|dc.description.abstract||In a large field experiment on an irrigated Templeton silt loam (Udic Ustochrept) soil using a completely randomised block design with six replicates (each main plot = 30 x 20m), main plot treatments of wheat and ryegrass grown for seed produced more dry matter (17-23 t cf. 16-17 t ha⁻¹) but less nitrogen (N) (130-230 kg N cf. 387-427 kg N ha⁻¹) than peas and white clover for seed. The application of ¹⁵N tracer to leguminous crops and ¹⁵N-labelled fertiliser to non-leguminous crops in micro-plots allowed the determination of biological N₂ fixation (BNF) and fertiliser N recovery, respectively. Significant proportions of N accumulated in white clover (90%) and peas (69%) was attributed to BNF. Apart from providing a cash crop in the short-term, white clover and peas fixed 327 and 286 kg N ha⁻¹ respectively. On the other hand, between 41-52% of the applied fertiliser N was recovered by ryegrass and wheat crops.
In the second year, residues from these main plot treatments and ¹⁵N-labelled residues in micro-plots were given subplot treatments of ploughing-in, burning, rotary hoeing or left as surface mulch in a split-plot design and their effects were measured using winter wheat as a test crop. In the third and fourth year sequential field experiments, all plots were ploughed after removing the residues of previous wheat crop and test crops of wheat were sown each year.
In the first sequential wheat crop results showed no significant interaction between crop residue and subplot treatments on grain yields. Grain yields followed the order: white clover > peas > ryegrass > wheat. For residue management grain yields of ploughed, rotary hoed and burned treatments were higher than that of mulched treatment. More than 80% of variations in grain yield were related to differences in the amount of N present in residues and the established plant populations. In the second sequential wheat crop, significantly lower grain yield was obtained under wheat residue compared with other residues and the mulched treatment produced significantly higher grain yield than other treatments. In the third sequential crop, grain yield was not affected by management treatments but ryegrass and wheat residues produced higher grain yields than those of leguminous residues. Differences in grain yields obtained in the second and third sequential wheat crops were explained by either a slow decomposition of the mulched residues or a slow release of the N immobilised in the non-leguminous crop residue of the first year experiments probably due to their relatively high C/N ratio.
Nitrogen benefits of crop residues to subsequent wheat crops were estimated as proportion of N accumulated in wheat crop derived from crop residue-N. Of the total N taken up by the first sequential wheat crop, leguminous residues supplied between 29-57% of wheat N accumulated. While non-leguminous crop residues provided only 6-10%. These proportions decreased with successive sequential crops. In the first sequential wheat crop, management treatments provided no significant differences in N benefits from non-leguminous residues while rotary hoeing reduced N benefits of white clover residue-N. These effects were not shown in subsequent sequential wheat crops.
Nitrogen losses from crop residues were estimated after the second sequential wheat crop. Results obtained showed relatively higher proportions of leguminous residue-N were lost compared with those of non-leguminous residue-N. Also higher proportions of leguminous residue-N were lost when these residues were either ploughed, rotary hoed or burned compared with mulched treatments (e.g. clover residue, 36-40% vs 21-23%). Greater proportions of non-leguminous residue-N were lost when these residues were burned (33- 44%) compared with those under ploughed, rotary hoed or mulched treatments (20-27 %).
Estimated gross N balances for the four years of cropping (1993 to 1997) showed the importance of BNF by white clover in supplying N to the cropping system. Higher amounts of N fixed by white clover than peas provided a positive N balance for first two sequential wheat crops. In contrast, N balances were negative for all sequential wheat crops for peas and other antecedent non-leguminous residues. In general, residue burned treatment showed significantly higher negative N balances than other management treatments studied.
In addition, effects of crop residue and management on residue decomposition and some soil properties (pH, microbial biomass C and N, soil carbohydrates and soil N mineralisation) were determined. Residue decomposition and changes in soil properties were related to crop residue quality and their management practices.
Practical implications of the results obtained from this study for farmers and environmental managers were presented and discussed.||en