Maximizing genetic gain II
Walter R. Fehr and Walter P. Suza
Readings:
- Chapter 18 [PDF], Principles of Cultivar Development. Vol. 1: Theory and Technique, by Walter R. Fehr (Access the full book)
Introduction
The emphasis of the lesson is on the role of genotype x environment interaction on the effectiveness of selection for quantitative traits. The importance of this interaction is highly dependent on the trait under selection. For example, the relative difference among genotypes for the number of days from planting to maturity is much more consistent among environments than the differences among genotypes for seed or forage yield. As a result, the genotype x environment component in the genetic gain equation discussed in the previous chapter is smaller for days to maturity than for yield. Therefore, a breeder can use fewer environments to obtain reliable values for the maturity of a genotype than are necessary to determine the genetic potential of a genotype for yield.
From the farmers’ perspective, performance data in one season is used to select cultivars for planting the next season. When a trait has a low genotype x environment interaction, the relative differences among cultivars in one season will be similar in the coming season. For example, a cultivar that matures 10 days earlier than another in one season would be expected to mature earlier in the coming season, although not necessarily by 10 days. For yield, one cultivar that is higher yielding than a second cultivar in one season may be lower yielding than the second cultivar in the next season.
In conducting research to determine the importance of genotype x environment interaction in a breeding program, a significant interaction may be found in an analysis of variance. Whether or not the interaction is of importance to the breeder depends on which of the types (illustrated in chapter 18 of Principles of Cultivar Development) are involved. The most difficult interaction to deal with is one that results when the best genotypes in one environment perform less well than others in another environment. When this type of interaction occurs, the breeder generally uses less stringent selection in one environment when deciding which genotypes to advance to the next season of testing.
The breeder ultimately relies on test results from multiple locations and years for making the decision on whether a genotype merits release as a clonal or pure-line cultivar or as a component of synthetic or hybrid cultivar. The purpose of Applied Learning Activity 2 is to illustrate the challenge that genotype x environment interaction presents to the breeder for making selections.
Applied Learning Activity 2
The following data are from the M.S. thesis of Raechel Baumgartner at Iowa State University in which she evaluated the total tocopherol (Vitamin E) content in the oil of soybean lines. There were 20 lines with mid-oleate content of about 50% grown in a randomized complete-block design with two replications at each of three Iowa locations. The goal of the breeding program is to develop cultivars with high Vitamin E content.
The genotype x environment interaction was significant for population 2, but not for populations 1. Provide answers for each of the following questions and an explanation for each answer.
- What are two primary causes or types of genotype x environment interaction? How does each type affected selection of lines by a breeder?
- Which of the two types of interactions are responsible for the significant genotype x environment interaction for mid-oleate lines in population 2? It is possible to both types of interactions to be involved in a significant genotype x environment interaction.
- Based on the phenotypic correlations among the mean values for the individual lines, do you think that genotype x environment interaction would likely make it more or less difficult for a breeder to select the best lines in population1 than in population 2, even though the analyses of variance for population 1indicated that the interaction was not statistically significant?
- Which lines, if any, would you be comfortable selecting in each population based on one environment of data? Give the designation of the lines for each population that you would be comfortable selecting. Keep in mind that any genotype you advance for additional testing will utilize your financial resources that always are limiting in a breeding program.
- To minimize the impact of genotype x environment interaction on genetic gain in a breeding program, would it be more important to emphasize the number of environments used to evaluate lines or the number of replications at each environment? Use the genetic gain equation to defend your answer.
- How would the heritability of a trait relate to the amount of testing required to determine the genetic potential of an individual for that traits? Compare a trait of your choice that has a relatively low heritability and another that has a relatively high heritability. Use a real example from the literature or your own experience, not a hypothetical example.
Population |
Location | Carlisle | Rippey |
---|---|---|---|
1 | Ames | 0.20ns† | 0.41ns |
1 | Carlisle |
|
0.25ns |
2 | Ames | 0.86** | 0.85** |
2 | Carlisle |
|
0.94** |
- * significant at the 0.05 probability level
- ** significant at the 0.01 probability level
- † ns = not significant at the 0.05 probability level
Entry | Mean (Ames), mg kg-1 | Rank (Ames) | Mean (Carlisle), mg kg-1 | Rank (Carlisle) | Mean (Rippey), mg kg-1 | Rank (Rippey) | Overall Mean | Rank |
---|---|---|---|---|---|---|---|---|
418001 | 1646 | 20 | 1799 | 14 | 1820 | 12 | 1755 | 17 |
418002 | 1786 | 16 | 1924 | 8 | 1648 | 17 | 1786 | 16 |
418003 | 1824 | 13 | 1474 | 20 | 1111 | 20 | 1469 | 20 |
418004 | 2078 | 1 | 2010 | 3 | 1737 | 14 | 1942 | 8 |
418005 | 1677 | 19 | 1945 | 6 | 2019 | 4 | 1880 | 9 |
418007 | 2041 | 4 | 1952 | 4 | 1993 | 8 | 1995 | 3 |
418008 | 1843 | 12 | 1877 | 10 | 1789 | 13 | 1836 | 12 |
418010 | 1925 | 11 | 1805 | 12 | 1633 | 18 | 1787 | 15 |
418011 | 1949 | 9 | 1950 | 5 | 1980 | 9 | 1960 | 6 |
418012 | 1798 | 14 | 1788 | 15 | 1927 | 10 | 1837 | 11 |
418014 | 1948 | 10 | 1926 | 7 | 1652 | 16 | 1842 | 10 |
418016 | 1681 | 18 | 1640 | 17 | 1702 | 15 | 1674 | 18 |
418017 | 1949 | 8 | 2012 | 2 | 2044 | 2 | 2002 | 2 |
418018 | 2035 | 5 | 1820 | 11 | 2008 | 6 | 1954 | 7 |
418019 | 1978 | 7 | 1485 | 19 | 1993 | 7 | 1819 | 14 |
418020 | 1727 | 17 | 1726 | 16 | 1489 | 19 | 1647 | 19 |
418022 | 2046 | 3 | 1533 | 18 | 2324 | 1 | 1968 | 5 |
418024 | 1995 | 6 | 1909 | 9 | 2032 | 3 | 1979 | 4 |
418026 | 1788 | 15 | 1803 | 13 | 1865 | 11 | 1819 | 13 |
418027 | 2053 | 2 | 2017 | 1 | 2008 | 5 | 2026 | 1 |
Mean, Ames | Mean, Carlisle | Mean (Rippey) | Overall Mean | |
---|---|---|---|---|
LSD 0.05 | 220 | 777 | 642 | 340 |
LSD 0.01 | 295 | 1040 | 859 | 451 |
Entry | Mean (Ames), mg kg-1 | Rank (Ames) | Mean (Carlisle), mg kg-1 | Rank (Carlisle) | Mean (Rippey), mg kg-1 | Rank (Rippey) | Overall Mean | Rank |
---|---|---|---|---|---|---|---|---|
419002 | 1885 | 10 | 1800 | 10 | 1992 | 7 | 1892 | 9 |
419006 | 1730 | 19 | 1621 | 18 | 1734 | 17 | 1695 | 19 |
419007 | 1917 | 8 | 1816 | 9 | 2002 | 6 | 1912 | 8 |
419008 | 1809 | 16 | 1631 | 17 | 1689 | 19 | 1710 | 18 |
419009 | 1987 | 4 | 1898 | 3 | 2108 | 1 | 1998 | 3 |
419010 | 2006 | 3 | 1886 | 5 | 2071 | 4 | 1988 | 4 |
419011 | 2030 | 2 | 1957 | 1 | 2075 | 3 | 2021 | 2 |
419012 | 1953 | 7 | 1865 | 7 | 2017 | 5 | 1945 | 5 |
419013 | 1862 | 13 | 1597 | 19 | 1729 | 18 | 1729 | 17 |
419014 | 1844 | 14 | 1737 | 13 | 1857 | 13 | 1813 | 13 |
419015 | 1711 | 20 | 1580 | 20 | 1665 | 20 | 1652 | 20 |
419017 | 1954 | 6 | 1891 | 4 | 1972 | 9 | 1939 | 7 |
419018 | 1875 | 11 | 1671 | 16 | 1795 | 16 | 1780 | 15 |
419020 | 2135 | 1 | 1924 | 2 | 2091 | 2 | 2050 | 1 |
419021 | 1785 | 18 | 1701 | 14 | 1842 | 15 | 1776 | 16 |
419022 | 1901 | 9 | 1745 | 11 | 1943 | 10 | 1863 | 10 |
419023 | 1798 | 17 | 1686 | 15 | 1896 | 12 | 1793 | 14 |
419025 | 1979 | 5 | 1876 | 6 | 1977 | 8 | 1944 | 6 |
419026 | 1838 | 15 | 1841 | 8 | 1908 | 11 | 1862 | 11 |
419027 | 1863 | 12 | 1738 | 12 | 1851 | 14 | 1817 | 12 |
Mean, Ames | Mean, Carlisle | Mean (Rippey) | Overall Mean | |
---|---|---|---|---|
LSD 0.05 | 116 | 88 | 69 | 73 |
LSD 0.01 | 155 | 118 | 93 | 98 |
References
Fehr, W. R. (ed). 1987. Principles of Cultivar Development. Vol 1. Theory and Technique. McGraw-Hill, Inc., New York.