Calibration and performance evaluation of the STICS crop model for simulating timothy growth and nutritive value

• First calibration of STICS model for timothy.
• Equations used in CATIMO model for calculating nutritive value indices successfully used and calibrated in STICS.
• Good overall model performance (LAI, biomass and nutritive value) for various sites and climatic years.
• Better model performance for simulating the primary growth than the regrowth.

Several models for simulating the growth and nutritive value of forage crops have been developed in past decades but most of them are specific to one crop species and focus only on agronomic outputs. In this study, we calibrated and evaluated the performance of the STICS crop model for the simulation of the growth and nutritive value (cell wall concentration and digestibility, in vitro true dry matter [DM] digestibility, and N concentration) of timothy (Phleum pratense L.) under Canadian growing conditions. Equations used for calculating nutritive value attributes were derived from the CATIMO model (CAnadian TImothy MOdel). The calibration included two datasets, one for eastern Canada (cultivar Champ) and one for western Canada (cultivar Climax). Very good agreement between simulations and measurements was obtained for the calibration dataset (RMSE% ≤10% for all variables). Twenty-two datasets including leaf area index (LAI), forage DM yield, N concentration, cell wall concentration and digestibility, and in vitro true DM digestibility were then used to evaluate model performance. The model simulated LAI, forage DM yield, nitrogen concentration and nutritive value of the primary growth fairly well for all sites (RMSE% <30%), but was less successful for the regrowth (i.e. average RMSE of forage DM: 66.5 and 85.9 g m−2 for primary growth and regrowth, respectively). The next step in this research will be to perform, multi-annual simulations to evaluate crop reserve simulations and improve regrowth simulation.