Simulating biomass accumulation and yield of yam (Dioscorea alata) in the Upper Ouémé Basin (Benin Republic)- I. Compilation of physiological parameters and calibration at the field scale

In order to evaluate the potential of yam (Dioscorea alata) production to meet future demand, accurate estimations of crop growth and yield by simulation models is required. The Environmental Policy Integrated Climate (EPIC) model simulates the effects of weather, soil characteristics, tillage, fertilization, irrigation and other management practices on crops. The present study was conducted to build both a crop parameter file for yam (D. alata), and to calibrate the EPIC model to simulate the growth and yield of yam under different management practices.After calibration, the model accurately simulated the effect of fertilizer on yam yield, as indicated by relatively low simulation mean residual errors (ME), ranging from 0.15 to 0.40 Mg ha−1 with a Mean absolute error (MR) ranging from 4.3 to 9.7%. Overall, there was a good correlation between simulated and observed mean yields during 2005 and 2006 (R2 = 0.69, p < 0.01). We conclude that the EPIC model can simulate the total biomass of yam with a relative accuracy acceptable for many applications. Its accuracy could be improved by a more detailed modeling of phonological development. However, for regional scale application, the models accuracy appears to be sufficient.