Page 15 - Korn_alii_2006

M. Korn et al. • Sister species within Triops cancriformis

parsimony-informative sites and five singletons observed (for We also include the Portuguese and Spanish populations
sequences from only European specimens, i.e. excluding the of the former T. c. mauritanicus in T. mauritanicus. However,
Japanese sequence: four singletons; if more sequences from further studies are needed to validate their substructure, posi-
Japanese specimen were included the alignment position 25 tion and status, and thus we refrain here from assigning them
would be parsimony informative, see section ‘Identification formal subspecific names. We do not have access to sufficient
of haplotype groups’). In contrast, the T. c. mauritanicus line- samples for genetic studies from the Iberian Peninsula to
age is very diverse. The sequences reveal 23 parsimony- undertake this investigation yet [most samples available in
informative sites and one singleton. Within the subclade collections for morphology are conserved in (only 70%)
T. c. simplex there are already three parsimony-informative sites. denatured alcohol or formalin, which degrades DNA, or are
The two main lineages based on T. c. cancriformis and too old]. Although the sister clade formation of the haplotype
T. c. mauritanicus are reciprocally monophyletic sister groups groups ‘Portugal’ and ‘Gitanilla’ appears in most reconstruc-
(Fig. 4). The two lineages have diverged by an average of tions (Fig. 4 and other topologies not shown), it may be an
2.9–3.3% in the 16S gene (Table 4a) and 4.1–4.3% in the 12S artefact of long branch attraction. One reconstruction
gene (Table 4c). This clear division of the T. cancriformis (PHYML tree of the separate subset of 16S sequences) did show
samples into two main lineages is in strong contrast to the a different clade formation (‘Portugal’ sister to T. m. simplex),
current taxonomy, which classifies T. cancriformis into three indicating the need for further study. Since the divergence
subspecies of equal rank. Rather, our genetic data support a among these Iberian samples is of the same magnitude as
classification into two main lineages of subspecific or even their differentiation from T. m. mauritanicus (Table 5), we
specific rank, in which the mauritanicus lineage also contains expect that further morphological and genetic analyses may
the clearly distinguished subclade of North African simplex differentiate further subspecies in Iberia.
samples. This subclade (including the atypical population As mentioned above, T. cancriformis and T. mauritanicus
from pool 063, Kairouan) is morphologically clearly separated lineages have diverged by an average of 2.9–3.3% in the 16S
from the remaining mauritanicus lineage by the much smaller gene (Table 4a). Unfortunately, no sequence data were
size of the furcal spines and was described as a separate available from Triops newberryi, the cryptic adelphotaxon of
species by Ghigi (1921, 1924; telson morphology, including T. longicaudatus (Sassaman et al. 1997), and no 16S sequences
furcal spines, was the most important source of characters were available from T. australiensis. The latter represents
used by Longhurst 1955 to separate Triops species). Its the closest relative of the adelphotaxa T. longicaudatus and
distinct position is also reflected by two autapomorphic T. newberryi and thus the 16S distances among these three taxa
substitutions in our 16S rDNA dataset (alignment positions are expected to be much lower than the smallest value (6.0%)
#115: A instead of G, #281: A instead of T; Appendix 1). indicated among recognized Triops species in Table 4a. This
We therefore reinstate the mauritanicus lineage to full is in line with the distance range observed with 12S sequences,
species status as Triops mauritanicus Ghigi, 1921, stat. rev., whereby the value of the distance between T. longicaudatus
with two subspecies in North Africa. Triops mauritanicus and T. australiensis is the lowest (6.1%; Table 4c). These two
mauritanicus (described by Ghigi 1921; from specimens species are morphologically similar (both species always
collected in Morocco and now held in the MNHN, Paris) have a completely reduced second maxilla, which is unique
is restricted to western Morocco north of the High Atlas, in Notostraca), but are nevertheless recognized by all authors
including the western ridges and mountain slopes of the latter. as distinct species. Occurring on different continents, it is not
The Moroccan and Tunisian populations of the former surprising that their divergence is slightly higher than the
T. c. simplex are here treated as T. m. simplex Ghigi, 1921, syn. one observed between T. cancriformis and T. mauritanicus
and stat. nov., on the basis of page priority and the Principle (4.1%; Table 4c), which both occur in the western Palaearctic.
of First Reviser (Algerian, Libyan, Egyptian, Sudanese and Further comparative data are available for Lepidurus
Arabian populations of the former T. c. simplex possibly also species from North America and Europe for the same DNA
belong to this subspecies). The clade T. m. mauritanicus fragment as used in the present study. Mantovani et al. (2004)
(T. c. mauritanicus from Morocco in Fig. 4) can be distinguished indicated that genetic distances between L. a. apus and
from the Iberian populations of the former T. c. mauritanicus L. a. lubbocki are of the same order of magnitude as those
by the strong reduction in size of dorsal carina spines in most observed between American Lepidurus species and furthermore
populations, the very long furcal spines (Fig. 5A) and the do not represent a monophyletic clade (the closest relative of
extremely high variability in the number of dorsal carina L. a. apus is L. arcticus, not L. a. lubbocki), which is why we per-
spines within most populations (Fig. 4D). Its monophyletic form distance comparisons with the two taxa separately. Among
status is genetically reflected by two autapomorphic substitu- well-recognized taxa of Lepidurus, 16S sequence divergences
tions in our 16S rDNA dataset (alignment positions #377: G (Table 4b; p-distance) may be as low as 2.8% (between L. arcticus
instead of A, #422: T instead of C; Appendix 1). and L. a. apus; 4.6% in 12S), 3.1% (between Lepidurus lemmoni

© 2006 The Authors. Journal compilation © 2006 The Norwegian Academy of Science and Letters • Zoologica Scripta, 35, 4, July 2006, pp301–322 315