Three cycles of recurrent restricted phenotypic selection (RRPS) were carried out on two cultivars each of timothy (Phleum pratense L.) and orchardgrass (Dactylis glomerata L.). The selections were conducted at the E. A. Lods Agronomy Research Center of Macdonald Campus of McGill University from 1984 to 1988. Individual plant selections were based on dry matter yields of 25 plants. Within each of 16 grids, the five highest yielding genotypes were selected (i.e. 80 genotypes overall) and allowed to interpollinate. Equal weights of seeds from each of the 80 selected plants were mixed in each cycle. Evaluation trials were carried out in both sward and spaced stand from 1989 to 1993 to assess the effectiveness of selection in improving dry matter yield. Several other morphological and physiological characteristics of cycles were recorded to find possible associated changes. / Three cycles of RRPS were not effective in increasing forage dry matter yields. The lack of response may be due to a narrow genetic base in the source populations and/or the population size under selection may have been too small, resulting in inbreeding depression. Another explanation for the ineffective selection may be non-random pollination among selected plants. The selection for forage yield produced some minor changes in other characteristics. There was a tendency towards increased plant height and circumference in orchardgrass; however, for timothy, height and circumference were reduced by selection. Cycle 3 of Avon orchardgrass was one day later in maturity and there was also a tendency for later maturity for timothy.

No description available.