Effect of amount and quality of previous crop residues on the nitrogen fertiliser response of a wheat crop

Abstract

Burr medic (Medicago polymorpha cv. Circle Valley) at 3 sowing rates (3, 12, and 50 kg/ha), narrow leaf lupins (Lupinus angustifolius cv. Illyarrie) at 2 sowing rates (30 and 100 kg/ha, with the high rate sown early or late) and oats (Avena sativa cv. Winjardie) at 2 sowing rates (20 and 60 kg/ha, with the high rate sown early or late), were planted on a deep, infertile siliceous sand at Badgingarra in 1988, to produce residues of varying quantity and quality. Residues produced (minus seed) varied in quantity from 0.70 t/ha (medic 3 kg/ha) to 4.22 t/ha (lupins 100 kg/ha sown early) when measured in late summer. Carbon to nitrogen ratios (C/N) ranged from 37.3 to 58.5 (medic), 28.3 to 40.4 (lupins), and 78.6 to 112.0 (oats). Seed present in the residues ranged from 0.18 to 0.47 t/ha (medic) and from 0.32 to 0.57 t/ha (lupins). The ranges in C/N values were 8.3-9.3 (medic) and 9.4-10.0 (lupin). There was no grain in the oats residues. The residues were incorporated into the soil during the sowing operation of a wheat crop in 1989. Subplots of 6 rates of N as ammonium nitrate (0, 14, 27, 54, 82, and 163 kg N/ha) were included in each block. Wheat dry matter and grain yields were higher following lupins and medic than following oats, and although there were large responses to N fertiliser following all 1988 crops, the N application had little effect on the difference between oats and legumes. In the absence of N fertiliser, yields and N uptake by the 1989 wheat crop were higher following lupins than following medic, due to hardseededness of medic seeds. There were no significant differences within 1988 crop groups. There were good correlations between dry matter and grain yield of wheat without N fertiliser and C/N value of the residues. The correlations were better when seed was not included in the calculation (r2 = 0.86 and 0.80, respectively) than when the seed was included (r2 = 0.79 and 0.64). Similar correlations were obtained between dry matter and grain yields and total N content of the residues. Correlations were lower when seed was omitted (r2 = 0.64 and 0.75) than when it was included (r2 = 0.70 and 0.81). Yields decreased with increasing C/N and increased with total N content of the residues. Wheat grain N concentrations fell with the first increment of N fertiliser, which produced a large yield increase, but rose to high levels at high rates of N.