Development of the Moorepark St Gilles grass growth model (MoSt GG model): A predictive model for grass growth for pasture based systems

Abstract In humid-temperate regions grazed grass is the most economical means of feeding ruminant livestock. Grass growth is often highly variable and therefore difficult to predict. It is influenced by many factors including climatic conditions, soil type and soil nutrients. The Moorepark St. Gilles Grass Growth model (MoSt GG model) is a dynamic model developed in C++ describing daily grass growth at the paddock level. It was developed by adapting an existing grass growth model to include a nitrogen (N) component and a soil water component. The model is effective in grazing and cutting scenarios. Inputs include weather data, grazing management decisions and N fertiliser application. Outputs include daily grass growth, soil mineral N content, grass N uptake, grass N content and NO3− leaching. The MoSt GG model was evaluated against measured data using 2 years data from an experimental farm; the predicted and measured biomass for each paddock was compared. The mean root mean square prediction error (RMSPE) at the measurement level was 505 kg DM/ha. When averaged by week of year, the RMSPE was reduced to 321 kg DM/ha. The MoSt GG model was also evaluated against a range of management scenarios including N fertiliser application rate (from 0 to 650 kg N/ha per year) and defoliation management, and weather conditions. The grass growth response to N fertiliser application was, on average, 9.6 kg DM/kg N applied with a minimum response of 0.8 kg DM/kg N applied and a maximum response of 16.2 kg DM/kg N applied, which is in the range of previously published studies. The MoSt GG model responds to daily weather conditions, patterns and methods of sward defoliation, and describes daily variations in soil mineral and organic N content.