Nitrogen release patterns from decomposing shoot residues of Tephrosia candida, Crotalaria grahamiana, Mucuna pruriens, Macrotyloma axillare, Macroptillium atropurpureum and Desmodium intortum were studied in the laboratory for a period of 22 weeks in a sandy clay soil and 10 weeks in a clay soil using a leaching tube technique. The residual effect of soil incorporated shoot residues of T. candida, T. vogelii, C. grahamiana, M. pruriens and C. juncea on maize yield was evaluated at four sites each in the high rainfall zone (mean precipitation 2100 mm year¿1) and low rainfall zone (mean precipitation 800 mm year¿1) of Bukoba District, Tanzania. N mineralised from the legume residues ranged from 24 to 61% of the initial N after 22 weeks in a sandy clay soil and ¿1 to 34% after 10 weeks in a clay soil. The N mineralisation rates of the residues decreased in both soils in the order M. atropurpureum>M. axillare>C. grahamiana>D. intortum>T.¿candida>M. pruriens and were mostly strongly related to (polyphenols+lignin)-to-N ratio, lignin-to-N ratio and lignin. Relative to the control, legume residues resulted in two and threefold increase in maize grain yield i.e. from 1.1 to 3.2 Mg ha¿1 and from 1.4 to 3.8 Mg ha¿1 in a high and low rainfall zone respectively. However, maize yield response to legume residues was limited when compared with application of 50 kg N ha¿1 of mineral fertiliser. The % fertiliser equivalency (%FE) of legumes ranged between 25 and 59% with higher values recorded with C. grahamiana. At harvest, apparent N recoveries in maize ranged between 23 and 73% of the N applied in the legume residues. Highest recovery was found with application of C. grahamiana and least recovery from T. candida residues. These results suggested that application of legume residues alone might not be sufficient to meet N requirements and to achieve the yield potential of maize crop in Bukoba soils unless supplemented with small doses of mineral fertilisers.
- alley cropping system