The Gene Pool Concept as a Basis For Cultivar Selection and Recommendation1
- A. R. Pedersen,
- E. H. Everson and
- J. E. Grafius2
Plant breeders are greatly concerned with the number and location of test sites necessary to assess cultivar performance. Yield trials rarely contain the same entries over sites and years, and this greatly reduces the flexibility needed to compare advanced germplasm with current commercial cultivars over a period of time. In this paper we provide an alternative method to the statistical procedures normally used to compare cultivars. Winter wheat (Triticum aestivum L.) nurseries in Michigan were used to develop the concept of a gene pool base, which allows comparisons over years and sites without regard to particular entries in the sample. It is proposed that as long as entries in the test are a representative sample of the population gene pool, the sample mean can be used as an index of comparison. Yearly additions and subtractions to the gene pool over the relatively short time span of six years did not greatly affect the average reaction to the environment. The regression of cultivar mean yield on nursery mean yield demonstrated that 53 to 95% of the variation in cultivar performance could be explained by site mean performance, with most cultivars falling in the 70 to 85% range. The data points were distributed in a continuum from high to low yielding sites. In graphing the regression of a cultivar on site mean yields, the designation of points by location could be used to suggest which cultivars should be grown in each environment and yield potential range. Locations were identified that had consistently high, low, intermediate, or extremely variable yield ranges over years. Any minimal set of test sites should include the extremes to provide an adequate estimate of the regression line.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © . .