The use of regression techniques for predicting the response of peas to environment.

Six pea varieties were grown in seven environments over two consecutive years and the results analysed using regression techniques. In both years a degree of linearity of response to environment was established for many of the characters recorded. Comparisons were made between the values observed in the second year and those predicted from the results of the first year, using 't' tests. As a result of these comparisons it appears that, due to poor linearity of response and/or different degrees of response in different years, accurate predictability of genotype behaviour was not possible.