The number of cultivars of perennial ryegrass has steadily increased s
ince the early 1960s. Perennial ryegrass has a self-incompatible, cros
s-pollinated breeding system and thus, each cultivar is a heterogeneou
s population of individual genotypes. The ability to accurately distin
guish large numbers of genetically heterogeneous populations from one
another, based solely on morphological data and agronomic performance,
is limited. Random amplified polymorphic DNA (RAPD) markers offer a n
early unlimited supply of molecular traits on which to base distinctiv
e characterizations of plant populations. Analysis of molecular varian
ce (AMOVA) provides a statistical tool which partitions molecular mark
er variation within and among populations and performs significance te
sting of differences. The objective of this research was to survey RAP
D marker variation within and among populations of perennial ryegrass
to characterize breeding germplasm and commercial cultivars using AMOV
A. A range of genetic diversity was represented by choosing 18 grass p
opulations that included related crosses, derived selections, and unre
lated ecotypes. For each population, 10 individuals were examined for
33 RAPD markers produced with two primers. Markers present in all indi
viduals of a particular population but not present in any other indivi
dual of another population (fixed marker difference) were not observed
among the 18 grass populations, so characterizations were based on ma
rker frequency differences among populations. RAPD markers analyzed wi
th AMOVA closely reflected the breeding histories of the perennial rye
grass germplasm or commercial cultivars. This was particularly true fo
r relationships structured on a population basis. Results indicated th
at high selection intensity substantially altered the distribution of
RAPD markers between parent and progeny populations thereby obscuring
measurements of genetic relatedness. Furthermore, this study observed
a limited ability to separate closely related populations. Evidence fo
r the narrow germplasm base of turfgrass cultivars was observed; altho
ugh, many perennial ryegrass populations appeared to retain levels of
within population variability comparable with that of land race ecotyp
es. The methods of analysis presented should provide a supplement to t
raditional morphological and agronomic data for plant variety protecti
on and for measuring genetic diversity within breeding programs.