An integrated approach to model the establishment, water use and growth of new perennial pasture species

Perennial pastures increase the sustainability of Australian cropping systems by increasing out-of season water use, improving ground cover and helping to avoid dryland salinity. Medicago sativa (lucerne) is the most common perennial legume pasture and is adapted to a wide range of environments, but it does not persist well in areas with acid soils or low summer rainfall (Cocks 2001). Under these scenarios, native herbaceous perennial legumes, such as Cullen australasicum (Cocks 2001), and the exotic perennial pasture legume Bituminaria bituminosa var. albomarginata (common name albo-tedera) are potential alternatives to the currently limited perennial forage options. However, the growth and adaptability of these legumes to high-input Australian agricultural systems are likely to differ from currently used pasture species and their pasture potential has not been widely studied, tested or predicted. This paper describes the modelling of three aspects of the performance of these species, with the overall aim of predicting their performance across Australian agricultural areas. First, we quantified the early performance of these legumes, including early establishment and first summer survival, at three locations in the wheatbelt of Western Australia (WA): Buntine, Merredin and Newdegate. Lucerne was used as a reference. Initial analysis showed significant variability in their early adaptability and Principal Component Logistic Regression (PCLR) modelling revealed that temperature and a number of soil parameters (moisture, pH, available sulfur, available aluminium, conductivity and organic carbon) were all important in determining the early establishment and summer survival of these legumes under field conditions. Second, a mechanistic physiological growth model was developed to estimate the daily net photosynthesis based on light decay down the canopy (Thornley 2002). Respiration and senescence rates were incorporated in the model to estimate the daily biomass loss and thus daily net dry matter production was estimated. The influence of drought on leaf expansion and senescence was accounted for through the changes in soil water potential. Thermal time and soil water potential were used to simulate the leaf area index development under drought. Finally, crop phenology and growth parameters measured for the two species under a diverse range of soil and climatic conditions are being used to modify the existing Agricultural Production Systems Simulator (APSIM) lucerne module to simulate B. bituminosa var. albomarginata and C. australasicum growth and production in APSIM. Once the models have been validated, they will be used to predict the performance of C. australasicum and B. bituminosa var. albomarginata under a wide range of environmental conditions in order to identify potentially suitable regions in WA, and across the country, for further trials and eventual adoption of these perennial pasture legumes. In addition, simulations will be done to predict future performance under expected climate change scenarios.