Assessing the Potential for Zone‐Specific Management of Cereals in Low‐Rainfall South‐Eastern Australia: Combining On‐Farm Results and Simulation Analysis. (14th January 2016)

• Record Type: Journal Article
• Authors: Hoffmann, M. P.; Llewellyn, R. S.; Davoren, C. W.; Whitbread, A. M.
• Is Part Of: Journal of agronomy and crop science. Volume 203:Number 1(2017)
• Page Start: 14
• Page End: 28
• Publication Date: 2016-01-14
• Subjects: APSIM -- plant available water capacity -- precision agriculture; Agronomy -- Periodicals; Crop science -- Periodicals; 630.5
• DOI: 10.1111/jac.12159
• Languages: English
• ISSNs: 0931-2250

Abstract

Abstract: In the low‐rainfall region of south‐eastern Australia, distinctive soil types reflecting the typical landscape of higher elevated dunes and swale zones at the bottom can be found within one field. Different soil characteristics cause consequently large variability in cropping productivity between soils and across seasons. To assess the possibilities for zone‐specific management, five farmer fields were zoned into a dune, mid‐slope and swale zone. For each site, zone yields were mapped over 2 years and soil properties were surveyed. This information was used to parameterize and validate the APSIM model for each zone. Field‐measured PAWC increased from the dune to the swale zone. On‐farm results and simulation analysis showed distinctive yield performance of the three designed zones. However, yield is not related to PAWC, it is rather a complex relationship between soil type, fertility and rainfall. While in high‐rainfall years, the swale zones yielded higher due to higher soil organic carbon content and less drainage losses, the dune zones performed better in the low‐rainfall years due to lower evaporation losses. This study emphasizes that in this specific environment where soil variation in texture and subsoil constraints strongly influence crop performance, mechanistic crop models and long‐term field observations are necessary for better understanding of zone‐specific performance, and simple linear relationships across years or sites are not useful.