| dc.description.abstract |
The responses of wheat to various rates of application of nitrogen fertiliser were compared following lupins (WL) or wheat (WW) in the previous year. Results covered 10 sites and trials were carried out between 1979 and 1984. giving 26 site-year comparisons. The trials were on sandy or duplex (sand over clay) soils. The overall yield of WL was greater than WW on 21 occasions (significant in 10 cases, P < 0.05), less than WW on 2 occasions (both significantly) and there was no difference on 3 occasions. When no nitrogen was applied the advantage of WL was 41% (WL 1.20 t/ha and WW 0.85 t/ha). The response to nitrogen differed between trials; over all trials a quadratic model best described the responses. There were significant interactions between rotation and nitrogen rate in 10 comparisons. In 4 of these cases, response to nitrogen with WW was greater than with WL and these response curves approached I another but did not meet, indicating that both residual nitrogen from the lupins and some other benefit from the lupins were involved. In 5 cases the greater response on WW resulted in convergence with the WL response curve, suggesting that residual nitrogen explained all of the benefit of lupins. In these cases the amount ofnitrogen fertiliser required to bring the yield of WW to that of the WL without nitrogen ranged from 20 to 47 kg N/ha, with a mean of 37 kg N/ha. Parallel response curves were observed In 12 cases, indicating involvement of factors other than residual nitrogen in the response to lupins, e.g. disease cleaning effects or slow release of nitrogen throughout the season. The type of response could not be related to particular site characteristics. When quadratic coefficients were averaged to give 'average response curves', similar rates of applied nitrogen were required for maximum yields in both WW and WL, and the maximum yields were 1.23 t/ha for WW and 1.41 t/ha for WL. The average advantage of WL over WW, in the absence of nitrogen fertiliser. was 350 kg/ha. |
|