Page 2 - Korn_alii_2006

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

example, the number of segments may vary by three or even the geographical distribution of the subspecies T. c. simplex.
four within a sample of specimens from one locality (Linder Longhurst (1955) reported it only in northern Africa, from
1952; Longhurst 1955) and partially reduced or extended, Ceuta to Egypt. Recently, the range of this subspecies has
spiral-shaped body rings can also be observed (Linder 1952). been extended to the Arabian Peninsula, where it was found
This poses great difficulties for morphological classification at a single locality in Yemen (Thiéry 1996). According to
of the group (e.g. Linder 1952), resulting in a high number Longhurst (1955), the populations occurring in the northern
of species (more than 60) being described, sometimes on the part of the Iberian Peninsular belong to T. c. cancriformis.
basis of just a single specimen. However, Margalef (1953) and Alonso (1985, 1996) regarded
The Notostraca were revised by Linder (1952) and these northern Iberian populations as belonging to
Longhurst (1955). The latter author reduced the number of T. c. simplex. They also reported the absence of T. c. cancriformis
species to nine, four of which were in the genus Triops. from the Iberian Peninsula. This discordance deserves
More recent studies have concentrated on North American further investigation, for Longhurst (1955) did stress that it
populations, using allozyme electrophoresis and DNA sequence might be difficult to ascribe single specimens to T. c. cancriformis
analysis. These studies revealed the existence of cryptic species or T. c. simplex. The same perhaps might be true for whole
among North American Lepidurus and Triops (Sassaman et al. populations, given the high morphological variability inherent
1997; King & Hanner 1998; Rogers 2001). to this group. The problems that arise in morphological
In temperate Europe and the Mediterranean region, determinations within Triops may even be apparent at the
T. cancriformis is the predominant species of Triops. Mantovani species level. For some southern African populations, there is
et al. (2004) concluded that there were no cryptic species no consensus regarding their affiliation to T. cancriformis
among T. cancriformis but they only investigated populations (e.g. Barnard 1929) or Triops numidicus (e.g. Longhurst 1955).
of the subspecies T. c. cancriformis. However, at present After re-examination of the type specimens, Hamer & Rayner
three subspecies are recognized, all of which occur in the (1995) again classified these populations as T. cancriformis, but
Mediterranean region. Originally, four species were described without assigning them to one of the recognized subspecies.
(Ghigi 1921) for these Mediterranean populations: Triops Thus, the status of these southern African populations remains
mauritanicus, Triops simplex, T. cancriformis and Triops apulius, to be investigated.
of which the first three were later treated as subspecies of In this study, we use 16S and 12S rDNA sequences to
T. cancriformis (Longhurst 1955), while T. apulius was regarded investigate the phylogenetic relationships among the sub-
as a synonym of either T. c. cancriformis, or less probably, species of T. cancriformis and to clarify their distributions in
T. c. simplex (Longhurst 1955). the western Mediterranean region. We compare the sequence
The subspecies differ markedly in sex ratio. Typical data with key morphological characters and with reproduc-
T. c. cancriformis populations are either unisexual or female-biased, tive mode.
whereas populations of T. c. simplex and T. c. mauritanicus
have equal distribution of the sexes (Eder & Hödl 2003; Materials and methods
Scanabissi et al. 2005). Ignorance of the fact that sex ratios in Taxon sampling
T. cancriformis appear to be strictly linked to subspecies led to We attempted to acquire as many different samples of
the simplified assumption of a geographical parthenogenesis Triops cancriformis from Europe and North Africa as possible
in T. cancriformis (Zaffagnini & Trentini 1980) with a latitu- (locality data are listed in Table 1). We used both wild-caught
dinal gradient, as only T. c. cancriformis occurs in Central and specimens and specimens raised in the laboratory from eggs
northern Europe, whereas the other subspecies are restricted from sediments. Most samples were conserved in absolute
to more southern regions. However, if there is a geographical ethanol until extraction (a few samples were fixed in 70%
gradient in the distribution of reproductive modes in Euro- ethanol). Tissue vouchers were deposited in the ‘Tissue’ col-
pean populations of T. cancriformis, it is directed rather from lection of the ‘Museum fuer Tierkunde’ (Dresden, Germany)
west to east than from north to south, as, for example, no under the MTD-TW numbers listed in Table 1. Voucher
gonochoric populations (i.e. those that have an obligately specimens (with the exception of those loaned from a private
outcrossing mode of reproduction, with separate male and collection, see Acknowledgements section for details) from
female individuals) of this species have been reported from the morphological analyses were deposited in the ‘Inverte-
mainland Italy. brates’ collection of the same museum, under the numbers
Triops cancriformis cancriformis has a wide distribution, MTD Crus 2624–MTD Crus 2666. Sequences were submitted
from Europe and western Russia through the Middle East to GenBank with accession numbers AM183821–AM183917
to northern India (Longhurst 1955). The range of for 16S and AM184165–AM184184 for 12S sequences.
T. c. mauritanicus covers north-west Africa, southern Spain and Sequences already available in GenBank were also included
the island of Menorca (Longhurst 1955). Authors disagree in in the phylogenetic analyses (see Table 2).

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