Contribution of legumes to wheat productivity in Mediterranean environments of central Chile

The potential benefits of using a legumes have not been evaluated in the rainfed, cropping systems of central Chile where wheat (Triticum aestivum) is typically grown in rotation with oats (Avena sativa) using high rates of nitrogen (N) fertilizer. Two experiments were initiated to compare unfertilized legume-wheat cropping sequences to the current oat–wheat system without and with applications of N fertilizer (160 or 207 kg N ha−1). One experiment was located in the interior dryland of central Chile (average annual rainfall 650 mm), and the other in the Andean foothills (average annual rainfall 1200 mm). Treatments were established in different areas of the same two experimental sites in consecutive years (2008 and 2009) resulting in four separate trials evaluating legume effects on wheat yield (2009 or 2010). Estimates of N2 fixation determined using the 15N natural abundance technique indicated that around 21 kg shoot N was fixed for every tonne (t) of shoot dry matter accumulated by either narrow-leaf lupin (Lupinus angustifolius), field peas (Pisum sativum), yellow lupin (Lupinus luteus), or white lupin (Lupinus albus) grown for grain, and by vetch (Vicia atroporpurea) grown in association with oats for green manure. When the amounts of N2 fixed were adjusted to account for N in the nodulated roots, narrow-leaf lupin was calculated to have returned the lowest average net inputs of fixed N to the system following grain harvest (26 kg N ha−1) and field pea the highest (192 kg N ha−1). By comparison vetch in the green manure treatments where all above-ground biomass were incorporated into the soil returned 67 kg fixed N ha−1. Depending upon location and year, growing grain legumes increased the subsequent N uptake by wheat by up to 60 kg N ha−1 (average 35 kg N ha−1) and vetch + oats green manure up to 103 kg N ha−1 (average 50 kg N ha−1). Wheat yields after grain legumes ranged from 2.4 to 3.0 t ha−1 in the interior dryland and 5.4–6.4 t ha−1 in the Andean foothills which were respectively 72–110% (90% average) and 69–83% (75% average) of the yields achieved by N fertilized wheat. Wheat yields after the vetch + oats green manure were 1.8 and 7.9 t ha−1 representing 60% and 102% of the yield of N fertilized wheat. In the absence of N fertilizer or legumes wheat yields after oats were around half that of the N fertilized wheat. It was concluded that grain legumes or legume-based green manures provide opportunities to greatly reduce the reliance of wheat production upon N fertilizer.

► Legumes fix large amounts of N2 and make substantial contributions to the N fertility of cropping soils in two Mediterranean environments of Chile. ► Wheat yields following grain legumes were 72–110% of the yield achieved by wheat grown following an oats crop and receiving high amounts of N fertilize. ► The use of a vetch + oats green manure in the Andean foothills enhanced wheat yield above the levels attained after grain legumes (102% cf 75% of N fertilized wheat). ► The inclusion of grain legumes in the rotation provides opportunities to greatly reduce the reliance of wheat production upon N fertilizer in the rainfed cropping systems of central Chile.