The amount of genetic variation among and within tall fescue (Festuca
arundinacea Schreb.) cultivars is an interest of geneticists and breed
ers. However, limited information is available on this. The objective
of this study was to investigate the genetic diversity of tall fescue
germplasm based on RFLP data. Twenty-four probes from a tall fescue Ps
tI-genomic library were hybridized to EcoRI- or HindIII-digested genom
ic DNA from nine turf and seven forage cultivars, Twenty plants per cu
ltivar were individually analyzed. Results showed variation within cul
tivars was high, varied among cultivars and probes, and increased with
the number of parental clones (PN) up to 10 and then plateaued, indic
ating additional genetic variation could not be detected in cultivars
with more than 10 parents. The number of bands was significantly diffe
rent among cultivars and related to the within cultivar genetic divers
ity, but not to the PN. Probes not hybridizing to the P genome detecte
d less variation. Genetic distances calculated from 4, 8, 12, 16, and
20 plants were similar within cultivars, implying full random mating i
n tall fescue cultivar development. The between cultivar genetic dista
nces, however, decreased in larger sample sizes and became consistent
when sample size reached 16 plants. Results suggested 16 randomly chos
en plants was the minimum number to maintain adequate genetic diversit
y in these tall fescue cultivars. The associations of the 16 cultivars
from cluster analysis agreed with avail able pedigree information. Tu
rf cultivars tended to be clustered while forage cultivars were more d
ivergent.