Cover crops in dry regions have been often limited by low nutrient and water-use efficiency. This study was conducted during 3.5 years to determine the effect of replacing bare fallow by a cover crop on yield, N uptake, and fate of labeled fertiliser in an intensive maize production system. Three treatments were studied: barley ( Hordeum vulgare L.), vetch ( Vicia villosa L.) and bare fallow during the intercropping period of maize ( Zea mays L.). All treatments were irrigated and fertilised following the same procedure, and a microplot in each plot was established with 210 kg N ha -1 of double labeled ammonium nitrate. Crop yield and N uptake, soil mineral N (N min), and recovery of 15N in plant and soil were determined after maize harvest and killing the cover crop. Replacing bare fallow with cover crops did not affect subsequent maize yield but affected N uptake. Vetch increased N supply by legume residues after the second year, and the N content in grain by the third. Nitrogen recover from fertiliser was not affected by treatment and averaged 46%. Barley recovered more 15N during the autumn-winter period than vetch or fallow. Under representative conditions, average barley N content was 47, vetch 51, and spontaneous vegetation content 0.8 kg N ha -1. Recovery of 15N in barley comprised 19% of total N content in aerial biomass, while only 4% in vetch. Vetch enhanced soil 15N recovery more than other treatments, suggesting its presence in a fairly stable organic fraction unavailable for maize uptake or lost. Replacing bare fallow by a cover crop only reduced fertiliser losses in a year with abundant precipitation. Nevertheless, reduction in soil N min in vetch and bare fallow treatments was similar, showing that N losses can be reduced in this cropping system, either by replace bare fallow with barley or smaller N fertiliser application to maize.
