Relationship between yield of grain sorghum (Sorghum bicolor) and soil salinity under field conditions

Three field experiments using grain sorghum (Sorghum bicolor), an important dryland summer crop on the Liverpool Plains in northern New South Wales, w ere conducted: (i) to determine the effect of dryland salinity on the yield of commercial crops at 2 sites; (ii) to see if ridging the soil would amel iorate the problem; and (iii) to compare 16 commercial varieties for tolera nce to dryland salinity. Grain sorghum was shown to be more severely affected by dryland salinity th an most literature would suggest. Over 3 seasons and 2 sites, sorghum yield was reduced by 50% at soil electrical conductivity (saturation extract, EC e) levels as low as 2.8 dS/m whereas advisory literature indicated a salini ty threshold (no yield reduction) for sorghum of 6.8 dS/m, and 50% yield re duction at 9.9 dS/m. Current advisory literature is based on research where salinity was artificially imposed after plants were established in non-sal ine soil. The measurements described in this paper were on sorghum sown int o saline soil. Soil and crop management strategies (ridging the soil or choosing a toleran t variety) showed limited potential for improving yields of grain sorghum o n saline soil. At one site, the ECe varied widely across the paddock but little down the s oil profile at any sampling point. Hence, analysing the surface soil would indicate the salinity hazard. However, at a second site, where ECe levels i n the surface soil were low (< 2 dS/m) everywhere, ECe at soil depths of 1 m varied widely (from 2 to 15 dS/m) across the paddock. Soil sampling to as sess salinity hazard before crop planting should therefore include the enti re root zone.