Soil mining is a major problem of low technology farming and replenishing such loses is one of the problems farmers face. Trials conducted in 2004 and 2005 at Samaru, (11011’N, 07038’E and 686m a.s.l) in the northern Guinea savanna ecology of Nigeria, estimated the response of maize (Zea mays L.) and soyabean (Glycine max (L.) Merr.) intercrops to green manure from the soyabean companion crop at various nitrogen fertility levels. Treatments were a factorial combination of four maize/soyabean mixtures in replacement and additive mixtures as well as their sole crops and four nitrogen rates arranged in randomized complete block (RCBD) in three replicates. Treatments improved the soil’s physical and chemical (soil C: 2 – 10 %; total N: over 100%; CEC: 20-80%) properties in 2005. Maize in replacement mixture had heavier cobs and 100-seed weight than in additive mixtures; while two plants excelled three plants per stand in these yield components and grain yield per hectare. Soyabean in additive mixture had heavier pods and 100-seed weight and seed yield per hectare than in replacement mixture. Crops in three row plots had significantly heavier pods, 100-seed weight, and seed yield than those in two row plots. Yield and yield parameters of maize increased significantly with increase in the rate of N applied while for soyabeans the 60 and 90 kg N ha−1 were similar in yield and yield parameters. Mixtures gave higher than expected LER values of mixed crops. Maize LER values at 60 and 90 kg N ha−1 were higher than 1.00, making soyabean a bonus crop. Total LER for the crops showed that crops in mixtures performed better than combined sole crops at 90 kg N ha−1. Agronomic efficiency was highest at 12:2(A) cropping pattern and 30 kg N ha−1, while replacement mixtures had higher efficiency values at low N application and the 12:2(A) had higher efficiency value at higher N rates.
