SEA Working Paper 99/03
Dryland Salinity Control: What Do We Usefully Know (in 1999)?
Steven Schilizzi and Gavin White
Agricultural and Resource Economics, The University of W.A., Nedlands, 6907
Abstract
In a year long survey of the available literature, we investigated how much of the research work on dryland salinity in Australia could be used for informing farm management decisions, policy formulation, and bioeconomic modelling. Out of over 2000 references screened, we found a little over 50 to be of direct use, providing some 200 pieces of useful information. Focusing on the relationships between land use and salinity, we organised the information into three categories, distinguishing: differences between trees, crops and pastures; impacts of tree species, density and plantation design; and impacts of rotations with crops and pastures. We also distinguished between biological and mechanical rehabilitation strategies (ground-cover vs. pumping, drains and contour banks). For each category we identified what was considered to be somewhat known and what was as yet poorly known. We found that the main types of useful information came as quantitative relationships, numerical parameter ranges for specific site conditions, identification of factors that do make a difference, scale effects, and acceptable simplifications. We conclude from this survey that a reallocation of research effort is desirable towards the production of more management-related knowledge. Such an effort could enhance the effectiveness of salinity and, more generally, sustainability indicators for land use.
Purpose and objectives
Thanks to special funding from the Faculty of Agriculture, University of Western Australia, we were able in the year spanning across 1997-98 to survey the existing literature on dryland salinity in Australia, and particularly in Western Australia. More than 2000 references, both published and unpublished, were screened, including an international database search, and many authors were personally interviewed. We found that much of the literature was in unpublished form, mostly because the authors considered the results of their work still needed confirmation.
This survey had a very specific purpose. We wanted to tease out of the usually very technical, scientific studies the knowledge that could be used for practical reasons. We had three in mind: farmers’ management decisions, salinity policy formulation, and bioeconomic modelling. Information was tagged "useful" if we thought it could contribute to any of these three objectives. We had three questions directing our search: 1) What do we (somewhat) know? 2) What do we yet need to know? 3) What must research emphasise? Moreover, we were targeting "usefulness" to specific questions: Can we assess the costs of salinity? What are the costs and benefits of salinity abatement strategies? How do these change over time in relation to hydrological dynamics? What is the extent of the risks and uncertainties that face farmers? What site-specific parameters make a difference?
Organisation of information
Using these screening criteria, we found a little over 50 references to be of direct use, providing some 200 pieces of useful information. These were all tabulated in an Excel workbook and tagged to their specific sources. We found it convenient to organise the information according to the relationships between land use and salinity. We distinguished three categories: 1) differences between trees, crops and pastures; 2) impacts of tree species, tree density and plantation design; and 3) impacts of crop and pasture rotations. We also distinguished rehabilitation strategies depending on whether they were biological (using ground-cover) or mechanical (pumping, drains and contour banks).
We identified three key types of factors affecting land use and salinity. Site-specific factors included rainfall, hydrogeology (deep water movements and storage), slope, and soil types, affecting water retention. Ground-cover factors included plant density coverage (biomass/ha), interception coverage (leaf area index), root system depth, mass and extension, and seasonality (winter vs. summer events). Management factors included the location, the timing, the type and the intensity of activities.
Salinity was found to be measured in three different ways, possibly leading to different estimates of "salinity": watertable depth; soil electro-conductivity (dS/m), and total dissolved salts/volume (mg/L), also expressible as salinity relative to average sea-water (% 30g/L). Note that watertable depth is dependent on location and timing, and obtaining reliable spatio-temporal maps is very costly.
We found that useful information as defined above could be into five categories:
Land use and salinity
Trees, crops and pastures.
Research carried out to date showed several aspects had been usefully investigated: the relationship between rainfall and groundwater level, distinguishing between short term and long term effects; water uptake rates by different species, especially relating rainfall to plant cover and evapotranspiration; effect of salinity on plant transpiration rates; the effect of salinity on plant growth; and salinity change under tree cover in higher rainfall areas (>600 mm/year).
Issues we considered to be yet poorly known included: the rate of salinity change under trees in lower rainfall areas (<500 mm/year); the effect of salinity on tree productivity; best location of trees on slopes (up, mid or down slope); errors due to scale effects, particularly regarding whole vs. part catchment areas; how farmers’ decision variables affect catchment variables; area and species-specific risks of species being salted out, drowned by excess waterlogging, or dying of drought (affecting the percentage surviving after N years); and many interaction effects, e.g. how catchment size affects uncertainties.
Tree species, density and plantation design
There is useful information on how tree species and density affect water draw-down rates, how tolerant certain trees are to salinity of soil and groundwater (in terms of survival), and on the time required for trees to claim saline land to levels before clearing or logging. However, figures are sensitive to area rainfall.
Much less is known regarding the effects of plantation design, storativity and transmissivity parameters and site-specificity and spatial variability of these and other values. There is a need for representative case studies to accumulate in order for key parameter ranges to be identified for different categories of sites. This would also improve comparability between studies, which at present remains weak.
Crop and pasture rotations
This approach to salinity control is motivated by a faster potential for adoption, since it relies on current farming systems and entails low investment costs.
Research has produced good information regarding relationships between time of sowing, seeding rates and water use by lupins; evapotranspiration rates for different species in different climate-soil conditions; and has showed that, in general, recharge control is more effective than discharge control.
Far less is known about how farm management decisions affect water use efficiency of the agricultural system; on how increased water use affects production, and what are the conditions for a strong positive correlation; the effects of the timing of farming operations; how well one can extrapolate from lupins to other legumes; and how rotational sequences affect recharge control. These deficiencies place limits on the predictive power of simulation models such as ‘AgET’.
Rehabilitation strategies
Biological techniques, based on use of ground-cover by plants
Most of the work published in this field has focused on the water draw-down capacity of trees and pastures, both perennial (like lucerne) and annual. Other work which may be ongoing was not yet published at the time of our survey.
There is still much to be known regarding the productivity of reclaimed sites, the rate at which site productivity recovers, the time needed for recovery to some pre-specified reference state, and the carrying capacity of agroforestry pastures.
Mechanical techniques: pumping, drainage, and contour banks
There has been much work in this area, presumably because results can be obtained more quickly. Much information exists on the drawdown and radial action of pumping and on the conditions for contour banks to be effective. Pumping can be an attractive option if it is not too costly, if the aquifer is not too complex, and, most importantly, if the salty effluent can be safely discharged - a condition likely to be hard to meet in most cases. Contour banks have been shown to be effective, and necessary, in the initial stages of saline waterlogging.
There are however other aspects that are less researched: the time needed and rate at which saline waterlogged land can regenerate in terms of productive capacity; the effects of water barriers as opposed to water conductors on vertical and horizontal water movements (e.g. dikes, holes, and quartz veins); the impact of soil storativity and transmissivity on the efficiency of pumping.
Conclusions
The main conclusion of this study is that if the research efforts invested so far are to be of practical use in controlling dryland salinity, some research effort must be redirected to management-related knowledge production. This means a reassessment of the trade-off between basic research and applied research. More effort must be focused on helping decision-makers at all levels, farmers, catchment managers and policy makers, together with improved risk assessment of management strategies.
Our survey also suggests that an increased coupling of observational and experimental fieldwork with modelling is needed. This has been done in hydrology, but less so in other fields. There is a special need to improve our knowledge about the rates at which things happen and what affects the dynamics, and to focus more on scale effects and threshold values.
Citation: Schilizzi, S. and White, G. (1999). Dryland Salinity Control: What Do We Usefully Know (in 1999)? SEA Working Paper 99/03, http://www.general.uwa.edu.au/u/dpannell/dpap9904f.htm
For a copy of the complete paper, contact Steven Schilizzi at schilizz@cyllene.uwa.edu.au
The SEA News index is at http://welcome.to/seanews