A lack of complementarity for water acquisition limits yield advantage of oats/vetch intercropping in a semi-arid condition

Oats (Avena sativa L.) and hairy vetch (Vicia villosa) are well adapted crop species for production in semi-arid environments, such as in Inner Mongolia, China, where due to variable rainfall, farmers do not apply fertilizer. We hypothesized that the use of a mixture of a cereal and a legume could enhance yields under these low input conditions, because integrating an N-fixing legume in the system could mitigate N limitation for the cereal and enhance its growth. A nine-year (2008–2016) field experiment was set up with three treatments: sole oats, sole vetch and oats/vetch strip intercropping. These cropping systems were grown continuously in the same plots, to allow accrual of long-term effects. Yields and water use were quantified in years 7–9 of the experiment (2014 to 2016). With a 50/50 ratio of the area sown to the two species, the intercropped oats had a relative yield of 0.59 and intercropped vetch had a relative yield of 0.45. Oats was the dominant crop characterized by a relative yield per plant of 1.18, compared to a relative yield per plant of vetch of 0.89. However, the land equivalent ratio (LER), expressing the comparative efficiency of land use in intercropping, and the water equivalent ratio (WER), the comparative system level water use efficiency of the intercrop relative to sole crops, were both not significantly different from one. Thus we reject the hypothesis that oat/vetch intercropping increases land productivity and water use efficiency. From differences in results in years with more rainfall and years with less rainfall, we infer that yields of both species are mostly limited by water availability. On average over the three years, the yield disadvantage of vetch was fully compensated by the yield advantage of oats, due to a lack of complementarity for water acquisition. This conclusion can be generalized to the testable prediction that species selection for productive intercropping should focus on achieving complementarity for traits that interact with the factor most constraining productivity, which was rainfall in this particular crop system under the conditions of the study.