Forage Yield Precision, Experimental Design, and Cultivar Mean Separation for Alfalfa Cultivar Trials
2000
Casler, Michael D. | Undersander, Daniel J.
As alfalfa (Medicago sativa L.) cultivars have become more numerous in recent years, the issue of precison of alfalfa forage yield determinations has become more important. The objective of this study was to develop a set of recommendations for improving the precision of alfalfa cultivar forage-yield estimates. Inferences were derived from 49 alfalfa cultivar trials conducted at 13 Wisconsin locations between 1984 and 1996. Although randomized complete block designs were sometimes effective, spatial analysis offers considerable potential for improved precision. Increasing the number of replicates was expected to be more effective than increasing plot size. Trial data should be discarded only when severe and irreversible biological or physical disturbances are present. If researchers feel the need to discard data or entire trials on the basis of low statistical precision as an additional criterion, the decision should be based on (i) nonsignificant F-tests for cultivars of all individual years and for the combined over-years analysis or (ii) an unusually high mean square error relative to the trial mean or range among cultivar means. Trial data should not be rejected based on the coefficient of variation (CV). Genotype × environment interactions followed patterns based largely on expectations of edaphic and climatic environmental differences. Inherent precision of alfalfa cultivar trials did not influence cultivar rankings per se.
显示更多 [+] 显示较少 [-]