excluded from most of Europe. It probably started
recolonisation after 18000 BP (the height of the Wiirm
III), in a pincer-like movement, along a western and an
eastern front (Fig. 1). In the east, this movement may
have originated from Ponto-Caspian relicts, and in the
West, from Maghrebian populations having crossed
north via the Gibraltar area. It appears that in the west,
on the Iberian peninsula, dispersal was much slower
than in the east, and that topography was most likely
responsible for that. In the Black Sea and the Caspian,
a large number of sizeable rivers (the Danube, Don,
Dnjepr, Wolga, etc.) debouch, which have northerly or westerly oriented valleys. These valleys must
have been extensively used by dispersing s. torvicornis. Conversely, on the Iberian peninsula, where rivers
have a predominantly east-west orientation, at right
angles to the direction of dispersal, the rate of advance
of the colonists was greatly slowed, and the barrier of
the Pyrenees has not been crossed even today. Likewise, the Danube Valley in the east was a prominent
pathway of dispersal, but the Dinarian Alps to the west
of it constituted an important barrier to dispersal, leaving the coastal zone unoccupied. 1
The mechanism of dispersal remains to be determined, but the upstream migratory tendency, using
river valleys as corridors, is strongly suggestive of
zoochoria and anemochoria. It remains unclear, however, why Italy and Corsardinia were never colonized,
except if long-distance dispersal is usually limited to an
accumulation of numerous small, individual steps. In
that case, we have to accept that the crossing at Gibraltar took place during the climatic optimum around 6000
years ago. During the early holocene eustatic lowering
of the mediterranean (ca 18000-12000 BP), the distance between Tunisia and Silicy-Italy was sufficiently
reduced to make a small-step policy credible, but temperatures then were probably too low for S. torvicornis
to occur on the mediterranean shores.
Support for this hypothesis is provided by the
fact that the species Branchipus schaefferi and Chirocephalus diaphanus, unlike Streptocephalus, do occur
in continental Europe including Italy, Corsardinia,
and Sicily (Cottarelli & Mura, 1979; 1983), as well
as in northern Africa (Hartland-Rowe, 1967; Thiery,
1986). Both genera are cold-resistant but still capable
of occurring in warmer climate belts during the cooler
part of the year. They therefore probably succeeded in
I Vekhoff (1993) presents convincing arguments in support ofthe
role of river vallies in the zoogeography of the Russian phyllopod
fauna.
283
the north- south crossing earlier, during the cold period
of the low eustatic sea level. They may not even have
been excluded from southern Europe at the height of
the glaciation.
A much stronger test of the pincer-shaped migration pattern of S. torvicornis is possible, however, and
will be explained hereafter.
S. t. bucheti lives in Iberia and N. Africa; S. t. torvicornis lives in E. Europe and the Ponto-Caspian
If it is true that the European populations of S. torvicornis moved in on two fronts, the borderline popUlations
on either side should have been separated in space
and time the longest, thus showing the largest amount
of morphological differenciation. If, conversely, the
western European hiatus were of recent age, the borderline popUlations should be closely related, and show
little or no morphological difference.
The test was primarily performed using populations from Spain (Zaragoza) and from Slovakia (Maly
Hores), but extended to popUlations from northern
Africa, the Levant and the Arabian peninsula. Males
were used, because of their diagnostic antennal characters. Figures 2-7 show that the double wave progression was confirmed by the facts: eastern European populations show a thumb with spines extending beyond
the major inflexion, and protuberances on the (anterior)
peduncle; western (Iberian) popUlations show a thumb
with a spine row terminating before the major inflexion, and a smooth anterior peduncle. Populations from
Morocco and northern Algeria conform to the Spanish ones, suggesting a recent common origin; further
to the centre and South of the Sahara, differences in
spine shape and size occur (see Dumont et aI., 1991),
but the pattern (spines not extending beyond the major
inflexion, no protuberances on the anterior peduncle)
remains consistent; even on the Arabian peninsula
(Yemen), where the spines on the thumb are smaller, the pattern remains consistent. A population from
the shorelands of the eastern mediterranean (Israel),
conversely, perfectly agreed with the Slovakian one in
morphology.
A taxonomic consequence of this finding is that
both morphological series can now be reassigned to
previously described but debated taxa, viz. S. t. torvicornis Waga in the north-east as far south as the Levant,
and S. t. bucheti Daday in the west, south and perhaps
in Arabia.
recolonisation after 18000 BP (the height of the Wiirm
III), in a pincer-like movement, along a western and an
eastern front (Fig. 1). In the east, this movement may
have originated from Ponto-Caspian relicts, and in the
West, from Maghrebian populations having crossed
north via the Gibraltar area. It appears that in the west,
on the Iberian peninsula, dispersal was much slower
than in the east, and that topography was most likely
responsible for that. In the Black Sea and the Caspian,
a large number of sizeable rivers (the Danube, Don,
Dnjepr, Wolga, etc.) debouch, which have northerly or westerly oriented valleys. These valleys must
have been extensively used by dispersing s. torvicornis. Conversely, on the Iberian peninsula, where rivers
have a predominantly east-west orientation, at right
angles to the direction of dispersal, the rate of advance
of the colonists was greatly slowed, and the barrier of
the Pyrenees has not been crossed even today. Likewise, the Danube Valley in the east was a prominent
pathway of dispersal, but the Dinarian Alps to the west
of it constituted an important barrier to dispersal, leaving the coastal zone unoccupied. 1
The mechanism of dispersal remains to be determined, but the upstream migratory tendency, using
river valleys as corridors, is strongly suggestive of
zoochoria and anemochoria. It remains unclear, however, why Italy and Corsardinia were never colonized,
except if long-distance dispersal is usually limited to an
accumulation of numerous small, individual steps. In
that case, we have to accept that the crossing at Gibraltar took place during the climatic optimum around 6000
years ago. During the early holocene eustatic lowering
of the mediterranean (ca 18000-12000 BP), the distance between Tunisia and Silicy-Italy was sufficiently
reduced to make a small-step policy credible, but temperatures then were probably too low for S. torvicornis
to occur on the mediterranean shores.
Support for this hypothesis is provided by the
fact that the species Branchipus schaefferi and Chirocephalus diaphanus, unlike Streptocephalus, do occur
in continental Europe including Italy, Corsardinia,
and Sicily (Cottarelli & Mura, 1979; 1983), as well
as in northern Africa (Hartland-Rowe, 1967; Thiery,
1986). Both genera are cold-resistant but still capable
of occurring in warmer climate belts during the cooler
part of the year. They therefore probably succeeded in
I Vekhoff (1993) presents convincing arguments in support ofthe
role of river vallies in the zoogeography of the Russian phyllopod
fauna.
283
the north- south crossing earlier, during the cold period
of the low eustatic sea level. They may not even have
been excluded from southern Europe at the height of
the glaciation.
A much stronger test of the pincer-shaped migration pattern of S. torvicornis is possible, however, and
will be explained hereafter.
S. t. bucheti lives in Iberia and N. Africa; S. t. torvicornis lives in E. Europe and the Ponto-Caspian
If it is true that the European populations of S. torvicornis moved in on two fronts, the borderline popUlations
on either side should have been separated in space
and time the longest, thus showing the largest amount
of morphological differenciation. If, conversely, the
western European hiatus were of recent age, the borderline popUlations should be closely related, and show
little or no morphological difference.
The test was primarily performed using populations from Spain (Zaragoza) and from Slovakia (Maly
Hores), but extended to popUlations from northern
Africa, the Levant and the Arabian peninsula. Males
were used, because of their diagnostic antennal characters. Figures 2-7 show that the double wave progression was confirmed by the facts: eastern European populations show a thumb with spines extending beyond
the major inflexion, and protuberances on the (anterior)
peduncle; western (Iberian) popUlations show a thumb
with a spine row terminating before the major inflexion, and a smooth anterior peduncle. Populations from
Morocco and northern Algeria conform to the Spanish ones, suggesting a recent common origin; further
to the centre and South of the Sahara, differences in
spine shape and size occur (see Dumont et aI., 1991),
but the pattern (spines not extending beyond the major
inflexion, no protuberances on the anterior peduncle)
remains consistent; even on the Arabian peninsula
(Yemen), where the spines on the thumb are smaller, the pattern remains consistent. A population from
the shorelands of the eastern mediterranean (Israel),
conversely, perfectly agreed with the Slovakian one in
morphology.
A taxonomic consequence of this finding is that
both morphological series can now be reassigned to
previously described but debated taxa, viz. S. t. torvicornis Waga in the north-east as far south as the Levant,
and S. t. bucheti Daday in the west, south and perhaps
in Arabia.