Page 13 - Asylv_molars_Pmax_revised_2014_01

13

partially
counterbalanced
the
insular
effect,
tending
to
promote
morphological
divergence.
Yet,
in

this
case,
the
divergence
between
this
lineage
and
its
Italian
relative
occurs
parallel
to
Pmax.
The

respective
role
of
isolation
and
adaptation
are
difficult
to
disentangle
in
the
case
of
this
endemic

lineage.

Latitudinal
trend
parallel
to
Pmax

Superimposed
to
the
differentiation
among
clades,
a
latitudinal
trend
in
molar
shape
has
been

documented
in
the
European
wood
mouse.
It
corresponds
to
broad
teeth
towards
the
South
and

slender
teeth
towards
the
North
(Renaud
and
Michaux
2007),
echoing
a
trend
in
mandible
shape

found
parallel
in
several
species
of
wood
mice
(Renaud
and
Michaux
2003).
As
most
clinal
geographic

trend,
it
has
been
interpreted
as
a
response
to
selection
(e.g.
Cardini
et
al.
2007b),
possibly
related
to

variations
in
food
resources
across
Europe.
Such
a
response
to
a
presumably
moderate
selection

might
be
facilitated
if
occurring
along
Pmax.

Yet,
the
interpretation
of
any
clinal
variation
as
a
result
of
the
action
of
selection
has
been

challenged,
suggesting
that
isolation
by
distance
among
populations
connected
by
spatially
limited

gene
flow
might
produce
similar
clinal
variations
(Vasemägi
2006).
Evolution
along
Pmax
might

equally
facilitate
the
occurrence
of
a
morphological
latitudinal
cline
through
such
a
process.
By

favoring
morphological
evolution
along
few
directions,
due
to
selection
as
due
to
random
processes,

the
channeling
role
of
Pmax
makes
difficult
to
disentangle
both
processes
(Ackerman
and
Cheverud

2004).

Parallel
evolution
on
islands:
exemplary
cases
of
the
channeling
role
of
Pmax

Compared
to
the
relatively
limited
divergence
between
lineages
separated
since
one
million
years,

insular
evolution
displays
amazing
cases
of
molar
shape
divergence,
as
exemplified
on
Porquerolles,

despite
close
genetic
relatedness
with
wood
mice
from
the
nearby
mainland
(Michaux
et
al.
2002).

Insular
divergence,
however,
is
not
systematic
and
several
islands
display
no
or
weak
differentiation

(Noirmoutier,
Port-­‐Cros,
Ibiza,
Corsica).
In
most
of
these
cases,
these
islands
are
densely
populated,

for
some
of
them
since
antiquity
(Vigne
and
Valladas
1996),
and
the
frequency
of
human
travels

might
have
promoted
exchanges
among
wood
mice
populations
and
dampen
phenotypic
divergence.

In
contrast,
pronounced
insular
divergence
was
overall
displayed
on
small
islands
(Porquerolles,
~13