SEA Working Paper 98/04

The dynamics of phase farming in dryland salinity abatement

Ute Mueller¹, Steven Schilizzi² and Tuyêt Tran¹

¹Department of Mathematics, Edith Cowan University, Mt Lawley 6050 Australia
²Agricultural and Resource Economics, The University of Western Australia, Nedlands 6907 Australia

The problem

This paper investigates the economics of recursively degrading and rehabilitating farmland. The type of land degradation is salinity, measured by depth of saline watertable, and is brought about by continuous cropping. Rehabilitation is considered to be achieved by tree plantations. The authors examine the extent to which such a system of successive cropping and tree plantation phases makes economic sense in the wheatbelt conditions of Western Australia (rainfall 300-600 mm/year). In particular, given reasonable assumptions on site conditions, they identify the optimal lengths of each phase and how farm profits are affected. Although this problem has clear practical implications, the basic motivation behind this work is in solving a technical difficulty. Given that switching from one phase to another involves up-front costs, computing the optimal duration of each phase is no trivial matter. Indeed, the problem was solved only for three phases: an initial cropping phase, followed by a rehabilitation phase, followed by a second and final cropping phase. The overall maximum time span was set equal to 100 years.

Investigating the problem with a computer model

This phase farming problem assumes from the start a specific type of situation: trees can reverse the salting process, before being themselves affected, by drawing back down the saline watertable. Watertable draw-down rates in the wheatbelt are as yet poorly known. Here they are assumed to range from 2 to 4 cm/year. Tree density, which varies in the model, affects draw down rates. Rising rates under cropping range between 4 and 14 cm/year. These values are best-guess extrapolations from observed situations in the wetter parts of south-western Australia, and may be over (or under) estimates.

The starting point is one where salinity has already become a potential problem to the farm: the initial watertable depth is 4 metres and is rising. For simplicity, the cropping phase assumes a unique crop (wheat). The tree phase can be interpreted as a dense block plantation, or as an agroforestry scheme, where part of the area previously sown to wheat is planted to trees. Location-wise, trees are assumed to be planted on ground slightly higher than the lowest, where crop yields are first affected. Implicitly, therefore, two adjacent areas are considered. Salt will affect crop yields on lower land before it affects trees slightly up-slope.

Crop yields are reduced by rising watertable levels. The assumption is that crops are not affected up to a point, after which yields fall from a maximum (when not affected by salt) down to zero in proportion to the rise in watertable level. The starting value (-4m) is just this threshold point. Trees draw down the water level (in the form of an S-curve) depending on tree density and the maximum achievable depth (-6m). The growth function of trees roughly represents that for oil mallees in the wheatbelt.

Trees have value through their rehabilitation impact and their product, leaves for oil extraction for the solvent industry (as researched by J. Bartle, CALM). However, economic returns from trees are much lower than from crops. The short term losses from replacing crops with trees are thus high. Also, switching from one phase to another is costly: land preparation, tree planting and protection, crop seed bed preparation. The data used in the study are summarised in table 1 below.

Finally, waiting is costly and the value of time is important to farmers. Accordingly, future returns are discounted, giving more weight to short term than to long term benefits.

The optimal phase farming system is assumed to maximise total farm profits over 100 years, given costs and prices and effects on land productivity over time.

Results

The qualitative results are not surprising, but confirm common sense intuition. The numerical values obtained are more interesting, given the plausible data used as parameters. However, numerical results must not be taken at face value; rather, this study focused on the sensitivity of farm profits and phase durations to changes in discount rates and watertable movements, given prices and costs for both phases.

Lower discount rates shorten the optimal time to rehabilitation and make this phase longer, ultimately driving up total profits. Lowering the annual discount rate from 10% to 4% shortened the first cropping phase from 64 to 25 years, lengthened the rehabilitation phase from 9 to 15 years, and lengthened the second cropping phase from 27 to 60 years. Total discounted profits just about doubled. Reversing the story, high discount rates make it profitable in the short run to let salinity drive down farm productivity, the cost of which, however, ends up being higher than the perceived short term benefits.

A quicker rate in the rise of watertable level under cropping phases reduces total profits, shortens the optimal time to rehabilitation, lengthens the rehabilitation phase, and shortens the second cropping phase. Given the data in table 1, an increase from 4 to 14 cm/year, in an economically optimal strategy at 8% discount, shortened the first cropping phase from 37 to 21 years, lengthened the tree phase from 11 to 17 years, shortened the second cropping phase from 52 to 21 years, and drove down total profits by one third. The total duration of the three phases was optimally reduced from 100 to 59 years. Note however that these numbers, as well as the ones below, vary with the discount rate and other set parameters.

A quicker rate of draw-down by trees during the rehabilitation phase induces, in an optimal strategy, a shorter first cropping phase, a shorter rehabilitation phase, and a longer second cropping phase. An increase in draw-down rates from 2 to 4 cm/year shortened the first cropping phase from 29 to 17 years and the tree phase from 12 to 7 years, lengthening the second cropping phase from 59 to 76 years. Profits, however, increased by only 8%. Although this lower impact on profits simply reflects a smaller absolute change in water rates, they do represent realistic technical efforts in choosing and adapting new, more water-efficient species.

Conclusions and qualifications

This study shows that phase farming is very sensitive to farmers’ discount rates and to the rates of vertical watertable movements as a function of cropping or rehabilitation. It signals the importance for policy makers to help farmers reduce their discount rates, possibly through ‘green’ (lower) interest rates for investments in rehabilitation, and to encourage them to increase the areas planted to trees. This can be achieved by increasing the value of trees, by improving the end-use of their products and by extending their marketable potential.

Though complex, the solution to the optimal phase farming problem considered in this study has relied on a number of drastic simplifications. The most important are that prices and costs have unrealistically been assumed constant over time, no uncertainties have been included, tree growth is simplified, a single crop is considered, and tree density is not allowed to vary continuously. Furthermore, no management variable is included in the cropping phase.

To say the least, much more work is needed to yield deeper insights into the bio-economic mechanisms that can lead to economically efficient phase farming systems as a means to combat on-farm land degradation.

Table 1 : Parameter values used in the model

Reference

Bartle J.R., Campbell C., and White G. (1996). 'Can trees reverse land degradation?' Paper to the Australian Forest Growers Conference, Mt. Gambier, SA.

Citation: Mueller, U., Schilizzi, S. and Tran, T. (1999). The dynamics of phase farming in dryland salinity abatement, Australian Journal of Agricultural and Resource Economics 43(1): 51-67. http://www.general.uwa.edu.au/u/dpannell/dpap9810f.htm

SEA News issue #2

The SEA News index is at http://welcome.to/seanews

Copyright © Ute Mueller, Steven Schilizzi, Tuyêt Tran 1998
Last revised: May 21, 2003.