Extrapolation domain approach, based on a matrix of three categorical variables (growing degree days, aridity index and temperature seasonality) was used to define agro-climatic zones. Based on crop harvested area distribution and the agro-climatic zones, a total of 15 (maize) and 16 (soybean and wheat) reference weather stations (RWS) were selected, accounting for 78, 81, and 70% of national area for these crops, respectively. 

Crops simulations were performed using locally calibrated and validated DSSAT V4.5 crop models. Management practices for each RWS buffer zone were retrieved from local agronomists. Dominant soil series were identified for each RWS buffer based on data provided by the Soil Division of INTA. The simulated crop rotations were: 2-y soybean-maize, and 2-y soybean-wheat-late soybean. Separate simulations were performed to estimate Yp and Yw. Simulations assumed no limitations to crop growth by nutrients. Yield gap (Yg) was calculated as the average difference between Yw or Yp and Ya (from Agricultural Ministry of Argentina), based solely on the last 7 harvest years (2006-2012).

National rainfed maize, soybean, and wheat yield potential is 11.7, 3.9, and 4.8 t/ha, current farm yield is 6.9, 2.6, 3.1 t/ha, which is 59, 68, and 64% of potential yield, respectively. Average yield gap is 4.8, 1.2, and 1.7 t/ha for maize, soybean, and wheat, respectively.