A study on the methodolgy of cultivar evaluation based on yield trial data with special reference to winter wheat in Ontario

Loading...
Thumbnail Image

Date

Authors

Yan, Weikai

Journal Title

Journal ISSN

Volume Title

Publisher

University of Guelph

Abstract

Improvement in the methodology of yield trial data analysis is an important aspect of improving breeding efficiency. The key to effective cultivar evaluation based on yield trial data is to correctly grasp the pattern of genotype by environment (GE) interaction and to develop a selection strategy in accordance with it. One outcome of this study is the development of a "GGE biplot", which graphically displays the genotypic main effect (G) and the GE interaction contained in the multi-environment trial (MET) data. The GGE biplot can be used to effectively address the following questions. (1) Are there crossover GE interactions in the data? (2) Which cultivars are best for a given location environment? (3) Which locations are more suitable for a given cultivar? (4) Can the genotypes be meaningfully grouped? (5) Can the environments be meaningfully grouped? If multi-year MET data are available, (6) can the target environment be divided into different mega-environments? If not, (7) what are the more discriminating and representative test environments? (8) What are the higher yielding and more stable genotypes? These questions can be directly answered by examining a GGE biplot. In addition, if external information, i.e., covariables other than yield, about the genotypes and the test environments are available, the following question can also be addressed: (9) what are the traits/characteristics that make-up a superior cultivar? And (10) what are the environmental factors that make up a better test environment? Another outcome of this study is the proposal of a new measure of cultivar performance--YREM, which is the yield relative to (i.e., divided by) the environmental maximum. It is a simple and intuitive measure of cultivar performance that is relatively independent of cultivar attendance. It provides quantitative criteria for selection/culling based on data from a single or multi-location trial in a single year. Application of the GGE biplot technique and the concept of YREM to the 1989 to 1999 Ontario winter wheat performance trial data revealed the following insights into winter wheat in Ontario. (1) Crossover genotype by location (GL) interaction occurred every year; the loss of yield due to crossover GE interaction was as high as 26~40% of the attainable yield. (2) The crossover GL interaction was largely unrepeatable, however. (3) Nevertheless, the yearly GL interaction patterns suggested that the Ontario winter wheat growing region could be divided into two mega-environments: Eastern Ontario as one and Western and Southern Ontario as the other. (4) Plant height and maturity were found to be the major genotypic causes of GE interaction, while temperatures in the winter (December to March) and summer (May to July) months are the major environmental causes of the GL interaction. In Southern and Western Ontario, shorter and earlier cultivars seemed to be more adapted; but in Eastern Ontario taller and later cultivars are more favored. (5) Resistance to various diseases, especially to septoria leaf blotch, was frequently found to contribute to higher average yield, and thus should be an important breeding objective. (6) Analysis using YREM indicated that the multi-year average YREM of adapted cultivars was >0.89. The power of a single year MET is to discard genotypes with average YREM < 0.84 and to promote genotypes with average YREM > 0.94; The data of single trial can only be used to discard genotypes with YREM < 0.60~0.74.

Description

Keywords

cultivar evaluation, genotype by environment interaction, selection strategy, GGE biplot, breeding efficiency, multi-environment trial data

Citation