CYTOGENETIC STUDIES OF POA. II. THE PAIRING OF CHROMOSOMES IN SPECIES AND INTERSPECIFIC HYBRIDS

Abstract

CHROMOSOME NUMBERS in the bluegrasses extend over an extreme range of ploidy from a few relatively restricted diploid species (2n=14) through to the more widespread species many of which have chromosome numbers ranging between 6-ploid (2n=42) and 13 or 14-ploid (2n=98) (Miintzing, 1933, 1940; Akerberg, 1942; Hartung, 1946; Nygren, 1950; Grun, 1954). At least three factors may influence chromosome pairing of plants belonging to such highly polyploid groups. First are the homologies of the chromosomes as determined by their gene structure or arrangement; second, mechanical complications arising from high polyploidy itself; and third, the action of physiological and environmental influences on chromosome behavior. The cytological make-up of plants even of non-hybrid origin is usually slightly unbalanced so that univalents are a typical feature of meiotic metaphase of all polyploid bluegrasses thus far examined. In the course of the Poa breeding program conducted by the Carnegie Institution numerous interspecific hybrids and hybrid derivatives have been produced. The present cytological survey of some of these plants is aimed at clarifying the relationships of chromosome pairing both within species and in interspecific hybrids, and at gaining some insight into the mechanisms governing pairing. MATERIALS AND METHODS.-Anthers from Carnoy-fixed florets were smeared in a specially prepared propiono-carmine stain (Grun, 1952). Fixations were made of culms of vigorous plants grown either in cultivated gardens or pots at Stanford or Mather, California, and in a few cases, of potted plants at the Earhart Laboratory of the California Institute of Technology at Pasadena. The average univalent frequency was determined by counting the number of unpaired chromosomes in a minimum of twenty sporocytes per plant. Standard errors of means reported indicate the possible random variation of the mean to be expected in 99 out of 100 trials. Univalent frequencies of most plants fell into normal distributions, but frequencies of those plants having nearly regular pairing often were Poisson distributions, a skewness that resulted from the fact that many cells contained no or very few univalents. Standard errors were, accordingly, calculated using the formula appropriate to the distribution of each population. The