Relationships between maize yield and growth parameters in a long-term fertilization experiment.

摘要

In a long-term maize monoculture experiment set up on the active ingredient equivalence principle, changes in the yield components were investigated over a period of three years (2005-2007) as a function of the fertiliser treatments, and the values of the growth parameters HI, LAI, NAR and CGR were calculated using the classical method of growth analysis. The results indicated that optimum N supplies and the year effect made a substantial contribution both to the grain number per ear and to the thousand-kernel weight. In the course of correlation analysis, both Pearson's correlation coefficient and multiple regression analysis demonstrated that the grain yield was in close positive correlation with these yield components, and with the maximum value of dry matter production and the harvest index. The two yield components explained 76% of the grain yield, and the effect of thousand-kernel weight was around 3.75 times as great as that of the grain number per ear ( beta =0.721 vs. 0.192). On the basis of partial correlation analysis, the maximum value of total dry matter and the thousand-kernel weight were jointly responsible for around 60% of the variance in maize grain yield. Analysis using the "Enter" method showed that the two yield components explained 62% and 59% of the grain yield in wet years (R2₂₀₀₅=62.3%; R2₂₀₀₆=58.8%), while in the dry year neither the thousand-kernel weight nor the grain number per ear had a significant effect on the yield (R2₂₀₀₇=4.5%).