Soil property differences and irrigated‐cotton lint yield—Cause and effect? An on‐farm case study across three cotton‐growing regions in Australia

Abstract

AbstractThe average lint yield of irrigated cotton in Australia ranges from 2270 to 3700 kg/ha, but yields vary substantially between farms and also between fields on the same farm. Differences in soil properties may cause these yield variations. Identifying which factors are causal and what management can be implemented to mitigate the impacts should help optimize inputs and improve profits. During the 2018–2019 summer cotton‐growing season, a paired‐field comparison approach was used to investigate and improve the understanding of soil property‐induced irrigated cotton yield differences within five farms across three regions of NSW, Australia. The paired fields at each farm recorded an average lint yield difference of >284 kg/ha (measured in 2018–2019 or 5‐year average lint yield). Several soil properties differed between the paired fields at each farm comparison. The soil organic carbon stocks were higher in the higher‐yielding fields at all the farm comparisons and the normalized lint yield percentage was positively correlated with soil organic carbon stocks. Soil sodicity was higher in the lower‐yielding fields at 3 of the 5 comparisons. Results for most soil nutrient tests were above the recommended critical concentrations for Australian cotton production. A stepwise linear regression excluding soil nutrients that were above soil test critical values for crop response and below crop toxicity levels indicated the lint yield was positively correlated with SOC stocks and negatively correlated with sodicity and bulk density. No earthworms were detected during visual soil assessment or soil sampling across all the sites. Visual soil assessment was not a sensitive predictor of cotton crop performance. Comparing soil properties using a paired field approach may assist cotton growers in understanding the factors behind yield differences. A similar strip comparison approach could be adopted for within‐field variability by dividing the fields into discrete performance zones and assessing the soil properties of each zone separately.