Members of the Polypodium vulgare complex are so similar morphological
ly and chromosomally that delimiting diploid species and determining t
he origin of allopolyploid derivatives has been difficult. Isozyme ana
lyses supported published hypotheses of two major diploid lineages wit
hin the complex but rejected all but two hypotheses for allopolyploid
origins. New hypotheses were erected and the original complex of five
diploid species, three polyploid species, and two polyploid cytotypes
swelled to seven diploid and seven polyploid species. Subsequent isozy
me analyses supported the new hypotheses for allopolyploid origins and
continued to affirm earlier hypothesis of two diploid lineages within
the complex. At the populational level, isozyme analyses demonstrated
that most diploids conformed to Hardy-Weinberg expectations, but ther
e was some indication of inbreeding. A synthesis of the data suggested
that glaciation drove allopatric diploids into sympatry and initiated
allopolyploid speciation. Two of the allopolyploids had features that
suggested a more complex ancestry. Some populations of P. hesperium c
ontained isozyme and morphological features suggesting an earlier orig
in than the remainder of the populations. Apparent gene silencing in P
. vulgare suggested that this largely European species had an earlier
origin than others in the complex. Revised phylogenetic hypotheses wer
e used to extrapolate patterns of character evolution with greater pre
cision. The phylogenetic distribution of ''sporangiasters'' confirmed
earlier hypotheses concerning the development of these novel structure
s from sporangial initials.