Open AccessDevelopmental stability in oats ( A vena sativa L.): II. Quality characters
Author(s) -
GULLORD MAGNE,
AASTVEIT ARE HALVOR
Publication year - 1987
Publication title -
hereditas
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.819
H-Index - 50
eISSN - 1601-5223
pISSN - 0018-0661
DOI - 10.1111/j.1601-5223.1987.tb00270.x
Subject(s) - husk , avena , test weight , biology , principal component analysis , stability (learning theory) , cultivar , linear regression , mathematics , horticulture , agronomy , statistics , botany , machine learning , computer science
G ullord , M. and A astveit , A.H. 1987. Developmental stability in oats ( Avena sativa L.). II. Quality characters. — Hereditas 107: 65–74. Lund, Sweden. ISSN 0018–0661. Received October 31, 1986 The present paper deals with developmental stability of quality characters of Nordic oat lines and cultivars over years and locations. In one series during the years 1981–84 highly significant interactions between genotypes and locations were found for husk content, test weight and protein content, but not for 1000 grain weight and oil content. The interaction component of variation has been split up into ecovalences. linear regressions and deviations from linearity, and principal components for the characters with significant genotype location interaction. The main results are:1) The linear regression model did not fit the data sufficiently well for any of the quality characters. However, there seems to be good agreement between the parameter for deviation from regression, Wricke's ecovalence, the principal components and the PLS parameters. 2) The PLS procedure for husk content, test weight and protein content divided the cereal growing area into roughly the same regions as for grain yield. 3) Temperature and rainfall, especially between heading and maturity. had large influence on the genotype x location interactions for husk content, test weight and protein content. 4) There was a large genetic variation in developmental stability for husk content, test weight and protein content. 5). Developmental stability in one character could be associated with stability in other characters but not necessarily so. 6) There was no close genotypic relationship between the quality characters and stability.