Jump to Main Content
Evaluation of the agricultural production systems simulator simulating Lucerne and annual ryegrass dry matter yield in the Argentine Pampas and south-eastern Australia
- Ojeda, J.J., Pembleton, K.G., Islam, M.R., Agnusdei, M.G., Garcia, S.C.
- Agricultural systems 2016 v.143 pp. 61-75
- Lolium multiflorum, Medicago sativa, alfalfa, correlation, crop models, data collection, dry matter accumulation, environmental management, farming systems, forage crops, growth and development, livestock, pastures, phenology, plant growth, production technology, range management, soil types, transpiration, weeds, Argentina, Australia
- Modelling plant growth provides a tool for evaluating interactions between environment and management of forage crops for pasture-based livestock systems. Consequently, biophysical and farm systems models are becoming important tools for studying production systems that are based on forage crops. The Agricultural Production Systems Simulator (APSIM) is a model with the potential to compare the growth of annual forage crops and perennial pastures. However, information is limited about how accurately the Lucerne and Weed modules represent the growth and development of forage crops and pastures under different managements, soil types and environments in South America. This study evaluated the capacity of APSIM to simulate the growth rates and predict the dry matter (DM) yield of Lucerne (Medicago sativa L.) and annual ryegrass (Lolium multiflorum Lam.) in contrasting climatic regions of Argentina. In addition, at several Australian locations, DM yields of both crops were simulated to ensure that possible changes to the model not interfere with the robust APSIM performance that was already shown in south-eastern Australia. Initial simulations for Lucerne and ryegrass were made with original Lucerne and Weed modules of APSIM, respectively. Simulated DM yield was then compared with field data collected from the same crops grown in five locations in the Argentine Pampas and seven locations in south-eastern Australia over 5 of years. APSIM predicted DM yield of Lucerne at each harvest with reasonable accuracy [0.59, 0.77 and 0.77 for R2, correlation coefficient and concordance correlation coefficient (CCC), respectively]. However, these statistics improved when the DM yield was analysed by annual accumulation, with values of 0.87, 0.93 and 0.92 for R2, correlation coefficient and CCC, respectively. APSIM, generally, over-predicted DM yield of annual ryegrass at the first harvest. Nonetheless, when the Weed module was modified through changes in phenology and transpiration efficiency, performance improved (values of 0.89, 0.94 and 0.93 for R2, correlation coefficient and CCC, respectively). This study showed that annual DM yield of Lucerne can be successfully modelled by the APSIM Lucerne module without any modifications, using a crop modelling approach. However, successfully modelling of Lucerne DM yield by harvest will require further development of the model. Moreover, modification of model parameters associated with phenology and transpiration was required to enable the Weed module of APSIM simulate growth and yield of annual ryegrass in a range of geographic locations within the Argentine Pampas.