Premium
Mega‐environment Analysis and Test Location Evaluation Based on Unbalanced Multiyear Data
Author(s) -
Yan Weikai
Publication year - 2015
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/cropsci2014.03.0203
Subject(s) - biplot , variety (cybernetics) , statistics , mega , computer science , biology , data mining , mathematics , genotype , genetics , physics , astronomy , gene
Mega‐environment analysis and test location evaluation are two important issues for effective crop variety evaluation through multilocation variety trials. These must be done based on multiyear multilocation variety‐trial data, which are usually highly unbalanced. This paper presents a new graphical approach for conducting mega‐environment analysis and test location evaluation utilizing unbalanced multiyear variety trial data. It consists of three steps: (i) generating a G (genotypic main effect) plus GE (genotype × environment interaction), or GGE, biplot using a missing‐value estimation procedure and treating each location–year combination (trial) as an environment; (ii) summarizing the interrelations among test locations (L) in a GGL + GGE biplot, which is the same GGE biplot imposed with the test locations. The placement of a test location in the biplot is defined by the coordinates of all environments at the location; and (iii) summarizing any subregion (S) (i.e., mega‐environment) differentiation revealed in Step 2 in a GGS biplot, which is the same GGE biplot imposed with the subregions. The placement of a subregion in the biplot is defined by the coordinates of all environments in the subregion. The same GGL + GGE biplot can also be used to visualize the ability and stability of each test location to represent a target mega‐environment. Yield data from the 2006–2012 Quebec oat ( Avena sativa L.) registration and recommendation trials were analyzed as a demonstration.