of males among the hundreds of individuals that have
been examined in both field samples and laboratory
rearings (Mattox, 1950; Mattox & Velardo, 1950). No
other species of Cyzicus, or any Cyzicidae for that
matter, has a populational structure that is known to
lack males, nor has androdioecy been demonstrated
in any cyzicid either. Were C. gynecia to have been
separated from its nearest living relative by a series of
other androdioecious species now extinct, it could be
expected to differ morphologically from other extant
forms. It is more likely, however, that it is merely a
recent derivation of a parthenogenetic form directly
from a sexual ancestor. One possible candidate would
be Cyzicus mexicanus. Cyzicus gynecia has a relatively limited geographic range in the north-central United States (Wiltshire & Hazelwood, 1972), overlapping
extensively the geographic range of Cyzicus mexicanus
(Mattox, 1957). No distinct characters were provided
in the original description (other than the absence of
males) to diagnose C. gynecia from the highly variable C. mexicanus. (Indeed, the original placement of
C. gynecia in Caenestheriella is questionable, since
that genus was defined by Daday (1914) on the basis
of a sexual dimorphism that cannot be evaluated in
a unisexual species.) If C. gynecia is derived from
C. mexicanus, or from a recent common ancestor, it
is predicted to be composed entirely of Sis (heterogametic) females (Fig. 4). A mutation with the effect of
suppressing meiosis during gametogenesis, such as the
first polar body fusion observed in Limnadia lenticularis (Zaffagnini, 1969), would account for the origin
of such a life cycle. The hypothesis that C. gynecia is
recently derived from another sexual species is further
testable by analysis of, for example, genetic differentiation among C. gynecia populations and genetic distance comparisons to populations of C. mexican us.
The newly described case of a unisexual species
of Leptestheria in Colombia (Roessler, 1995b) would
seem to represent a parallel circumstance. Although a
detailed characterization of this species remains to be
completed, several aspects of its biology appear to be
comparable to that of C. gynecia. There are no other
indications of unisexuality elsewhere in the genus, or in
the entire family (Table 1), and the one known species
has a very limited geographic distribution (Roessler,
1994b).1t is quite conceivable that this species is also
recently derived from a sexual ancestor by a mechanism involving the suppression of meiosis in heterogametic females.
59
Cyclestheriidae
The evolution of cyclical parthenogenesis from an obligately sexual ancestor undoubtedly involved a more
complex sequence of changes than those postulated
for the evolution of unisexuality in limnadiids and
cyzicids. Hebert (1987) has presented a scenario for
the evolution of cyclic parthenogenesis in cladocerans starting from a presumed ancestral condition of
bisexual reproduction and genetic sex determination.
His model involves a complex sequence of changes
in both reproductive mode and the mechanism of sex
determination. The evolutionary pathway is initiated
by retention of direct-developing (subitaneous) eggs in
the brood chamber that develop into a mixture of male
and female offspring. This condition is followed by a
change in the mechanism of sex determination from
genetic determination to environmental control, leading to the production of single-sexed clutches (either
male or female) of direct-developing eggs. Clutches
of diapausing eggs would still produce mixtures of
male and female offspring. He then postulates a mutation that suppresses meiosis during the production of
direct-developing, but not diapausing, eggs. Natural
selection, operating on the conditional sex determination, then differentially modulated sex ratios to eliminate males from the diapause phase of the life cycle
and restrict their occurrence to particular times during
the direct-development phase.
Although Hebert's (1987) scenario could apply to
the evolutionary history of Cyclestheria, a simpler
model (Fig. 4) should be explored first. My model
supposes initially the suppression of meiosis to create a unisexual (parthenogenetic) life cycle, similar to
that inferred for Cyzicus gynecia and Leptestheria sp.
(Fig. 4). This step is followed by the retention of directdeveloping eggs within the brood chamber; the switch
between retaining subitaneous eggs versus producing
diapausing eggs being under environmental influence.
A second mutation in the physiological process leading
to diapause eggs restores a modified form of egg production that prevents segregation of homologs. One
possible mechanism would be restoration of normal
meiosis, but accompanied by fusion of the first two
cleavage nucleii during embryonic development. The
resulting offspring would be mixtures of homozygous
males and homozygous females. These individuals,
reproducing by normal sexual methods, would produce a new generation of diapausing eggs that were
uniformly heterozygous for the sex-determining locus.
This model requires several mutational changes to pro-
been examined in both field samples and laboratory
rearings (Mattox, 1950; Mattox & Velardo, 1950). No
other species of Cyzicus, or any Cyzicidae for that
matter, has a populational structure that is known to
lack males, nor has androdioecy been demonstrated
in any cyzicid either. Were C. gynecia to have been
separated from its nearest living relative by a series of
other androdioecious species now extinct, it could be
expected to differ morphologically from other extant
forms. It is more likely, however, that it is merely a
recent derivation of a parthenogenetic form directly
from a sexual ancestor. One possible candidate would
be Cyzicus mexicanus. Cyzicus gynecia has a relatively limited geographic range in the north-central United States (Wiltshire & Hazelwood, 1972), overlapping
extensively the geographic range of Cyzicus mexicanus
(Mattox, 1957). No distinct characters were provided
in the original description (other than the absence of
males) to diagnose C. gynecia from the highly variable C. mexicanus. (Indeed, the original placement of
C. gynecia in Caenestheriella is questionable, since
that genus was defined by Daday (1914) on the basis
of a sexual dimorphism that cannot be evaluated in
a unisexual species.) If C. gynecia is derived from
C. mexicanus, or from a recent common ancestor, it
is predicted to be composed entirely of Sis (heterogametic) females (Fig. 4). A mutation with the effect of
suppressing meiosis during gametogenesis, such as the
first polar body fusion observed in Limnadia lenticularis (Zaffagnini, 1969), would account for the origin
of such a life cycle. The hypothesis that C. gynecia is
recently derived from another sexual species is further
testable by analysis of, for example, genetic differentiation among C. gynecia populations and genetic distance comparisons to populations of C. mexican us.
The newly described case of a unisexual species
of Leptestheria in Colombia (Roessler, 1995b) would
seem to represent a parallel circumstance. Although a
detailed characterization of this species remains to be
completed, several aspects of its biology appear to be
comparable to that of C. gynecia. There are no other
indications of unisexuality elsewhere in the genus, or in
the entire family (Table 1), and the one known species
has a very limited geographic distribution (Roessler,
1994b).1t is quite conceivable that this species is also
recently derived from a sexual ancestor by a mechanism involving the suppression of meiosis in heterogametic females.
59
Cyclestheriidae
The evolution of cyclical parthenogenesis from an obligately sexual ancestor undoubtedly involved a more
complex sequence of changes than those postulated
for the evolution of unisexuality in limnadiids and
cyzicids. Hebert (1987) has presented a scenario for
the evolution of cyclic parthenogenesis in cladocerans starting from a presumed ancestral condition of
bisexual reproduction and genetic sex determination.
His model involves a complex sequence of changes
in both reproductive mode and the mechanism of sex
determination. The evolutionary pathway is initiated
by retention of direct-developing (subitaneous) eggs in
the brood chamber that develop into a mixture of male
and female offspring. This condition is followed by a
change in the mechanism of sex determination from
genetic determination to environmental control, leading to the production of single-sexed clutches (either
male or female) of direct-developing eggs. Clutches
of diapausing eggs would still produce mixtures of
male and female offspring. He then postulates a mutation that suppresses meiosis during the production of
direct-developing, but not diapausing, eggs. Natural
selection, operating on the conditional sex determination, then differentially modulated sex ratios to eliminate males from the diapause phase of the life cycle
and restrict their occurrence to particular times during
the direct-development phase.
Although Hebert's (1987) scenario could apply to
the evolutionary history of Cyclestheria, a simpler
model (Fig. 4) should be explored first. My model
supposes initially the suppression of meiosis to create a unisexual (parthenogenetic) life cycle, similar to
that inferred for Cyzicus gynecia and Leptestheria sp.
(Fig. 4). This step is followed by the retention of directdeveloping eggs within the brood chamber; the switch
between retaining subitaneous eggs versus producing
diapausing eggs being under environmental influence.
A second mutation in the physiological process leading
to diapause eggs restores a modified form of egg production that prevents segregation of homologs. One
possible mechanism would be restoration of normal
meiosis, but accompanied by fusion of the first two
cleavage nucleii during embryonic development. The
resulting offspring would be mixtures of homozygous
males and homozygous females. These individuals,
reproducing by normal sexual methods, would produce a new generation of diapausing eggs that were
uniformly heterozygous for the sex-determining locus.
This model requires several mutational changes to pro-