first published online 28 October 2009, doi: 10.1098/rspb.2009.12722772010Proc. R. Soc. B
John R. Stewart, Adrian M. Lister, Ian Barnes and Love Dalén
space and time
Refugia revisited: individualistic responses of species in
References
http://rspb.royalsocietypublishing.org/content/277/1682/661.full.html#related-urls
Article cited in:
http://rspb.royalsocietypublishing.org/content/277/1682/661.full.html#ref-list-1
This article cites 77 articles, 13 of which can be accessed free
Subject collections
(121 articles)molecular biology
(1430 articles)evolution
(1306 articles)ecology
Articles on similar topics can be found in the following collections
Email alerting service hereright-hand corner of the article or click
Receive free email alerts when new articles cite this article - sign up in the box at the top
http://rspb.royalsocietypublishing.org/subscriptionsgo to:Proc. R. Soc. BTo subscribe to
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
Review
Refugia revisited: individualistic responses
of species in space and time
John R. Stewart1, Adrian M. Lister1, Ian Barnes2 and Love Dale´n2,*,†
1
Department of Palaeontology, Natural History Museum, Cromwell Road, London SW7 5BD, UK
2
School of Biological Sciences, Royal Holloway University of London, Egham, Surrey TW20 0EX, UK
Climate change in the past has led to significant changes in species’ distributions. However, how individual
species respond to climate change depends largely on their adaptations and environmental tolerances.
In the Quaternary, temperate-adapted taxa are in general confined to refugia during glacials while coldadapted
taxa are in refugia during interglacials. In the Northern Hemisphere, evidence appears to be
mounting that in addition to traditional southern refugia for temperate species, cryptic refugia existed
in the North during glacials. Equivalent cryptic southern refugia, to the south of the more conventional
high-latitude polar refugia, exist in montane areas during periods of warm climate, such as the current
interglacial. There is also a continental/oceanic longitudinal gradient, which should be included in a
more complete consideration of the interaction between species ranges and climates. Overall, it seems
clear that there is large variation in both the size of refugia and the duration during which species are
confined to them. This has implications for the role of refugia in the evolution of species and their
genetic diversity.
Keywords: cryptic refugia; Quaternary; climate change; speciation; geographical distribution
1. INTRODUCTION
It has long been recognized that the geographical ranges
of species have expanded and contracted in a cyclical
manner according to the climatic changes of the Quaternary
(Darwin 1859, pp. 364–382; Hewitt 1996; Bennett &
Provan 2008). The places where species persist during
glaciations have generally been described as refugia. Isolation
within such glacial refugia, and the timing and
mode of expansion from them, have become topics of
increasing importance in our understanding of evolutionary
processes such as adaptation, speciation and
extinction. Understanding how species have responded
to past climate changes, and where they endured periods
of adverse climates, also has relevance for models forecasting
how current climate change may affect species.
The subject of Quaternary refugia is therefore of interest
to a variety of researchers including palaeoecologists,
population geneticists and conservation biologists.
Owing to the breadth of this array of interested scientists,
however, there is confusion about the meaning of the
refugium concept.
We propose here that Quaternary refugia should be
defined as the geographical region or regions that a
species inhabits during the period of a glacial/interglacial
cycle that represents the species’ maximum contraction in
geographical range. This is a flexible definition that
accommodates species that are adapted to different climatic
conditions, while at the same time highlighting
the idea that species in general respond to climatic
changes independently of each other (Taberlet et al.
1998; Stewart 2008; see also the literature on vegetational
change reviewed by Hewitt 1996). We nonetheless consider
it useful to identify and discuss different categories
of refugium, based both on general geographical location
and whether the refugium is inhabited by a temperate or a
cold-adapted species. The oceanic–continental gradient,
with its corresponding variation in key parameters, will
be considered in greater detail owing to its relevance to
organisms of the last cold stage of the Pleistocene. Furthermore,
we discuss the effects of differing refugial
sizes and varying lengths of time during which populations
are restricted to refugia. We also consider the
degree to which different species have the same refugia
and the fate of populations outside refugia during the
contraction phase. Finally, the role of refugia in species
evolution is discussed, with examples illustrating different
possible scenarios. We have concentrated on the mid- to
high-latitude Palaearctic as this area has a relatively
well-documented history and has been the locus of pioneering
studies on the biotic effects of glacial/interglacial
cycles (e.g. Hewitt 1996, 1999, 2000), while recognizing
that many analogous studies have been made in North
America (Avise 2000; Swenson & Howard 2005).
2. PREVIOUS USE OF THE REFUGIUM CONCEPT
The glacial refugium hypothesis has dominated studies of
ice age biogeography for some time (e.g. Holder et al.
1999). This concept sees the cold, glacial, phases of
Earth’s recent history as being the primary forcers of
population divergence and, in some cases, speciation
(Hewitt 1996). This emphasis on the cold phases comes
partly from the origin of the refugium concept, which
* Author for correspondence (love.dalen@nrm.se).
†
Present address: Molecular Systematics Laboratory, Swedish
Museum of Natural History, 104 05 Stockholm, Sweden.
Proc. R. Soc. B (2010) 277, 661–671
doi:10.1098/rspb.2009.1272
Published online 28 October 2009
Received 17 July 2009
Accepted 5 October 2009 661 This journal is q 2009 The Royal Society
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
arose from a consideration of the contraction phase of
vegetation during glacial stages (Bennett & Provan 2008).
It has also led to a general assumption that many organisms
are pushed southwards as the glaciated north becomes
inhospitable to many life forms. In recent phylogeographic
studies, different organisms have been shown to expand
out of various European peninsulae and other southern
refugia at the end of the last ice age (Taberlet et al. 1998;
Hewitt 1999, 2000, 2004). This picture was, however,
complicated by the suggestion that cryptic northern refugia
had existed in the Late Pleistocene for some temperate
organisms (Bilton et al. 1998; Willis et al. 2000; Stewart &
Lister 2001; Stewart 2003). In addition, it has been
argued that some, or most, peninsular populations were
areas of endemism rather than refugia (Bilton et al. 1998;
Stewart 2003; Bennett & Provan 2008; Bhagwat & Willis
2008; Provan & Bennett 2008).
Over recent years, the use of the glacial refugium concept
has broadened, and has frequently been applied to coldadapted
species such as lemmings (Fedorov & Stenseth
2001, 2002), rock ptarmigan (Lagopus muta; Holder et al.
1999), mountain sheep (Loehr et al. 2005), mountain
avens (Dryas integrifolia; Tremblay & Schoen 1999) and
white spruce (Picea glauca; Anderson et al. 2006). The problem,
in our view, with assigning glacial refugia to coldadapted
species is that they generally have a larger
distribution during cold stages than during periods of
warm climate (e.g. Stewart & Lister 2001; Dale´n et al.
2005; Stewart & Dale´n 2008). Although vicariance
events may have resulted from the growth of ice sheets
during cold stages, or rising sea levels during warm
periods, such isolated populations can hardly be viewed
as refugial, since the species would have had large distributions
elsewhere (Musil 1985; Tyrberg 1991; Stewart
et al. 2003, etc.). Instead, we argue that since the range
of cold-adapted species is at its minimum during periods
of warm climate, such species are in refugia during interglacials.
Our refugial concept also excludes ‘range shifts’
that do not entail a significant contraction of area; population
or genetic ‘bottlenecks’, which imply reduced
population size and will often result from refugial contraction
but do not entail it; populations left in isolation as the
species range contracts but that rapidly extirpate as climate
worsens; and species that permanently occupy a small
range. We also do not treat an area per se as a refugium
except insofar as it contains refugial populations of one
or more species.
An important category of refugia are the cryptic refugia
as defined by Stewart & Lister (2001). Cryptic refugia are
refugia situated at different latitudes or longitudes than
would normally be expected, and often resemble climatic
islands in which conditions differ favourably from the surrounding
areas. Rull (2009) has revived his concept of
‘macrorefugia’ and ‘microrefugia’ for larger and smaller
refugial areas, subsuming cryptic refugia within microrefugia.
However, his concept of microrefugia covers a
broader range of phenomena, including widespread but
low-density populations, and hypothesized large numbers
of small, isolated populations, than we consider here;
and his definition of microrefugia would include any
area with a small, isolated population, whereas we limit
the refugial concept to the contraction phase of a species’
expansion–contraction cycle. Finally, although (as
discussed below) cryptic refugia will often be smaller
than conventional refugia, small size is not integral to
their definition.
Recently, the value of continuing to use the refugium
concept has been challenged by Bennett & Provan
(2008), who point out that there are many and complex
ways in which species respond to climatic and environmental
change, and that the refugial concept has lacked
clear definition and has been used in confusingly different
ways. While we agree with those points, we believe that
the refugial concept, as we have defined and limited it,
remains important for a variety of reasons. The cycle of
expansion and contraction into refugia (as we define
them) has a particular importance in species-level evolution,
including its significance in determining the
pattern of genetic variation in a species. Issues of refugium
size are also important, both because they lead to
testable predictions of the effects of refugial contraction
across taxa and because of their relevance to extinction
risk, both in the past and in the conservation biology of
today. Finally, refugia remain important as the source
populations from which species expand their ranges at
the onset of more favourable conditions.
3. SPATIAL AND TEMPORAL CATEGORIES
OF REFUGIA
The concept of refugium used here is the area occupied
by an individual species, not the area occupied by a
whole community of species as in some studies (e.g.
Whittington-Jones et al. 2008). We propose to classify
refugia first from a temporal perspective, where species
can be broadly viewed as having either glacial or interglacial
refugia. Second, we divide refugia into further categories
based on their geographical location. The categories of
refugia, therefore, include the traditional southern refugia
and the equivalent polar refugia for cold-adapted species,
as well as cryptic refugia to the north or south of the
main areas into which populations contract (figure 1).
This classification is reminiscent of that used by
Thienemann (1950). We also introduce a new dimension,
the continental/oceanic gradient, and discuss refugia for
organisms adapted to these respective conditions.
Examples of each type of refugium are given below and
are listed in table 1. It is acknowledged that for some
species it may be difficult to determine whether they
had larger or smaller ranges during glacials or interglacials.
This would be particularly the case for taxa with
broad ecological ranges (e.g. the wolf Canis lupus), and
for species with meagre fossil records (e.g. many insects).
It should also be noted that the definitions given here are
for the Northern Hemisphere and that these would have
their mirror image in the Southern Hemisphere.
(a) Glacial refugia
(i) Southern refugia
These are the traditionally accepted refugia for temperate
species during glacial phases, which in general comprise
the southern portion of the species’ distribution during
warm climatic phases such as the current interglacial. In
Europe, southern refugia are generally located within
the Iberian, Italian and Balkan peninsulas. The identification
of these refugia was initially based on
palaeoecological evidence (Huntley & Birks 1983;
Bennett et al. 1991) and was later confirmed through
662 J. R. Stewart et al. Review. Refugia revisited
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
phylogeographic studies, which showed that many extant
populations further north are derived from southern
regions (Hewitt 1996, 1999, 2000). This pattern of glacial
survival in the South, followed by post-glacial
recolonization of northern regions, seems to be a general
pattern among a variety of temperate taxa, including
plants, insects and vertebrates (Hewitt 2001). However,
different species’ expanded populations seem to be
derived from different southern refugia, suggesting that
species have responded individualistically to the increases
Figure 1. Schematic map showing some types of refugia for Europe and western Asia. Interglacial refugia for cold-adapted
species are shown in blue, glacial refugia for temperate species in red. Long-term refugia, indicated by dark blue/red, are a
subset of all refugia that are inhabited throughout at least one full glacial/interglacial cycle. The areas shown in paler colour
are refugia in the sense that they are inhabited during the contraction phase, but are not inhabited during the expansion
phase owing to the spread of ice sheets during glacials (cold-adapted species), or excessive temperatures and/or too high aridity
during interglacials (temperate species). Also shown, in yellow, are interglacial refugia along the oceanic/continental gradient,
with a continental refugium in the east and cryptic refugia further west. The ice sheet for the Last Glacial Maximum is taken
from Ehlers & Gibbard (2004). The diagram is schematic; not all of the refugia would have been occupied simultaneously, but
the ranges are based on real examples taken from table 1.
Review. Refugia revisited J. R. Stewart et al. 663
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
in habitat availability brought on by the climatic changes
at the end of the last glaciation (Taberlet et al. 1998).
In the last decade, several studies have confirmed the
existence of similar southern refugia in North America
(e.g. Lacourse et al. 2005; Soltis et al. 2006), as well as
analogous northern refugia in the Southern Hemisphere
(Byrne 2008).
(ii) Cryptic northern refugia
Cryptic northern refugia are glacial refugia for temperate
taxa situated at higher latitudes than the expected areas of
suitable habitat to the South. The concept, as originally
conceived by Stewart & Lister (2001) was applied to
taxa that were not generally accepted as living in central
or northern Europe during the last glaciation. However,
the concept is in need of refinement as it has subsequently
been applied to non-temperate taxa (e.g. Pruett & Winker
2005). In fact, the original inclusion of pine (Pinus sylvestris)
living on the Norwegian coast during the last glaciation
(Stewart & Lister 2001), was already stretching the
definition of a cryptic northern refugium as this taxon is
not strictly temperate and is relatively cold-tolerant.
The cryptic northern refugium hypothesis has received
significant support since its publication, with phylogeographic
studies finding evidence for northern refugia in
various temperate organisms, including small mammals
(Wo´jcik et al. 2002; Jaarola & Searle 2003; Deffontaine
et al. 2005; Kotlik et al. 2006), ferns (Trewick et al.
2002), sedges (Tyler 2002a,b), snails (Haase & Bisenberger
2003; Pfenninger et al. 2003; Benke et al. 2009), freshwater
fishes (Ha¨nfling et al. 2002; Volckaert et al. 2002),
amphibians (Teacher et al. 2009) and reptiles (Carlsson
2003). Further evidence in support of the hypothesis
has come from the study of fossil pollen, plant macros
including wood charcoal, and mammals (Birks 2003;
Willis & van Andel 2004; Sommer & Nadachowski
2006; Caseldine et al. 2008).
(b) Interglacial refugia
(i) Polar refugia
Polar refugia are the high-latitude regions where coldadapted
species persist through interglacials. In the
Northern Hemisphere, polar refugia are located in the
northernmost parts of continental Eurasia and North
America, as well as in several islands in the Arctic
Ocean, for example Greenland, Svalbard, Wrangel
Island and the New Siberian Islands.
During the last glaciation, many cold-adapted species
had a larger distribution than they have today. Furthermore,
several studies on cold-adapted species have
identified genetic signatures of increase in population
size during the early stages of the last glaciation,
suggesting that these species had small population sizes
also during the last interglacial (Fedorov et al. 1999;
Flagstad & Røed 2003; Shapiro et al. 2004; Dale´n et al.
2005). It therefore seems likely that glacial expansions
and interglacial contractions were a recurrent pattern for
cold-adapted species during the Late Quaternary. This
supports the idea that cold-adapted species are in refugia
during interglacials, and thus respond to climatic changes
in the opposite way to temperate species. Several Arctic
species are at present confined exclusively to polar refugia,
for example Arctic fox (Alopex lagopus), lemmings
(Lemmus spp. and Dicrostonyx spp.), reindeer (Rangifer
tarandus) and muskox (Ovibos moschatus). The present
ranges of some of these species are large, but they are
still significantly reduced compared with their size
during the last glaciation.
(ii) Cryptic southern refugia
Cryptic southern refugia are interglacial refugia for coldadapted
species situated at lower latitudes. Today, and
presumably also during previous interglacials, these
refugia accommodate relict populations of formerly widespread
cold-adapted Pleistocene taxa such as mountain
Table 1. Classification of refugia together with examples identified by phylogeographic and other studies.
refugium type taxon study
temperate taxa in glacials
southern refugia U. arctos, Erinaceus sp., Chorthippus
parallelus, Quercus spp., Arvicola spp.,
Crocidura suaveolens, Triturus spp.,
Apodemus sylvaticus
Hewitt (1996, 1999, 2000), Taberlet et al.
(1998).
cryptic northern refugia Clethrionomys glareolus, Sorex araneus,
Microtus araneus, Asplenum ceterah,
Carex digitata
Wo´jcik et al. (2002), Trewick et al. (2002),
Tyler (2002a,b), Jaarola & Searle
(2003), Deffontaine et al. (2005) and
Teacher et al. (2009).
cold taxa in interglacials (e.g. present day)
polar refugia R. tarandus, A. lagopus, Dicrostonyx
spp., Lagopus spp.
Fedorov et al. (1999), Flagstad & Røed
(2003), Dale´n et al. (2005) and Stewart &
Dale´n (2008).
cryptic southern refugia L. muta, D. octopetala, B. nana,
H. lapponicus
Angus (1983) and Stewart & Lister
(2001).
continental-adapted taxa in interglacials (e.g. present day)
continental refugia Spermophilus spp., S. tatarica and
Ochotona spp.
no study but apparent from comparison
between fossil record and modern
distribution, e.g. Musil (1985).
cryptic refugia for continentaladapted
taxa (in more
oceanic areas)
S. subtilis, H. rhamnoides. no study but apparent from fossil record
and modern distribution, e.g.
Macdonald & Barrett (1993) and
Iversen (1973).
664 J. R. Stewart et al. Review. Refugia revisited
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
avens Dryas octopetala, dwarf birch Betula nana, rock
ptarmigan Lagopus mutus, Arctic hare Lepus timidus
and the water beetle Helophorus lapponicus (Angus
1983). Although not ‘cryptic’ in the original sense
(since their present ranges are well known), we nonetheless
retain the term to emphasize the parallel to cryptic
northern refugia. Refugial areas for these taxa generally
have a montane topography such as the Alps and
Pyrenees, although when situated at higher latitudes
the altitude can be lower. The Scottish highlands are
therefore included as a cryptic southern refugium for
rock ptarmigan L. mutus and red/willow grouse Lagopus
lagopus. Populations of cold-adapted species in cryptic
southern refugia are often surrounded during interglacials
by populations of temperate species that have
expanded from their glacial refugia. Most of the species
in these refugia also have disjunct distributions (polar
refugia) to the north (e.g. rock ptarmigan, mountain
avens, Arctic hare and H. lapponicus). Some, however,
are only known in the southern montane refugia (e.g.
ibex Capra ibex and chamois Rupicapra rupicapra) in
southern Europe, although they were also found in
more northerly rocky lowland areas, such as southern
Belgium (Stewart et al. 2003), during the Late
Pleistocene.
(c) The oceanic–continental gradient
One biogeographic dimension that is often ignored in discussions
on species’ response to the glacial/interglacial
cycle is the oceanic–continental axis. ‘Oceanic’ adaptation
implies more humid, less seasonably variable
climate; ‘continental’ adaptation, drier climate with
greater seasonal variation. This is often a longitudinal
perspective as opposed to the latitudinal aspect of northern
and southern refugia. We are unaware of any
phylogeographic studies explicitly dealing with this perspective
in detail. However, Eurasian mammal species
such as the ground squirrel Spermophilus spp., saiga antelope
Saiga tatarica and pika Ochotona spp. have more
restricted continental distributions in the Palaearctic
today, having had more extensive distributions, extending
to the British Isles, during parts of the last glaciation
(Musil 1985). In fact, it is this longitudinal gradient
that explains the expansion of steppic species and their
inclusion in the Late Pleistocene ‘steppe–tundra’
biome. We therefore propose that some species have
continental interglacial refugia.
Based on the existence of cryptic northern refugia for
temperate species and cryptic southern refugia for cold
species, it seems possible that a similar phenomenon
could exist along the longitudinal axis. For example,
one could expect species currently in eastern continental
refugia also to occur in cryptic refugia along the west
of Eurasia today. Possible examples are the southern
birch mouse (Sicista subtilis) in Hungary and Romania
(Macdonald & Barrett 1993) and some of the halophytic
plant species, such as sea buckthorn (Hippophae¨
rhamnoides), found along the Atlantic seaboard as well
as in the Asiatic steppe (Iversen 1973). Identifying
currently isolated populations as cryptic refugia requires
care, however, as in some cases their separation or even
existence may be the result of historical human impact
on the landscape. In theory, one might speculate that
these refugia would have a counterpart in refugia, and
cryptic refugia, for ‘oceanic-adapted’ species during
glacials, since the extension of arid climates during the
Late Pleistocene would have been as much of an impediment
to some taxa as the cold itself. However, we know of
no good example of a species that follows such a pattern.
The hippopotamus Hippopotamus amphibius, for example,
shows an oceanic distribution in the Palearctic during the
last interglacial, in that it spread no further east than
central Europe because of its intolerance of cold winters.
Its range contraction during the last glaciation, however,
was southward rather than westward, in this respect similar
to other temperate species occupying southern refugia.
The longitudinal and traditional latitudinal gradients
would thus work in tandem in defining the precise
location of a species’ refugium, depending on the species’
ecology. This agrees with the expectation that species will
respond individualistically, and not in concert, to climatic
changes (Taberlet et al. 1998).
4. OTHER CATEGORIES OF REFUGIUM
The question arises whether some temperate species
could be in refugia during warm periods such as interglacials,
and cold-adapted species during cold periods. As
discussed above, mere isolation is not enough to justify
describing a population as refugial. Hence, populations
of temperate species isolated on islands during interglacials,
or of cold-adapted species isolated through glacial
vicariance, for example along the coasts and on nunataks,
would not alone constitute refugial populations, as long as
total species range remained large. However, following
the definition of a refugium proposed in this paper,
some cold-adapted species may actually have a smaller
distribution during parts of glacials compared with interglacials
owing to the advance and retreat of glaciers and
continental ice sheets. In periglacial areas, habitat restriction
evidently contracted the ranges of even cold-adapted
species into refugia during the maximally severe phases
of glacial climate (e.g. woolly mammoths; Stuart
et al. 2004).
Some cold-adapted species endemic to mountainous
regions might constitute a further category. Montane
species generally have larger ranges during glacials when
they spread to lowland areas (Stewart et al. 2003). However,
some alpine species may be explicitly adapted to a
montane environment, and would thus not have
expanded into the surrounding lowlands during glacials.
Such alpine-adapted species could thus have larger distributions
during interglacials owing the expansion of
mountain glaciers during cold stages, and would instead
have been confined to nunataks or peripheral refugia
during glaciations (Holderegger & Thiel-Egenter 2008).
5. THE SIZE OF REFUGIA AND DURATION
OF OCCUPATION
Cryptic refugia are generally expected to be smaller than
the more traditional southern and polar refugia because
they are peripheral and are surrounded by unsuitable
habitats. Cryptic northern refugia may often comprise
sheltered habitats located in deeply incised valleys that
provided microclimates for temperate species, allowing
them to survive at latitudes where they would normally
Review. Refugia revisited J. R. Stewart et al. 665
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
have perished (Stewart & Lister 2001). Nonetheless,
recent work using back-casted species distribution
models has suggested relatively widespread distributions
for some small mammal species to the north of the traditional
southern refugia during the Last Glacial Maximum
(Fløjgaard et al. 2009).
Cryptic southern refugia are generally located in
mountainous regions, where the high altitude provides
cooler climates than the surrounding landscape, thus permitting
the local survival of cold-adapted species at
relatively low latitudes during interglacials. Most such
refugia will be smaller than the regions comprising polar
refugia, at least in the Northern Hemisphere, although
this is not always clear-cut since high-altitude plateaus
and mountain ranges can be quite large.
The Quaternary encompassed climatic cycles of differing
amplitudes and durations. This affected the length of
time temperate and cold-adapted species were confined to
refugia, which in turn can be expected to have had important
ecological and evolutionary consequences. One of
the major features of the Quaternary is the long-term climatic
cycling of the last 700 kyr. These cycles, with long
glaciations lasting up to 100 kyr and shorter interglacials
lasting 10–15 kyr, are thought to have been a major driving
force for population divergence in temperate species
(Hewitt 1996). However, embedded within these longscale
cycles are short-term climatic events that took
place on a millennial scale. These include warm interstadials
and particularly cold stadials during glacials, as well
as cool episodes during interglacials. These were first
documented through terrestrial pollen analyses and subsequently
correlated with detailed marine records
(Shackleton & Opdyke 1973; Tzedakis 1993). Since
then, further complexity has been detected in the cold
Heinrich events seen in marine sediments and the warm
Dansgaard–Oeschger events (Greenland interstadials)
identified in ice cores (Svensson et al. 2006).
The general expectation from the broad-scale pattern,
with long glacials and shorter interglacials, is that
temperate species spend much longer in refugia than
cold-adapted species. The effect of the shorter millennial-scale
fluctuations in climate is less clear, since these
are difficult to identify with precision in the terrestrial
record. However, it is probable that they also caused ecological
disturbances and gave rise to shorter term episodes
of refugial isolation and range expansions of temperate
and cold-adapted species during stadials and interstadials.
On the other hand, the duration of these
fluctuations may in some cases have been so short that,
even when climatically favourable, many species did not
expand their range fully. For example, slow-moving temperate
species in the Northern Hemisphere were probably
not able to expand as far north during interstadials
as expected, and vice versa for cold-adapted species
during stadials.
From an evolutionary perspective, the most important
refugial areas are geographical regions where a species has
persisted throughout a series of full glacial/interglacial
cycles (each 100–120 kyr in duration), since each full
cycle will usually have included a species’ maximum
and minimum distributions. The locations that a species
inhabits continuously for at least one full glacial/
interglacial cycle can be viewed as constituting ‘longterm
refugia’ (Stewart & Dale´n 2008), and we expect
that they will tend to harbour the greatest level of genetic
diversity within the species’ range. However, it is important
to note that not all refugia, as defined earlier, will
constitute long-term refugia. For example, many Arctic
regions that are inhabited by cold-adapted species
during interglacials, and thus are refugia by our definition,
are made uninhabitable during glacials by
advancing ice sheets (see earlier). In the same way,
some southern glacial refugia may become too hot or
arid for a temperate species to persist there during interglacials.
In the case of cryptic refugia, some may operate
over several climatic cycles, others only for one. This
will vary with the niche of the particular species and the
geographical and climatic characteristics of the area that
formed the refugium. Consequently, what we here
define as long-term refugia will represent a subset of all
refugia, and will include both polar and southern refugia,
as well as some cryptic refugia.
The combination of refugial size and duration has
implications for species persistence. The reason for this
is that a population’s ability to persist throughout a
period of adverse climate (cold or warm depending on
the species’ ecology), without becoming extinct owing
to demographic stochasticity or inbreeding (Lande
1988), depends on the putative refugium’s carrying
capacity as well as the duration of the climatic stage.
The probability that a population will survive throughout
a period of adverse climate is therefore inversely related to
the duration of confinement, and positively related to the
size of the potential refugium. This suggests that it could
be possible to define a ‘minimum refugium size’ required
for species persistence, which would be dependent on the
climatic interval (100 kyr glaciations, 10–15 kyr interglacials
or millennial-scale stadials/interstadials), as well as
the space requirements of the species in question
(figure 2). Some general predictions arise from this concept,
particularly for animals. First, one would expect
cryptic refugia, owing to their relatively small size, to be
less common for large-bodied species since the carrying
capacity is generally lower for such species. Second,
species with a large body size would be less likely to persist
in cryptic northern refugia compared with those in
cryptic southern refugia, since glaciations are normally
an order of magnitude longer than interglacials. Third,
one would expect a trophic effect, where species with a
high trophic level (e.g. carnivores), regardless of their
body size, would be less likely to survive through ‘adverse’
climatic periods in small patches of suitable habitat
(figure 2), because of low population size and limited
food base, a concept familiar from island biogeography
(MacArthur & Wilson 1967).
6. THE ROLE OF SPECIES INDEPENDENCE AND
THE CONGRUENCE OF REFUGIA
Recently, the individualistic (or independent) response of
species to climate change over several glacial cycles has
been discussed (Stewart 2008). This independence has
implications for the congruence of refugia for different
species. Clearly, refugia will often be congruent owing
to their similar climatic and environmental requirements,
rather than any species-specific interdependence.
Congruence can also occur when species have similar discontinuous
ranges resulting from different histories
666 J. R. Stewart et al. Review. Refugia revisited
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
(Soltis et al. 2006). Coevolved relationships may lead to
stronger congruence of distributional history, for example
between some insects and their food plants, or parasites
and their hosts. In many cases, however, species are
believed not to be highly interdependent. This suggests,
for instance, that the existence of trees need not be
accompanied by the herbivores often associated with
them. A small stand of deciduous trees in the north of
Europe during a glacial is unlikely to be accompanied
by the whole ecosystem associated with a deciduous
forest biome in the area today. However, if it is associated
with some species, it needs to satisfy their habitat
tolerances as well as being large enough, with an adequate
carrying capacity for the species. This variation in community
composition, together with the geographically
isolated nature of cryptic refugia, again recalls island
biogeography and can be expected to promote ecologically
adaptive evolution (Hewitt 1996, 1999; Stewart
2008).
7. THE FATE OF POPULATIONS
OUTSIDE REFUGIA
There is an outstanding question about the fate of
populations outside refugia when climatic changes lead
to refugial confinement. Bennett et al. (1991) used
data from the pollen record to show that many tree
populations in northern Europe became extinct at the
onset of the last glaciation. Correspondingly, a recent
study by Dale´n et al. (2007) showed that southern populations
of the Arctic fox (A. lagopus) did not contribute
genetically to present-day populations when temperatures
increased at the end of the last glaciation. The
results from these studies suggest that expanded populations
become extinct instead of tracking retreating
habitats by physically moving into the refugium
(Hewitt 1993, 1996; Lister 1997; Stewart et al. 2003;
Dale´n et al. 2007). This implies that populations in
long-term refugia are descended from individuals that
are already in place during the expansion phase, and
consequently that populations outside refugia make
little or no contribution to the long-term evolution of
the species. It should, however, be noted that the fate
of extra-refugial populations has only been investigated
in a limited number of taxa (Bennett et al. 1991;
Dale´n et al. 2007), and thus would benefit from further
study. Nonetheless, we note that rarity of habitat tracking
has the potential to explain the high degree of
population turnover described in several recent ancient
DNA studies (Barnes et al. 2002; Hofreiter et al. 2007;
Leonard et al. 2007). It also limits the value of phylogeographic
studies of taxa currently in refugia (i.e. Arctic
taxa). This is because it is not possible to analyse the
previously expanded populations without using ancient
DNA. A failure to take account of recently extirpated
populations may lead to erroneous conclusions.
The ultimate fate of a species in a contraction phase
may be complete extinction, and because species’ ranges
tend to contract in the direction of their refugia, the
long-term refugial areas will often be the eventual location
of the terminal populations (von Koenigswald 1999;
Lister & Stuart 2008).
8. EVOLUTION AND SPECIATION
It is clear that refugial phases are times when populations
will be in isolation and hence more prone to evolutionary
divergence. As discussed earlier, different types of refugia
have different characteristics that lend themselves to
hypotheses of population differentiation and even speciation.
These factors will be discussed in turn.
The first factor to consider is that glacials—considered
broadly as even-numbered marine oxygen isotopic
stages—are longer than interglacials, so that cold-adapted
and temperate species have been restricted to refugia for
different lengths of time. In addition, polar refugia,
being situated near the poles, will tend to be geographically
much closer together compared with southern
refugia. These two factors lead to an expectation of less
population divergence between populations in polar
refugia than those in southern refugia, in turn suggesting
that the opportunity for population differentiation is
greater in temperate species.
The question then arises whether refugial isolation can
lead to speciation. Lister (2004) concluded that several
hundred thousand years of isolation are normally required
for speciation to occur among mammals, although exceptions
exist in other groups (e.g. cichlid fishes; Johnson
et al. 1996). This implies that refugial isolation during
one glacial cycle would often be insufficient for speciation
to take place. However, it is tempting to speculate that
isolation in cryptic refugia could occasionally lead to
this kind of rapid evolution, as these populations fulfil several
of the requirements for allopatric speciation (Mayr
1954; Eldredge & Gould 1972), particularly ‘ecological’
speciation under strong adaptive selection (Hendry et al.
2007; Nosil et al. 2009). First, populations in glacial cryptic
northern and interglacial cryptic southern refugia,
occupying ‘pockets’ of suitable habitat in otherwise
unfavourable regions, are likely often to be smaller than
0
1
10
100
size of potential refugium
timetoextinction(kyr)
Figure 2. Conceptual figure showing the relationship
between size of a potential refugium and time to extinction
of the population (caused by demographic or genetic stochasticity).
Each line represents a range of areas occupied
by populations of a given species. The time required for
population extirpation is dependent on the size of the habitat
patch, and the horizontal dashed lines indicate the minimum
viable refugium size, i.e. the relative sizes required to survive
millennial-scale events, 10 kyr interglacials and 100 kyr glaciations.
The slope of the curve depends on several factors
such as body size, generation length and, as illustrated
here, trophic level. Dash-dotted curve, trophic level 0; solid
curve, trophic level 1.
Review. Refugia revisited J. R. Stewart et al. 667
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
their counterparts in southern/polar refugia. Second, it is
likely that the populations in these cryptic refugia are subjected
to different selective pressures than the populations
inhabiting the more traditional refugia, especially when
the refugial isolation is accompanied by a change in
climate. Third, extinction of predators and competitors
owing to environmental change and small patch size
could change community structure and thus alter the
species’ realized niche and hence the selective pressures
on species—the ‘New Neighbour’ hypothesis of Hewitt
(1996, 2000, 2001) and (Stewart 2008, in press). These
processes could lead to rapid adaptive divergence and, if
reproductive isolation were underway before refugial
populations expanded and met on climatic amelioration,
constitute the first steps towards speciation.
The potential for speciation would be higher for populations
in cryptic northern refugia during the longer
glacials than for cryptic southern refugial populations
during the shorter interglacials. It is therefore interesting
to consider the role of cryptic northern refugia for the
evolution of Arctic species, such as polar bear from
brown bear and the Arctic fox from swift fox. The phylogenetic
evidence clearly shows that the polar bear (Ursus
maritimus) evolved from the brown bear (Ursus arctos)
and that this probably happened within the last
200–300 kyr (Talbot & Shields 1996; Ho et al. 2008).
For the polar bear to have evolved its unique adaptations,
a brown bear population must have become exposed to
northern climates and oceanic habitats. We suggest that
such a population is likely to have lived in a cryptic northern
refugium as this would have provided an opportunity
for allopatric speciation in isolation from other brown
bears. Similarly, the origin of the Arctic fox (A. lagopus)
from the temperate swift fox (Vulpes velox) is thought to
have happened at approximately the same time as
(Geffen et al. 1992), or slightly earlier than (Sher
1986), the polar bear evolved from the brown bear.
Similar evolution is presumably ongoing in the cryptic
southern refugia of the Alps and Pyrenees today where,
for example, distinct subspecies of rock ptarmigan have
evolved (L. m. helvetica and L. m. pyrenaicus). It is less
likely, however, that temperate species (rather than just
subspecies) have evolved from arctic ancestors in cryptic
southern refugia, both because interglacials are shorter
than glacials and because there is greater niche occupancy
in the species-rich, long-held temperate regions compared
with the relatively recently originated arctic zone.
Another speciation scenario that has been invoked is
the differentiation within temperate species while in
different southern refugia (Hewitt 1996; Lister 2004).
Quaternary glacial cycles are believed to have promoted
population divergence, and sometimes even speciation,
among populations in the Mediterranean peninsulae, for
example, although this seems to have required isolation
on time scales exceeding a single glaciation (Hewitt
1996). Such isolation between southern refugia, despite
mixing of lineages further north during interglacials, is
thought to have been made possible by a lack of habitat
tracking at the onset of glaciations (Hewitt 1999). It has
also been pointed out that many species never expanded
from their southern ranges, which allowed for long-term
isolation between populations (Bilton et al. 1998),
although by our definition these are not refugia.
Speciation among populations in southern refugia might
generally be slower than in cryptic northern refugia, as
their population sizes are likely to be larger because of
the broader geographical area, which could have a tempering
effect on the rate of adaptive divergence (Mayr
1954). The importance of small population size in
evolution is, however, a topic of debate (Barton &
Charlesworth 1984; Coyne & Orr 2004). Furthermore,
since different southern refugia lie on approximately the
same latitude, their populations may have been subject to
similar selection pressures, especially given the additional
buffering effect provided by the complex topography of
many southern refugia (Tzedakis et al. 2002).
9. CONCLUSION
The subject of refugia is relevant to many areas of
ecology and evolutionary biology. Furthermore, the
individualistic nature of species’ responses to climate
change implies that the location of refugia varies according
to the climate as well as to the adaptations of
individual species or populations. We therefore suggest
that, in general, refugia can be classified as either glacial
or interglacial refugia. Glacial southern refugia are the
traditional low-latitude refugia for temperate taxa best
known from the work of Hewitt (1996, 1999, 2000),
whereas interglacial polar refugia harbour cold-adapted
taxa at high latitudes during warm periods, such as the
interglacial we are in today. However, owing to the complex
structure of environments and habitats across
space, it is also proposed that cryptic northern refugia
exist during glacials and equivalent cryptic southern refugia
during interglacials. The existence of the former has
had increasing support from phylogeographic (including
ancient DNA) studies of a wide range of organisms,
while cryptic southern refugia can be seen in areas such
as the Alps today. An additional dimension is the oceanic/continental
gradient, with continental-adapted taxa
in refugia during interglacials.
In general, cryptic refugia are smaller in size than
southern glacial or polar interglacial refugia. Furthermore,
the length of time during which organisms are
isolated in refugia differs between cold and temperate
taxa, since glacials are longer than interglacials. This disparity
among different types of refugia, as well as the
individualistic nature of species’ responses to climate
change, has several evolutionary implications. For
example, individualism may lead to new ecological associations
and interactions, which in themselves can impose
novel selective pressures on populations (Hewitt 1996,
2000, 2001), Stewart (in press). Also, populations inhabiting
small refugia, such as populations of temperate
species in cryptic northern refugia, are more likely to
become extinct. Such populations would not then act as
sources of expansion on climatic amelioration (Sommer &
Zachos in press). Paradoxically, however, long-term isolation
of small populations can lead to rapid population
divergence. Combined with the novel selection pressures
in peripheral refugia, this could potentially lead to speciation.
We therefore propose that isolation of temperate
species in cryptic northern refugia may have played an
important role in the origin of Arctic species.
L.D. acknowledges support from the Marie Curie Actions
grant FP6 041 545. We would like to thank Godfrey Hewitt
and two anonymous reviewers for their helpful comments.
668 J. R. Stewart et al. Review. Refugia revisited
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
REFERENCES
Anderson, L. L., Sheng Hu, F., Nelson, D. M., Petit, R. J. &
Paige, K. N. 2006 Ice-age endurance: DNA evidence
of a white spruce refugium in Alaska. Proc. Natl
Acad. Sci. USA 13, 12 447–12 450. (doi:10.1073/pnas.
0605310103)
Angus, R. B. 1983 Evolutionary stability since the Pleistocene
illustrated by reproductive compatibility between
Swedish and Spanish Helophorus lapponicus Thomson
(Coleoptera, Hydrophilidae). Biol. J. Linn. Soc. 19,
17–25. (doi:10.1111/j.1095-8312.1983.tb00773.x)
Avise, J. C. 2000 Phylogeography: the history and formation of
species. Cambridge, MA: Harvard University Press.
Barnes, I., Matheus, P., Shapiro, B., Jensen, D. & Cooper, A.
2002 Dynamics of Pleistocene population extinctions in
Beringian brown bears. Science 295, 2267–2270.
(doi:10.1126/science.1067814)
Barton, N. H. & Charlesworth, B. 1984 Genetic revolutions,
founder effects, and speciation. Annu. Rev. Ecol. Syst. 15,
133–164. (doi:10.1146/annurev.es.15.110184.001025)
Benke, M., Bra¨ndle, M., Albrecht, C. & Wilke, T. 2009
Pleistocene phylogeography and phylogenetic concordance
in cold-adapted spring snails (Bythinella spp.).
Mol. Ecol. 18, 890–903. (doi:10.1111/j.1365-294X.
2008.04073.x)
Bennett, K. D. & Provan, J. 2008 What do we mean by
‘refugia’? Quat. Sci. Rev. 27, 2449–2455. (doi:10.1016/
j.quascirev.2008.08.019)
Bennett, K. D., Tzedakis, P. C. & Willis, K. J. 1991 Quaternary
refugia of north European trees. J. Biogeogr. 18,
103–115. (doi:10.2307/2845248)
Bhagwat, S. A. & Willis, K. J. 2008 Species persistence in
northerly glacial refugia of Europe: a matter of chance
or biogeographical traits? J. Biogeogr. 35, 464–482.
(doi:10.1111/j.1365–2699.2007.01861.x)
Bilton, D. T., Mirol, P. M., Mascheretti, S., Fredga, K.,
Zima, J. & Searle, J. B. 1998 Mediterranean Europe as
an area of endemism for small mammals rather than a
source for northwards postglacial colonization.
Proc. R. Soc. Lond. B 265, 1219–1226. (doi:10.1098/
rspb.1998.0423)
Birks, H. H. 2003 The importance of plant macrofossils in
the reconstruction of Lateglacial vegetation and climate:
examples from Scotland, western Norway and Minnesota,
USA. Quat. Sci. Rev. 22, 453–473. (doi:10.1016/S0277-
3791(02)00248-2)
Byrne, M. 2008 Evidence for the multiple refugia at different
time scales during Pleistocene climatic oscillations in
southern Australia inferred from phylogeography. Quat.
Sci. Rev. 27, 2576–2585. (doi:10.1016/j.quascirev.2008.
08.032)
Carlsson, M. 2003 Phylogeography of the adder, Vipera
berus. In Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Sciences and Technology
849. Uppsala, Sweden: Acta Universitatis Upsaliensis.
Caseldine, C. J., McGarry, S. N. F., Baker, A., Hawksworth,
C. & Smart, P. L. 2008 Late Quaternary speleothem
pollen in the British Isles. J. Quat. Sci. 23, 193–200.
(doi:10.1002/jqs.1121)
Coyne, J. A. & Orr, H. A. 2004 Speciation. Sunderland, MA:
Sinauer Associates.
Dale´n, L., Fuglei, E., Hersteinsson, P., Kapel, C. M. O.,
Roth, J. D., Samelius, G., Tannerfeld, M. & Angerbjo¨rn, A.
2005 Population history and genetic structure of a
circumpolar species: the Arctic fox. Biol. J. Linn. Soc. 84,
79–89. (doi:10.1111/j.1095-8312.2005.00415.x)
Dale´n, L., Nystro¨m, V., Valdiosera, C., Germonpre´, M.,
Sablin, M., Turner, E., Angerbjo¨rn, A., Arsuaga, J. L. &
Go¨therstro¨m, A. 2007 Ancient DNA reveals lack of
postglacial habitat tracking in the Arctic fox. Proc. Natl
Acad. Sci. USA 104, 6726–6729. (doi:10.1073/pnas.
0701341104)
Darwin, C. R. 1859 On the origin of species by means of natural
selection, or the preservation of favoured races in the struggle
for life, 1st edn. London, UK: John Murray.
Deffontaine, V., Libois, R., Kotlı´k, P., Sommer, R.,
Nieberding, C., Pradis, E., Searle, J. B. & Michaux,
J. R. 2005 Beyond the Mediterranean peninsulas: evidence
of central European glacial refugia for temperate
forest mammal species, the bank vole (Clethrionomys
glareolus). Mol. Ecol. 14, 1727–1739. (doi:10.1111/
j.1365-294X.2005.02506.x)
Ehlers, J. & Gibbard, P. 2004 Quaternary glaciations—extent
and chronology. London, UK: Elsevier.
Eldredge, N. & Gould, S. J. 1972 Punctuated equilibria: an
alternative to phyletic gradualism. In Models in paleobiology
(ed. T. J. M. Schopf), pp. 82–115. New York, NY:
Freeman, Cooper & Co.
Fedorov, V. B. & Stenseth, N. C. 2001 Glacial survival of the
Norwegian lemming (Lemmus lemmus) in Scandinavia:
inference from mitochondrial DNA variation.
Proc. R. Soc. Lond. B 268, 809–814. (doi:10.1098/rspb.
2001.1584)
Fedorov, V. B. & Stenseth, N. C. 2002 Multiple glacial refugia
in the North American Arctic: inference from
phylogeography of the collared lemming (Dicrostonyx
groenlandicus). Proc. R. Soc. Lond. B 269, 2071–2077.
(doi:10.1098/rspb.2002.2126)
Fedorov, V. B., Fredga, K. & Jarrell, G. H. 1999 Mitochondrial
DNA variation and the evolutionary history of
chromosome races of collared lemmings (Dicrostonyx) in
the Eurasian Arctic. J. Evol. Biol. 12, 134–145. (doi:10.
1046/j.1420-9101.1999.00017.x)
Flagstad, Ø. & Røed, K. H. 2003 Refugial origins of reindeer
(Rangifer tarandus L.) inferred from mitochondrial DNA
sequences. Evolution 57, 658–670. (doi:10.1554/0014-
3820(2003)057[0658:ROORRT]2.0.CO;2)
Fløjgaard, C., Normand, S., Skov, F. & Svenning, C.
2009 Ice age distributions of European small
mammals: insights from species distribution modelling.
J. Biogeogr. 36, 1152–1163. (doi:10.1111/j.1365-2699.
2009.02089.x)
Geffen, E., Mercure, A., Girman, D. J., Macdonald, D. W. &
Wayne, R. K. 1992 Phylogenetic relationships of the foxlike
canids: mitochondrial DNA restriction fragment, site
and cytochrome b sequence analyses. J. Zool. 228, 27–39.
(doi:10.1111/j.1469-7998.1992.tb04430.x)
Haase, M. & Bisenberger, A. 2003 Allozymic differentiation
in the land snail Arianta arbustorum (Stylommatophora,
Helicidae): historical inferences. J. Zool. Syst. Evol. Res.
41, 175–185. (doi:10.1046/j.1439-0469.2003.00208.x)
Ha¨nfling, B., Hellemans, B., Volckaert, F. A. M. &
Carvallho, G. R. 2002 Late glacial history of the coldadapted
freshwater fish Cottus gobio, revealed by
microsatellites. Mol. Ecol. 11, 1717–1729. (doi:10.1046/
j.1365-294X.2002.01563.x)
Hendry, A. P., Nosil, P. & Rieseberg, L. H. 2007 The speed
of ecological speciation. Funct. Ecol. 21, 455–464.
(doi:10.1111/j.1365-2435.2006.01240.x)
Hewitt, G. 1993 Postglacial distribution and species
substructure: lessons from pollen, insects and hybrid
zones. In Evolutionary patterns and processes (eds D. R.
Lees & D. Edwards), pp. 97–123. London, UK:
Academic Press.
Hewitt, G. 1996 Some genetic consequences of ice ages,
and their role in divergence and speciation. Biol. J.
Linn. Soc. 58, 247–276. (doi:10.1111/j.1095–8312.
1996.tb01434.x)
Review. Refugia revisited J. R. Stewart et al. 669
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
Hewitt, G. 1999 Post-glacial recolonisation of European
biota. Biol. J. Linn. Soc. 68, 87–112. (doi:10.1111/
j.1095-8312.1999.tb01160.x)
Hewitt, G. 2000 The genetic legacy of the Quaternary ice
ages. Nature 405, 907–913. (doi:10.1038/35016000)
Hewitt, G. M. 2001 Speciation, hybrid zones and
phylogeography—or seeing genes in space and time.
Mol. Ecol. 10, 537–549. (doi:10.1111/j.1365–294X.
2001.01202.x)
Hewitt, G. M. 2004 Genetic consequences of climatic oscillations
in the Quaternary. Phil. Trans. R. Soc. Lond. B 359,
183–195. (doi:10.1098/rstb.2003.1388)
Ho, S. Y. W., Saarma, U., Barnet, R., Haile, J. & Shapiro, B.
2008 The effect of inappropriate calibration: three case
studies in molecular ecology. PLoS One 3, e1615.
(doi:10.1371/journal.pone.0001615)
Hofreiter, M., Mu¨nzel, S., Conard, N. J., Pollack, J., Slatkin,
M., Weiss, G. & Pa¨a¨bo, S. 2007 Sudden replacement of
cave bear mitochondrial DNA in the Late Pleistocene. Curr.
Biol. 17, R122–R123. (doi:10.1016/j.cub.2007.01.026)
Holder, K., Montgomerie, R. & Friesen, V. L. 1999 A test of
the glacial refugium hypothesis using patterns of mitochondrial
and nuclear DNA variation in rock ptarmigan
(Lagopus mutus). Evolution 53, 1936–1950. (doi:10.
2307/2640452)
Holderegger, R. & Thiel-Egenter, C. 2008 A discussion of
different types of glacial refugia used in mountain biogeography
and phylogeography. J. Biogeogr. 36, 476–480.
(doi:10.1111/j.1365-2699.2008.02027.x)
Huntley, B. & Birks, H. J. B. 1983 An atlas of past and present
pollen maps of Europe, 0–13,000 years ago. Cambridge,
UK: Cambridge University Press.
Iversen, J. 1973 The development of Denmark’s nature since the
last glacial. Danm. Geol. Unders. V 7C, 1–126.
Jaarola, M. & Searle, J. B. 2003 Phylogeography of field voles
(Microtus agrestis) in Eurasia inferred from mitochondrial
DNA sequences. Mol. Ecol. 11, 2613–2621. (doi:10.
1046/j.1365–294X.2002.01639.x)
Johnson, T. C., Scholz, C. A., Talbot, M. R., Kelts, K.,
Ricketts, R. D., Gideon, N., Beuning, K., Ssemmanda,
I. & McGill, J. W. 1996 Late Pleistocene desiccation of
Lake Victoria and rapid evolution of cichlid fishes. Science
273, 1091–1093. (doi:10.1126/science.273.5278.1091)
Kotlik, P., Deffontaine, V., Mascheretti, S., Zima, J.,
Michaux, J. R. & Searle, J. B. 2006 A northern glacial
refugium for bank voles (Clethrionomys glareolus). Proc.
Natl Acad. Sci. USA 103, 14 860–14 864. (doi:10.1073/
pnas.0603237103)
Lacourse, T., Mathewes, R. W. & Fedje, D. W. 2005 Lateglacial
vegetation dynamics of the Queen Charlotte
Islands and adjacent continental shelf, British Columbia,
Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol. 226,
36–57. (doi:10.1016/j.palaeo.2005.05.003)
Lande, R. 1988 Genetics and demography in biological conservation.
Science 241, 1455–1460. (doi:10.1126/science.
3420403)
Leonard, J. A., Vila´, C., Fox-Dobbs, K., Koch, P. L., Wayne,
R. K. & Van Valkenberg, B. 2007 Megafaunal extinctions
and the disappearance of a specialized wolf ecomorph.
Curr. Biol. 17, 1146–1150. (doi:10.1016/j.cub.2007.
05.072)
Lister, A. M. 1997 The evolutionary response of vertebrates
to Quaternary environmental change. In Past and future
rapid environmental changes: the spatial and evolutionary
responses of terrestrial biota (eds B. Huntley, W. Cramer,
A. V. Morgan, H. C. Prentice & J. R. M. Allen), pp. 287–
302. NATOASI Series I, no. 47. Berlin, Germany: Springer.
Lister, A. M. 2004 The impact of Quaternary Ice Ages on
mammalian evolution. Phil. Trans. R. Soc. Lond. B 359,
221–241. (doi:10.1098/rstb.2003.1436)
Lister, A. M. & Stuart, A. J. 2008 The impact of climate
change on large mammal distribution and extinction: evidence
from the last glacial/interglacial transition.
C. R. Geosci. 340, 615–620. (doi:10.1016/j.crte.2008.
04.001)
Loehr, J., Worley, K., Grapputo, A., Carey, J., Veitch, A. &
Coltman, D. W. 2005 Evidence for cryptic glacial refugia
from North American mountain sheep mitochondrial
DNA. J. Evol. Biol. 19, 419–430. (doi:10.1111/j.1420-
9101.2005.01027.x)
MacArthur, R. H. & Wilson, E. O. 1967 The Theory of island
biogeography. Princeton, NJ: Princeton University Press.
Macdonald, D. & Barrett, P. 1993 Mammals of Britain and
Europe. London, UK: Harper Collins Publishers.
Mayr, E. 1954 Change of genetic environment and evolution.
In Evolution as a process (eds J. Huxley, A. C.
Hardy & E. B. Ford), pp. 157–180. London, UK:
Allen & Unwin.
Musil, R. 1985 Paleobiogeography of terrestrial communities
in Europe during the Last Glacial. Acta Musei Nationalis
Pragae XLI B, pp. 1–2, 1–83.
Nosil, P., Harmon, L. J. & Seehausen, O. 2009 Ecological
explanations for (incomplete) speciation. Trends Ecol.
Evol. 24, 145–156. (doi:10.1016/j.tree.2008.10.011)
Pfenninger, M., Posada, D. & Magnin, F. 2003 Evidence
for the survival of the Pleistocene climatic change in
Northern refugia by the land snail Trochoidea geyeri
(Soo´s 1926) (Helicellinae, Stylommatophora. BMC
Evol. Biol. 3, 8. (doi:10.1186/1471-2148-3-8)
Provan, J. & Bennett, K. D. 2008 Phylogeographic insights
into cryptic glacial refugia. Trends Ecol. Evol. 23,
564–571. (doi:10.1016/j.tree.2008.06.010)
Pruett, C. L. & Winker, K. 2005 Biological impacts of climatic
change on a Beringian endemic: cryptic refugia in
the establishment and differentiation of the rock sandpiper
(Calidris ptilocnemis). Clim. Change 68, 219–240.
(doi:10.1007/s10584-005-158-4)
Rull, V. 2009 Microrefugia. J. Biogeogr. 36, 481–484.
(doi:10.1111/j.1365–2699.2008.02023.x)
Shackleton, N. J. & Opdyke, N. D. 1973 Oxygen isotope and
palaeomagnetic stratigraphy of Equatorial Pacific core
V28–238: Oxygen isotope temperatures and ice volumes
on a 105
year and 106
year scale. Quat. Res. 3, 39–55.
(doi:10.1016/0033-5894(73)90052-5)
Shapiro, E. et al. 2004 Rise and fall of the Beringian steppe
bison. Science 306, 1561–1565. (doi:10.1126/science.
1101074)
Sher, A. V. 1986 Olyorian land mammal age of northeastern
Siberia. Paleontogr. Ital. 74, 97–112.
Soltis, D. E., Morris, A. B., McLachlan, J. S., Manos, P. S. &
Soltis, P. S. 2006 Comparative phylogeography of
unglaciated eastern North America. Mol. Ecol. 15,
4261–4293. (doi:10.1111/j.1365-294X.2006.03061.x)
Sommer, R. S. & Nadachowski, A. 2006 Glacial refugia of
mammals in Europe: evidence from fossil records.
Mammal Rev. 36, 251–265. (doi:10.1111/j.1365-2907.
2006.00093.x)
Sommer, R. S. & Zachos, F. E. 2009 Fossil evidence and
phylogeography of temperate species: ‘glacial refugia’
and post-glacial recolonization. J. Biogeog. 36,
2013–2020. (doi:10.1111/j.1365-2699.2009.02187.x)
Stewart, J. R. 2003 Comment on ‘Buffered tree population
changes in a Quaternary refugium: Evolutionary implications’.
Science 299, 825a. (doi:10.1126/science.1079388)
Stewart, J. R. 2008 The progressive effect of the individualistic
response of species to Quaternary climate change:
an analysis of British mammalian faunas. Quat. Sci. Rev.
27, 2499–2508. (doi:10.1016/j.quascirev.2008.08.003)
Stewart, J. R. In press. The evolutionary consequence of the
individualistic response to climate change. J. Evol. Biol.
670 J. R. Stewart et al. Review. Refugia revisited
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from
Stewart, J. R. & Dale´n, L. 2008 Is the glacial refugium concept
relevant for northern species? A comment on Pruett
and Winker 2005. Clim. Change 86, 1–2. (doi:10.1007/
s10584-007-9366-9)
Stewart, J. R. & Lister, A. M. 2001 Cryptic northern refugia
and the origins of modern biota. Trends Ecol. Evol. 16,
608–613. (doi:10.1016/S0169-5347(01)02338-2)
Stewart, J. R., van Kolfschoten, M., Markova, A. & Musil, R.
2003 The mammalian faunas of Europe during oxygen
isotope stage three. In Neanderthals and modern humans
in the European landscape during the last glaciation, 60,000
to 20,000 years ago: archaeological results of the stage 3 project
(eds T. H. van Andel & S. W. Davies), pp. 103–129.
McDonald Institute Monograph Series. Cambridge,
MA: McDonald Institute.
Stuart, A. J., Kosintsev, P. A., Higham, T. F. G. & Lister,
A. M. 2004 Pleistocene and Holocene extinction
dynamics in giant deer and woolly mammoth. Nature
431, 684–689. (doi:10.1038/nature02890)
Svensson, A. et al. 2006 The Greenland Ice Core Chronology
2005, 15–42 ka. Part 2: comparison to other
records. Quat. Sci. Rev. 25, 3258–3267. (doi:10.1016/
j.quascirev.2006.08.003)
Swenson, G. & Howard, D. J. 2005 Clustering of contact
zones, hybrid zones, and phylogeographic breaks in
North America. Am. Nat. 166, 581–591. (doi:10.1086/
491688)
Taberlet, P., Fumagalli, L., Wust-Saucy, A. G. & Cosson, J.
F. 1998 Comparative phylogeography and postglacial
colonization routes in Europe. Mol. Ecol. 7, 453–464.
(doi:10.1046/j.1365-294x.1998.00289.x)
Talbot, S. L. & Shields, G. F. 1996 Phylogeography of brown
bears (Ursus arctos) of Alaska and paraphyly within the
Ursidae. Mol. Phylogenet. Evol. 5, 477–494. (doi:10.
1006/mpev.1996.0044)
Teacher, A. G. F., Garner, T. W. J. & Nichols, R. A. 2009
European phylogeography of the common frog (Rana temporaria):
routes of postglacial colonization into the British
Isles, and evidence for an Irish glacial refugium. Heredity
102, 490–496. (doi:10.1038/hdy.2008.133)
Thienemann, A. 1950 Verbreitungsgeschichte der Su¨sswassertierwelt
Europas. Die Binnengewa¨sser. Band XVIII. Stuttgart,
Germany: Schweizerbart’sche Verlagsbuchhandlung.
Tremblay, N. O. & Schoen, D. J. 1999 Molecular phylogeography
of Dryas integrifolia: glacial refugia and postglacial
recolonization. Mol. Ecol. 8, 1187–1198. (doi:10.1046/
j.1365-294x.1999.00680.x)
Trewick, S. A., Morgan-Richards, M., Russell, S. J.,
Henderson, S., Rumsey, F. J., Pı´nter, I., Barrett, J. A.,
Gibby, M. & Vogel, J. C. 2002 Polyploidy, phylogeography
and Pleistocene refugia of rockfern Asplenum ceterah:
evidence from chloroplast DNA. Mol. Ecol. 11, 2003–2012.
(doi:10.1046/j.1365-294X.2002.01583.x)
Tyler, T. 2002a Geographical distribution of allozyme variation
in relation to post-glacial history in Carex digitata,
a widespread European woodland sedge. J. Biogeogr. 29,
919–930. (doi:10.1046/j.1365-2699.2002.00698.x)
Tyler, T. 2002b Large-scale geographic patterns of genetic
variation in Melica nutans, a widespread Eurasian woodland
grass. Plant Syst. Evol. 236, 73–87. (doi:10.1007/
s00606-002-0235-8)
Tyrberg, T. 1991 Arctic, montane and steppe birds as glacial
relicts in West Palearctic. Orn. Verh. 25, 29–49.
Tzedakis, P. C. 1993 Long-term tree populations in
northwest Greece through multiple Quaternary
climatic cycles. Nature 364, 437–440. (doi:10.1038/
364437a0)
Tzedakis, P. C., Lawson, I. T., Frogley, M. R., Hewitt, G. M. &
Preece, R. C. 2002 Buffered tree population changes in a
Quaternary refugium: evolutionary implications. Science
297, 2044–2047. (doi:10.1126/science.1073083)
Volckaert, F. A. M., Ha¨nfling, B., Hellemann, B. &
Carvalho, G. R. 2002 Timing of the population dynamics
of bullhead Cottus cottus (Teleostei: Cottidae) during the
Pleistocene. J. Evol. Biol. 15, 930–944. (doi:10.1046/
j.1420-9101.2002.00469.x)
von Koenigswald, W. 1999 Hat der Mensch das Aussterben
der grossen pleistozanen Pflanzenfresser verursacht?
Kaupia 9, 193–201.
Whittington-Jones, G. M., Bernard, R. T. F. & Parker, D. M.
2008 Bushclumps as refugia for small mammals in
two Eastern Cape conservation areas. Afr. Zool. 43,
273–276. (doi:10.3377/1562-7020-43.2.273)
Willis, K. J. & van Andel, T. H. 2004 Trees or no trees? The
environments of central and eastern Europe during the
Last Glaciation. Quat. Sci. Rev. 23, 2369–2387.
(doi:10.1016/j.quascirev.2004.06.002)
Willis, K. J., Rudner, E. & Su¨megi, P. 2000 The full-glacial
forests of central and southeastern Europe. Quat. Res.
53, 203–213. (doi:10.1006/qres.1999.2119)
Wo´jcik, J. M., Ratkiewicz, M. & Searle, J. B. 2002
Evolution of the common shrew Sorex araneus: chromosomal
and molecular aspects. Acta Theriol. 47(Suppl. 1),
139–167.
Review. Refugia revisited J. R. Stewart et al. 671
Proc. R. Soc. B (2010)
on April 16, 2013rspb.royalsocietypublishing.orgDownloaded from