Legume N2 fixation was substantial in most of the experiments. Values exceeding 300 kg N/ha were recorded for irrigated pigeonpea and mungbean in the North West Frontier Province (NWFP), Pakistan, and for soybean in the Hills of Nepal. But more commonly, values were in the range 50-150 kg N/ha. The values for the Australian experiments were similar to those for Pakistan and Nepal.
Legumes increased yields in the cereal crops that followed. The majority of responses, relative to yields in a cereal monoculture, were in the range of 20 to 100%. Cereals also responded to inputs of fertiliser N, particularly in the low-N soils of Pakistan. Responses were commonly 100-200%.
Economic analysis, using gross margin, dominance and marginal analysis of a number of the rotation experiments, revealed large benefits of legumes - particularly chickpea and faba bean in Australia and soybean and black gram in Pakistan. The legume economic benefit reflected both its high grain price and impact on cereal yields.
The high productivity rotations incorporating legumes and fertiliser N increased soil organic fertility. For example, after 4 years of the chickpea-wheat rotation experiments in Pakistan, soil organic matter had increased between 28 and 56% through inclusion of chickpea and between 35 and 72% through inputs of fertiliser N in the wheat monoculture. Values for increases in soil total N were 22-56% (chickpea) and 66-100% (fertiliser N).
Both growth and N2 fixation models for faba bean and chickpea developed in APSIM are being refined as a result of the project's field experiments at Gatton in southeast Queensland and through verification using other data sets. A key outcome of the experimental program was that faba bean showed a greater capacity for N2 fixation than chickpea and, when soil nitrate supply far exceeded demand, spared soil nitrate. This finding has implications for APSIM's N2 fixation model.
APSIM-generated simulations of some NWFP Agricultural University data sets showed good agreement with observed data for wheat (-N) and mungbean, and reasonable predictions for responses of the wheat to both legumes and fertiliser N. APSIM proved an excellent tool for scenario analysis (e.g. fertiliser N rates, residue management) in such systems.
The xylem solute method for measuring legume N2 fixation was calibrated for cowpea, mungbean and black gram, adding to the list of calibrated species. Sampling procedures for the method have been simplified - previously 7-8 samplings were recommended, now it is 1-2.
Surveys of 107 farmer crops of soybean, lentil, chickpea, mash bean, pigeonpea, grasspea, faba bean and ground nut in Nepal showed that percentage nitrogen derived from N2 fixation values were almost always >50%, with generally higher values for the winter species than for the summer legumes. Total N fixed was primarily influenced by crop growth. Researchers estimated that about 30,000 tonnes N is fixed annually in Nepal by legumes, valued at A$30 million.
Quality assurance was introduced into the laboratory operations of the three Pakistani institutes. A report of the NWFP group, which may well represent all groups, noted that the ACIAR project had played a major role in the capacity building of its institutions. The research team was able to establish appropriate lab and field facilities for research, and to conduct training courses/workshops in Peshawar for researchers and students.