Premium
Procedure for Adjusting the Yield of Plots of Burley Tobacco (Nicotiana tabacum L.) for Differential Stands 1
Author(s) -
Gupton C. L.,
Archer L. E.
Publication year - 1973
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj1973.00021962006500010031x
Subject(s) - yield (engineering) , nicotiana tabacum , mathematics , plot (graphics) , statistics , compensation (psychology) , tobacco leaf , agronomy , horticulture , biology , agricultural engineering , engineering , biochemistry , materials science , gene , metallurgy , psychology , psychoanalysis
During 1968 and 1969 we conducted a study to determine whether the precision of yield estimates could be improved by adjusting plot yields for differences in stand of burley tobacco ( Nicotiana tabacum L.). Plants were removed from the plots at four different stages of growth and in five different patterns. Both the stage and pattern of plant removal had significant effects on the plot yield of cured tobacco. To derive an adjustment procedure for missing plants, we estimated the amount of compensation resulting from each stage and pattern of plant removal as the difference between the actual yield per plot and the expected yield, assuming no compensation. The proportion of the check plant average compensated for in a given pattern was computed as the ratio of the estimated amount of compensation to the average yield per check plant. A direct adjustment of plot yields to a full stand basis is possible after adding the equivalent number of plants resulting from compensation to the stand count. The efficacy of our procedure was determined by applying it to data from a yield and quality trial that had a poor stand in 1971. Relative efficiencies were 100, 129, 139, and 149 for no adjustment and adjustment by Crew‘s procedure, covariance analysis, and our procedure, respectively. To determine the nature of compensation by plants adjacent to missing plants, we computed standard partial regression coefficients, which were 0.2437 for plant height, 0.0180 for length of leaf, and 0.5972 for width of leaf. Plant height appears to exert a small influence on compensation, but the major influence is apparently the leaf width component of leaf size.