46
ed in my laboratory over the past few years. The goal is
to synthesize two apparently disparate lines of inquiry
into a unified framework to understand the evolutionary history of reproductive variation in clam shrimps
as a group, but moreover, to emphasize the utility of
integrating these two kinds of information in understanding the processes of evolution of reproductive
systems.
I will argue that a large body of data on sex ratio
variation among species of conchostracans is informative - in the light of new laboratory studies - about
modes of reproduction in individual species. These
data, when considered in a phylogenetic perspective,
suggest that unisexuality is derived from an ancestral
condition of obligate sexuality, and that it has arisen
at least four times. Studies on sex-determining mechanisms in obligately sexual species indicate that females
are heterozygous for one or a very few genetic factors
specifying gender. These genetic factors are differently expressed in androdioecious species (where females
are capable of self-fertilization) to produce two female
phenotypes, one heterozygous and one homozygous.
Natural selection favoring the homozygous female
phenotypes of androdioecious species would be sufficient to create a unisexual condition, and this evolutionary mechanism may explain the large number
of cases of unisexuality observed within the Limnadiidae. Three other cases of unisexuality, in cyzicids,
leptestheriids and cyclestheriids, however, probably
originated from different ancestors and evolved by different pathways.
Taxonomic conventions
The classification of clam shrimps is controversial at
all levels ranging from the diagnosis of species to that
of orders. Since some of the interpretations in this
review are based on evolutionary arguments which
may be influenced by my particular taxonomic conventions, I will here outline the most important of
these. First, I view the Lynceidae as the sister group to
all other conchostracans rather than as a distinct order
with unspecified affinities to conchostracans, cladocerans, and notostracans (Fryer, 1987). Second, I follow
Straskraba (1965a) in regarding Imnadia Hertzog and
M etalimnadia Mattox as genera of Limnadiidae rather
than warranting separate familial status (Botnariuc &
Orghidan, 1941; Marincek & Petrov, 1991, Roessler,
1991 b, 1995a). Finally, Cyzicus Audouin, as used here,
includes those species formerly classified as Caenestheriella Daday (following Straskraba, 1965b; Wiltshire, 1973; Forro & Brtek, 1984) but does not include
Caenestheria Daday. Williams (1980) treats all Australian Cyzicidae, including those described as Caenestheria, as Cyzicus. My reluctance to view Caenestheria as a synonym of Cyzicus stems from my studies on
larval morphology to be detailed elsewhere. Overall,
my classification of conchostracan genera differs from
Martin (1992) only in recognizing Paracyclestheria
Shen and Dai and not recognizing Caenestheriella.
Terminology of reproduction
The complexity of reproducti ve systems in animals and
plants has created a bewildering nomenclature (Bell,
1982). Much of the terminology is dependent upon
the cytological details of gametogenesis, an area of
conchostracan biology that has remained almost totally unexplored. The terminology in this review will
focus on functional properties of clam shrimp reproduction rather than the cytological specifics. Thus, I
will use the terms 'obligately sexual' and 'gonochoric'
interchangeably to refer to any system in which reproduction requires the union of gametes produced by two
different individuals. 'Selfing' and 'hermaphroditism'
will be used to refer to cases in which genetic evidence
implicates the union of two independently derived
gametes from the same individual, and 'parthenogenesis' will be used in its narrowest definition - the transmission of the mother's genotype, without modification, to her offspring.
The conchostracan male is recognizable by the
presence of one or two pairs of anterior limbs modified to form claspers with which to effect pair formation. Males clasp the ventral free edge of the
female carapace for extended periods of time, periodically inserting their body into the cavity between
the females' valves (Sars, 1896a; Gravier & Mathias, 1930; Valtonen, 1966; Martin, et aI., 1986;
Knoll, 1994). Sperm are ameboid (in one case rodshaped) (Wingstrand, 1978; Roessler & Sanchez,
1986; Roessler, 1995c) and are thought to be transferred to the female in a spermatophore-like package (Solowiow, 1927; Strenth, 1977, but, see Knoll,
1994).
The terminology for egg-bearing individuals is not
so evident. Females lack modifications of the anterior
limbs and are the producers and brooders of the eggs.
However, in some groups the females are not restricted to that role; rather they are hermaphroditic with
regions of the gonad differentiated to produce sperm.
Cytological evidence of hermaphroditism in Limnadia
ed in my laboratory over the past few years. The goal is
to synthesize two apparently disparate lines of inquiry
into a unified framework to understand the evolutionary history of reproductive variation in clam shrimps
as a group, but moreover, to emphasize the utility of
integrating these two kinds of information in understanding the processes of evolution of reproductive
systems.
I will argue that a large body of data on sex ratio
variation among species of conchostracans is informative - in the light of new laboratory studies - about
modes of reproduction in individual species. These
data, when considered in a phylogenetic perspective,
suggest that unisexuality is derived from an ancestral
condition of obligate sexuality, and that it has arisen
at least four times. Studies on sex-determining mechanisms in obligately sexual species indicate that females
are heterozygous for one or a very few genetic factors
specifying gender. These genetic factors are differently expressed in androdioecious species (where females
are capable of self-fertilization) to produce two female
phenotypes, one heterozygous and one homozygous.
Natural selection favoring the homozygous female
phenotypes of androdioecious species would be sufficient to create a unisexual condition, and this evolutionary mechanism may explain the large number
of cases of unisexuality observed within the Limnadiidae. Three other cases of unisexuality, in cyzicids,
leptestheriids and cyclestheriids, however, probably
originated from different ancestors and evolved by different pathways.
Taxonomic conventions
The classification of clam shrimps is controversial at
all levels ranging from the diagnosis of species to that
of orders. Since some of the interpretations in this
review are based on evolutionary arguments which
may be influenced by my particular taxonomic conventions, I will here outline the most important of
these. First, I view the Lynceidae as the sister group to
all other conchostracans rather than as a distinct order
with unspecified affinities to conchostracans, cladocerans, and notostracans (Fryer, 1987). Second, I follow
Straskraba (1965a) in regarding Imnadia Hertzog and
M etalimnadia Mattox as genera of Limnadiidae rather
than warranting separate familial status (Botnariuc &
Orghidan, 1941; Marincek & Petrov, 1991, Roessler,
1991 b, 1995a). Finally, Cyzicus Audouin, as used here,
includes those species formerly classified as Caenestheriella Daday (following Straskraba, 1965b; Wiltshire, 1973; Forro & Brtek, 1984) but does not include
Caenestheria Daday. Williams (1980) treats all Australian Cyzicidae, including those described as Caenestheria, as Cyzicus. My reluctance to view Caenestheria as a synonym of Cyzicus stems from my studies on
larval morphology to be detailed elsewhere. Overall,
my classification of conchostracan genera differs from
Martin (1992) only in recognizing Paracyclestheria
Shen and Dai and not recognizing Caenestheriella.
Terminology of reproduction
The complexity of reproducti ve systems in animals and
plants has created a bewildering nomenclature (Bell,
1982). Much of the terminology is dependent upon
the cytological details of gametogenesis, an area of
conchostracan biology that has remained almost totally unexplored. The terminology in this review will
focus on functional properties of clam shrimp reproduction rather than the cytological specifics. Thus, I
will use the terms 'obligately sexual' and 'gonochoric'
interchangeably to refer to any system in which reproduction requires the union of gametes produced by two
different individuals. 'Selfing' and 'hermaphroditism'
will be used to refer to cases in which genetic evidence
implicates the union of two independently derived
gametes from the same individual, and 'parthenogenesis' will be used in its narrowest definition - the transmission of the mother's genotype, without modification, to her offspring.
The conchostracan male is recognizable by the
presence of one or two pairs of anterior limbs modified to form claspers with which to effect pair formation. Males clasp the ventral free edge of the
female carapace for extended periods of time, periodically inserting their body into the cavity between
the females' valves (Sars, 1896a; Gravier & Mathias, 1930; Valtonen, 1966; Martin, et aI., 1986;
Knoll, 1994). Sperm are ameboid (in one case rodshaped) (Wingstrand, 1978; Roessler & Sanchez,
1986; Roessler, 1995c) and are thought to be transferred to the female in a spermatophore-like package (Solowiow, 1927; Strenth, 1977, but, see Knoll,
1994).
The terminology for egg-bearing individuals is not
so evident. Females lack modifications of the anterior
limbs and are the producers and brooders of the eggs.
However, in some groups the females are not restricted to that role; rather they are hermaphroditic with
regions of the gonad differentiated to produce sperm.
Cytological evidence of hermaphroditism in Limnadia