthermore planting of trees also has the effect of reducing surface runoff, with implications for river flows. The effect of tree planting on runoff is relatively immediate and can be potentially large.

There will be situations where catchment-scale tree planting is attractive, but those situations will be the exception rather than the rule. Large-scale re-vegetation will be more likely to occur where:

A classic example of an attractive opportunity for catchment-scale re-vegetation lies in the Collie catchment of Western Australia. Wellington Dam was constructed in the catchment in 1960 with the main purposes of supplying the water supply needs of Perth and Bunbury. The salinity levels of the streamflows have subsequently become highly salinised such that water from the Wellington Dam cannot be used for urban supplies. The Water and Rivers Commission of Western Australia believes that most of that additional salt load has been contributed by two small sub-catchments which are managed primarily for grazing. The cleared area across those two sub-catchments covers only about 16,000 hectares and catchment water balance modelling by the Water and Rivers Commission indicates that tree planting across that 16,000 hectares would lead to a greater reduction in salinity levels than those required to meet their water quality targets for urban supplies from the Wellington Dam.

For supplies equivalent to the safe minimum yield of the Wellington Dam, it would be necessary to spend $1,070 million for the lowest costing alternative water supply. When it is considered that those potable supplies could be achieved by planting trees across only 16,000 hectares, the mean level of benefit for each hectare of trees planted would be about $67,000. That is, on average the recharge reduction from each hectare planted to trees would lead to the avoidance of future capital expenditure for water supply headworks of about $67,000. The reasons for the seemingly nonsensical, continuation of agriculture in the problem sub-catchments of the Collie lie squarely at the political end of the spectrum.

The Collie catchment is an exception. Most salinised catchments across Australia are not well suited to trees, with low rainfall generally being the constraint. The Collie has good tree growing conditions plus a major external benefit. It is not very common to have either of those and the Collie has both. The Wanilla catchment represents the opposite extreme. The Wanilla catchment will not grow trees (nor lucerne and other perennial pastures species) with commercially acceptable yields and there are no substantial external benefits.

There are good technical options for some catchments and these are being implemented in a profitable manner. It is now generally accepted that there is a need to incorporate a significant coverage of perennial vegetation if we are to reduce significantly the level of leakage across the landscape. Trees are generally not going to be viable for large scale salinity treatment, and changing farming systems to substitute perennial pastures for annual pastures is the other way of reducing leakage. In addition, the two main engineering approaches to dryland salinity that have been used in Australia are surface drains and pumps.

An important observation by Read (2000) is that impacts for human consumers of salinised water may not be as high as previously thought since it may be much cheaper to treat the salinised water supplies rather than to control all of the dryland salinity in a catchment. Similarly, many environmental impacts may be treated more cheaply with engineering approaches rather than attempting to control all of the dryland salinity in a catchment (see for example, Lake Warden wetlands).

In terms of economic value, the two important types of externalities from dryland salinity across Australia would be:

The scale of the latter remains largely unknown, but judgements can be made about the former.

The Water Authority of W.A. estimated that reverse osmosis water treatment technology could be used at a cost equivalent to $1,300 per megalitre. At that cost, the entire water supplies of both Adelaide and Perth could be treated to excellent standards for a cost of the order of $340 million per year. This would be equivalent to increasing total water charges for water use in those cities by a factor of about 2 to 3. Reverse osmosis water treatment can produce drinking water attributes similar to those of a pristine mountain stream, even for appallingly degraded water resources such as those supplied presently to Adelaide. The reverse osmosis treatment process can remove taste problems associated not only with salinity, but also with other characteristics such as turbidity (which, interestingly, contributes more to poor taste than does salinity in the case of Adelaide). It seems that such an engineering approach would be much more cost effective, possibly even by orders of magnitude, than attempts at catchment-scale recharge control.

In terms of achieving an economically optimal mix of salinity control measures, it is concluded that:

Read (2000) has emphasised that most recharge control requires landholders to switch from annual crops or pastures to perennial plants, which generally involve more intensive farming systems. Most dryland salinity in Australia occurs on mixed wheat-sheep farms and the traditional farming systems have been based on a low level of inputs. Such low input farming has allowed reduced risks, particularly by providing greater flexibility for landholders to switch between cropping and grazing in response to changes in relative commodity prices. Adoption of perennial pastures greatly increases the level of farming inputs required, and this is a barrier to adoption since landholders do not wish to reduce their flexibility to switch from year to year between cropping and grazing enterprises.

The present state of knowledge suggests the following three groups with respect to likelihood of adoption:

Those who have no option but to live with the salt

As in Wanilla, at present there are no viable options to control salinity in any substantial way for much, possibly even the majority, of areas affected by dryland salinity in Australia. For that large area of Australia, the emphasis must remain on ‘living with the salt’. The hopes for these areas is either that new and better suited control measures are identified, or that exogenous shocks, such as substantial changes in a commodity price(s), lead to the present options becoming viable. This group is likely to be the largest, possibly comprising as much as 30 to 60 per cent of Australia’s dryland salinity.

Those who adopt substantial salinity control

As in the Kamarooka and Lake Warden catchments, landholders do adopt appropriate salinity control measures if they are profitable in their region and/or if the expected level of salinisation in the future is substantial. From the case studies, this group appears likely to be the larger of the remainder of landholders, concentrated particularly in Western Australia, where there is generally a greater justification for implementation of salinity control measures since the impacts are generally higher.

Those who could, but choose not to, adopt substantial salinity control

For example, landholders in the Upper Billabong Creek catchment. For some landholders the expected level of salinisation would not lead them to adopt salinity control options even though some of the presently available options would be marginally viable. They would prefer to retain their present farming systems and ‘live with’ the salt.

It is most promising to see the progress in the Kamarooka and Lake Warden catchments. This has emphasised that very severe salinity, such as is progressing in the Lake Warden catchment and elsewhere in Western Australia, is the like of a massive commodity price shock that is sufficient to achieve substantial adoption of salinity control by encouraging landholders to change farming systems. Many landholders have changed to farming systems that represent only a marginal improvement in profitability and which incur major difficulties for landholders. The change and willingness to accept those difficulties has been motivated by the need to protect against the future expansion of salinity on their properties.

The fallacy that widespread re-vegetation with tree plantations was technically and economically feasible led to a fairly uniform policy response over the past twenty years which emphasised trees for most areas affected by dryland salinity. It is now clear that this has been inappropriate. The major emphasis should be placed on targeting only those instances where other control measures are technically and economically attractive. Those other control measures are likely to comprise mainly farm-scale changes to farming systems as well as engineering approaches.

The other fallacy, that economic externalities were thought to be very substantial, led to conclusions that there should be a substantial amount of Government assistance for landholders who implement salinity control measures. Externalities are limited mainly to (unpriced) environmental impacts on surface waters at the downstream end of catchments. To the extent that those do justify substantial Government funding, then it is important to evaluate carefully whether it is less costly to use engineering solutions to protect the environmental assets at the downstream sites, rather than to change farming systems over enormous areas in the upper catchment.

There are relatively few off-site impacts for downstream farmers, nor for regional and urban buildings and infrastructure. The high incidence of local groundwater flow systems, and low transmissivity for other groundwater flow systems, means that such impacts would be affected mainly only by management of adjacent land, not by land management further afield in the upper catchment, as thought previously.

The finding that there is no viable and substantial salinity control presently suited to most of the area affected by dryland salinity means that Government funding must be directed at R&D aimed at providing a greater range of technical options. Options should be sought for immediate implementation, but others might be identified which could become viable at a later date due to exogenous changes. The more technology is on the shelf, the more chance it can be adopted if circumstances change.

ABARE. 1992, ‘Dryland salinity: some economic issues’, in Natural Resource Management, AGPS, Canberra.

Baker, P., Dawes, W., Probert, M. and Moore. 2001, ‘Assessment of salinity management options for Kamarooka, Victoria: Groundwater and crop water balance modelling’, CSIRO Land and Water project, undertaken for NL&WRA, Canberra.

Coram, J.E., Dyson, P.R., Houlder, P.A. and Evans, W.R. 2000, ‘Australian groundwater flow systems contributing to dryland salinity’, Bureau of Rural Sciences project, undertaken for NL&WRA’s Dryland Salinity Theme, Canberra.

George, R.J., Nulsen, R.A., Ferdowsian, R. and Raper, G.P. 1999, ‘Interactions between trees and groundwaters in recharge and discharge areas – A survey of Western Australian sites’, Agricultural Water Management, vol. 30, pp 91-113.

Hekmeijer, P., Dawes, W., Bond, W., Gilfedder, M., Stauffacher, M., Probert, M., Huth, N., Gaydon, D., Keating, B., Moore, A., Simpson, R., Salmon, L. and Stefanski, A. 2001, ‘Assessment of salinity management options for Kamarooka, Victoria: Groundwater and crop water balance modelling’, CSIRO Land and Water project, undertaken for NL&WRA, Canberra.

Pannell, D.J., McFarlane, D.J. and Ferdowsian, R. 1999, ‘Rethinking the externality issue for dryland salinity in Western Australia’, SEA Working Paper 99/11, GRDC Project UWA251, Agricultural and Resource Economics, University of Western Australia.

Short, R., Salama, R., Pollock, D., Hatton, T., Bond, W., Paydar, Z., Cresswell, H., Gilfedder, M., Moore, A., Simpson, R., Salmon, L., Stefanski, A., Probert, M., Huth, N., Gaydon, D. and Keating, B. 2001, ‘Assessment of salinity management options for Lake Warden catchments, Esperance, Western Australia: Groundwater and crop water balance modelling’, CSIRO Land and Water project’, Technical Report 20/00, undertaken for NL&WRA, Canberra.

Stauffacher, M., Bond, W., Bradford, A., Coram, J., Cresswell, H., Dawes, W., Gilfedder, M., Huth, N., Keating, B., Moore, A., Paydar, Z., Probert, M., Simpson, R., Stefanski, A. and Walker, G. 2000, ‘Assessment of salinity management options for Wanilla, Eyre Peninsular’, CSIRO Land and Water project, Technical Report 1/00, undertaken for NL&WRA, Canberra.

Tickle, P., Keenan, R., Walker, J. and Barson, M. 2000, ‘Can commercial tree planting help dryland salinity mitigation and meet greenhouse gas abatement objectives?’, Bureau of Rural Science, Canberra.