Source DocumentPrevious PageTable Of ContentsNext Page

Improving growth and yield of wheat and lupin crops on duplex soils in western Australia

R.K. Belford, D. Tennant and M. Dracup

Western Australia Department of Agriculture, Baron Hay Court, South Perth WA 6151

Crop yields on duplex (Dy) soils in Western Australia are lower than the potential, particularly in wet years, and often associated with marked spatial variability. In 1988 and 1989, we studied spatial variation in yields and the causes of this variation on a duplex soil site near Beverley in the 400 mm rainfall zone of WA. Crop yields on the site varied three-fold within the paddock, and we identified several soil physical and chemical problems: (i) fertility was low, particularly for copper and potassium, and pH in the sand was also low (<4.5 in CaC12) in some areas; (ii) there were large differences in early growth of shoots and roots (2), associated with the texture, cementing and porosity of the sand layer, but not with the more obvious factors of waterlogging, root disease, or soil strength; and (iii) permeability of the clay subsoil varied widely, which had large effects on infiltration, waterlogging, root growth and water extraction. In 1990, we tested agronomic options to overcome these soil constraints in order to reduce variability in crop growth and raise yields; these manipulations are reported here.


Treatments were imposed to correct nutrient deficiencies, raise pH, loosen the sand horizon by deep ripping, and improve the permeability of the clay by incorporating gypsum just above the clay; for comparison, a district practise (DP) treatment was included as control. Each treatment was laid out as a strip along the length of the site, and thus included the variability characterised in earlier work (1). Crops of wheat (Triticum aestivum cv. Kulin) and lupin (Lupinus angustifolius cv. Gungurru) were sown at the end of May.

Results and discussion

Comparison of the manipulative treatments was complicated by delayed sowing on parts of the deep ripped plots, and the underlying variability both across and between plots on the site. However, yields of wheat and lupins respectively were 2.6 and 1.1 t/ha for the control (DP) treatment; mean yields for the other treatments were 3.0 (wheat) and 1.4 t/ha (lupins). For wheat, coefficients of variation within each treatment were 35% for the DP treatment, but 21% for the other treatments. Overall the largest benefits came from improving fertility, with lesser gains from liming, ripping and gypsum.

The manipulative treatments reduced spatial variability in yield and raised yields, which were the two major objectives of the work. Yields of lupins were more variable than yields of wheat, and it is possible that adding lime raised pH to more than the optimum for L. angustifolius. Correcting nutrient deficiencies on these soils and using good agronomy (such as early sowing to minimise waterlogging damage and raise yield potential, the correct variety, and high and split levels of nitrogen for wheat), are likely to improve yields and their sustainability on these soils. The manipulative treatments to improve soil physical conditions gave little additional improvement in yield in 1990, but may have benefits in future years if subsoil permeability gradually improves through the use of gypsum.


Belford, R.K., Dracup, M. and Tennant, D. 1990. J. Agric. (WA) 31, 62-65.

Dracup, M., Gregory, P., Belford, R.K. and Tennant, D. 1992. Proc. 6th Aust. Agron. Conf., Armidale (these proceedings).

Previous PageTop Of PageNext Page