Strip intercropping increases land use efficiency but the effect on water use efficiency is less well-known. Here we develop a modelling method to simulate the growth of an intercrop taking into account the acquisition of light and water by the component species in order to calculate the efficacy of light and water acquisition in an intercropping system as compared to sole crops. The model is parameterized, calibrated and validated using data on wheat-maize intercropping in Gansu province, Northwest China. Observed above-ground biomass was 1630 g m−2 for sole wheat and 1334 g m−2 for intercropped wheat while it was 3023 g m−2 for sole maize and 2259 g m−2 for intercropped maize. The average water use was 405 mm in sole wheat, 595 mm in sole maize and 711 mm in wheat-maize intercropping. Based on observed yields, the land equivalent ratio (LER) was 1.59 and the water equivalent ratio (WER) was 1.14, where LER and WER express the relative amounts of land and water needed to achieve the yields obtained in a unit area of intercrop using sole crops. These results indicate that relay strip intercropping of wheat and maize achieves an increase of land use efficiency of 59% and of water use efficiency of 14%. Overall the intercrop model gave satisfactory predictions, with coefficients of efficiency (CE) in validation of 0.86–0.97, 0.90–0.95, 0.85–0.91 and 0.98 for biomass of sole wheat, sole maize, intercropped wheat, and intercropped maize, respectively. Overall CE of water use was 0.95. Simulated LER and WER were similar to observed LER and WER. The results show that intercropping could be used to obtain more yield on less land with less water. Policies that limit water use per unit land and prohibit the use of intercropping on the basis of its high water use per unit area may therefore be counter-productive. The model for intercrop growth and development under water limitation may be used for exploring production possibilities under land and water constraints.
- Water use efficiency