Premium
Moving‐Mean Analysis of Field Tests for Iron Efficiency of Soybean
Author(s) -
Diers B. W.,
Voss B. K.,
Fehr W. R.
Publication year - 1991
Publication title -
crop science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.76
H-Index - 147
eISSN - 1435-0653
pISSN - 0011-183X
DOI - 10.2135/cropsci1991.0011183x003100010013x
Subject(s) - statistics , replication (statistics) , selection (genetic algorithm) , biology , analysis of variance , analysis of covariance , mathematics , covariance , microbiology and biotechnology , veterinary medicine , zoology , computer science , medicine , artificial intelligence
Soil heterogeneity reduces the effectiveness of selection of soybean [ Glycine max (L.) Merr.] genotypes for Fe efficiency on calcareous soils. A study was conducted to determine if the moving‐mean analysis procedure would reduce the error variance in Fe‐efficiency tests and improve selection among genotypes. Chlorosiscores were obtained from the Iowa Soybean Yield Test (ISYT), which included 620 entries replication −1 in 1987 and 660 entries replication −1 in 1988, and from the Uniform Soybean Test (UST), which included 80 entries replication −1 in both years. The data from the tests were analyzed using an analysis of covuriance in which the eovariate for each plot was the mean of its adjacent plots. The use of the analysis of covariance reduced the error variance for the ISYT by 33% in 1987 and 14% in 1988, and for the UST by 21% in 1987 and 7% in 1988, compared with a conventional analysis of variance (ANOVA). Gains from selection were calculated for selection among entries common to both years in the ISYT by using moving‐mean adjusted and unadjusted entry means. No significant differences ( P > 0.05) between the two types of means were obtained for gains from selection. These results show that, although the moving‐mean analysis reduced the error variance, it did not adjust means in a way that would have resulted in improved selection compared with unadjusted means.