42
the latter area to North America" (Banarescu, 1990)
and "the American species group is probably derived
and diversified from a streptocephalid stock introduced
from Africa" (Wiman, 1979a), nowadays distribution of Streptocephalus seems better explained by the
vicariance model (Croizat et at., 1974). We consider two alternatives. The first, offered by Belk (1984),
proposes that given its absence from South America
and Australia, Streptocephalus had a Laurasian origin
(early Cretaceous, c. 105 mybp), and after the breakup
of Gondwanaland (late Cretaceous, c. 65-70 mybp)
dispersed through Africa, Madagascar, India, Sri Lanka, and Indonesia, with a major adaptative radiation
in temperate Africa. The alternative is that Streptocephalus had a Pangaean distribution, but the species
of South America and Australia later became extinct,
or have not been discovered yet.
Cladistics
In the first analysis, four most parsimonious cladograms of 25 steps in length, CI 0.88, and RI 0.92 were
found. The only difference between these trees is the
arrangement of the relationship between the species of
the sealii group. We propose one of them, as our working hypothesis of the phylogeny of the New World
species (Fig. 15). A closer relationship seems to exist
between S. sea Iii and S. woottoni, evidenced by the
presence of linguiform genital outgrowths, which are
absent in S. similis. Although S. sealii and S. simiUs have spinose cercopods, the linguiform outgrowths
seem to be more important. For example, closely related species like those of the torvicornis subgroup of the
Old World, and the species of the dichotomus group,
have these structures, while closely related species
such as S. bidentatus Hamer & Appleton, 1993 and
S. vitreus (Brauer, 1887), exhibit different types of
cercopods (Hamer & Appleton, 1993). The shape of
the spur of the thumb (character No.5), appears to be
a homoplastic character. The different arrangements
of the relationship of the members of the sealii group,
shown by the four trees, is reflected in the strict consensus tree, where there is no resolution for those species
(Fig. 16). However, well-defined monophyletic groups
are formed, which are in accordance with the classification of the species groups and subgroups, i.e. the
mackini and torvicornis subgroups of the torvicornis
group.
In the second analysis, eight most parsimonious
cladograms of 35 steps in length, CI 0.71, and RI 0.87
were found. Again, the difference between them is in
the arrangement of the relationship between the forms
of the sealii group. However, in the strict consensus
cladogram (Fig. 17) two discrepancies are found. First,
S. (gr. sealii) woottoni keys out as part of a monophyletic group together with members of the torvicornis group, and second, S. guzmani and S. mattoxi, are
separated from the other members of the torvicornis
subgroup, to form a monophy letic group with the members of the mackini subgroup (Fig. 17). Homoplasy is
noted in characters 5 (apex of spur), 19 (linguiform
outgrowths), and 22 (lateral lamella of the thumb).
Discussion
The results of the cladistic analyses support our hypothesis that S. kargesi is more nearly related to the African
species of the dendyi group than to American species,
and similarly for the members of the sealii and torvicornis group.
Contrary to a close relationship between S. kargesi and S. moorei indicated by Spicer'S (1985) cladogram, S. kargesi appears as one of the most primitive
Streptocephalus, and S. moo rei as one of the most
morphologically complex species in the New World,
related most closely with members of the mackini subgroup having biramous ovaries. Spicer's (1985) cladogram was influenced by the character state morphology of the tip of the frontal appendage. However, the
frontal appendage has been demonstrated to be subject
to intraspecific variability (Moore, 1958).
Our results agree with Spicer's (1985), that S.
mackini is far from representing the primitive stock
from which the other North American species were
derived, as speculated by Wiman (1979a). We agree
with the objection expressed by Spicer (1985) that
the ability to hybridize is not a decisive criterion for
understanding the phylogeny of the group, as Wiman
(1979a) suggested. Recently, Dubois (1988) proposed
that hybridization serves best as an indication of genus
membership. Under this point of view, therefore, the
ability to hybridize is a plesiomorphic feature of congeneric species. This is not to deny the importance
of studies on hybridization. To the contrary, they are
of a great value in defining generic identity and have
potential use for developing a measurement method
of "hybrid distance" between species (Dubois, 1988).
Hybridization studies clearly have a place in the study
on anostracan biology.
the latter area to North America" (Banarescu, 1990)
and "the American species group is probably derived
and diversified from a streptocephalid stock introduced
from Africa" (Wiman, 1979a), nowadays distribution of Streptocephalus seems better explained by the
vicariance model (Croizat et at., 1974). We consider two alternatives. The first, offered by Belk (1984),
proposes that given its absence from South America
and Australia, Streptocephalus had a Laurasian origin
(early Cretaceous, c. 105 mybp), and after the breakup
of Gondwanaland (late Cretaceous, c. 65-70 mybp)
dispersed through Africa, Madagascar, India, Sri Lanka, and Indonesia, with a major adaptative radiation
in temperate Africa. The alternative is that Streptocephalus had a Pangaean distribution, but the species
of South America and Australia later became extinct,
or have not been discovered yet.
Cladistics
In the first analysis, four most parsimonious cladograms of 25 steps in length, CI 0.88, and RI 0.92 were
found. The only difference between these trees is the
arrangement of the relationship between the species of
the sealii group. We propose one of them, as our working hypothesis of the phylogeny of the New World
species (Fig. 15). A closer relationship seems to exist
between S. sea Iii and S. woottoni, evidenced by the
presence of linguiform genital outgrowths, which are
absent in S. similis. Although S. sealii and S. simiUs have spinose cercopods, the linguiform outgrowths
seem to be more important. For example, closely related species like those of the torvicornis subgroup of the
Old World, and the species of the dichotomus group,
have these structures, while closely related species
such as S. bidentatus Hamer & Appleton, 1993 and
S. vitreus (Brauer, 1887), exhibit different types of
cercopods (Hamer & Appleton, 1993). The shape of
the spur of the thumb (character No.5), appears to be
a homoplastic character. The different arrangements
of the relationship of the members of the sealii group,
shown by the four trees, is reflected in the strict consensus tree, where there is no resolution for those species
(Fig. 16). However, well-defined monophyletic groups
are formed, which are in accordance with the classification of the species groups and subgroups, i.e. the
mackini and torvicornis subgroups of the torvicornis
group.
In the second analysis, eight most parsimonious
cladograms of 35 steps in length, CI 0.71, and RI 0.87
were found. Again, the difference between them is in
the arrangement of the relationship between the forms
of the sealii group. However, in the strict consensus
cladogram (Fig. 17) two discrepancies are found. First,
S. (gr. sealii) woottoni keys out as part of a monophyletic group together with members of the torvicornis group, and second, S. guzmani and S. mattoxi, are
separated from the other members of the torvicornis
subgroup, to form a monophy letic group with the members of the mackini subgroup (Fig. 17). Homoplasy is
noted in characters 5 (apex of spur), 19 (linguiform
outgrowths), and 22 (lateral lamella of the thumb).
Discussion
The results of the cladistic analyses support our hypothesis that S. kargesi is more nearly related to the African
species of the dendyi group than to American species,
and similarly for the members of the sealii and torvicornis group.
Contrary to a close relationship between S. kargesi and S. moorei indicated by Spicer'S (1985) cladogram, S. kargesi appears as one of the most primitive
Streptocephalus, and S. moo rei as one of the most
morphologically complex species in the New World,
related most closely with members of the mackini subgroup having biramous ovaries. Spicer's (1985) cladogram was influenced by the character state morphology of the tip of the frontal appendage. However, the
frontal appendage has been demonstrated to be subject
to intraspecific variability (Moore, 1958).
Our results agree with Spicer's (1985), that S.
mackini is far from representing the primitive stock
from which the other North American species were
derived, as speculated by Wiman (1979a). We agree
with the objection expressed by Spicer (1985) that
the ability to hybridize is not a decisive criterion for
understanding the phylogeny of the group, as Wiman
(1979a) suggested. Recently, Dubois (1988) proposed
that hybridization serves best as an indication of genus
membership. Under this point of view, therefore, the
ability to hybridize is a plesiomorphic feature of congeneric species. This is not to deny the importance
of studies on hybridization. To the contrary, they are
of a great value in defining generic identity and have
potential use for developing a measurement method
of "hybrid distance" between species (Dubois, 1988).
Hybridization studies clearly have a place in the study
on anostracan biology.