Refugia of marine ﬁsh in the northeast Atlantic during the ...A. J. Kettle et al.: Refugia of...

Clim Past 7 181ndash201 2011wwwclim-pastnet71812011doi105194cp-7-181-2011copy Author(s) 2011 CC Attribution 30 License

Climateof the Past

Refugia of marine fish in the northeast Atlantic during the lastglacial maximum concordant assessment from archaeozoologyand palaeotemperature reconstructions

A J Kettle1 A Morales-Muniz2 E Rosello-Izquierdo2 D Heinrich3 and L A Voslashllestad4

1Department of Earth Science SUNY-Oswego Oswego New York USA2Laboratorio de Arqueozoologia Universidad Autonoma de Madrid Madrid Spain3Boesselstrasse 9 24937 Flensburg Germany4Center for Ecological and Evolutionary Synthesis Department of Biology University of Oslo Oslo Norway

Received 7 June 2010 ndash Published in Clim Past Discuss 16 July 2010Revised 11 January 2011 ndash Accepted 28 January 2011 ndash Published 3 March 2011

Abstract Archaeozoological finds of the remains of ma-rine and amphihaline fish from the Last Glacial Maximum(LGM) ca 21 ka ago show evidence of very different speciesranges compared to the present We have shown how an eco-logical niche model (ENM) based on palaeoclimatic recon-structions of sea surface temperature and bathymetry can beused to effectively predict the spatial range of marine fishduring the LGM The results indicate that the ranges of ma-rine fish species now in northwestern Europe were displacedsignificantly southwards from the modern distribution chal-lenging an existing paradigm of marine glacial refugia Themodel presents strong evidence that there was an invasionof important fish through the Straits of Gibraltar in glacialtimes where they were exploited by Palaeolithic human pop-ulations around the western Mediterranean Sea The ENMresults are important for ongoing studies of molecular ecol-ogy that aim to assess marine glacial refugia from the ge-netic structure of living populations and they pose questionsabout the genetic identity of vanished marine populationsduring the LGM Economically the approach may be usedto understand how the ranges of exploited fish species maybe displaced with the future climate warming The researchpresents a challenge for future archaeozoological work to de-limit the glacial refugia and to verify palaeoclimatic recon-structions based on deep-sea core records

Correspondence toA J Kettle(anthonykettleoswegoedu)

1 Introduction

To predict how the geographic ranges of marine species innorthwest Europe will change with future climate warmingit is important to understand their environmental responsecharacteristics and how their spatial distributions changedduring the climate fluctuations of the Pleistocene glaciationsAlthough the spatial distribution of marine species may belargely determined by water temperatures (eg Kucera etal 2005 Lenoir et al 2010) the complicated interactionsof different marine species within ecosystems ndash with evi-dence of step-like regime shifts (Beaugrand and Reid 2003)ndash make predictions of future range modification difficult (Be-lyea 2007) The best way to predict the future response ofmarine ecosystems may be to understand changes under theextreme climate conditions of the past Much palaeoclimaticinformation exists for the cold conditions of the Last GlacialMaximum (LGM) ca 21 000 calender years before present(Mix et al 2001 Sarnthein et al 2003) as well at the mid-Holocene warm period 6000 years ago (Bradley 1999) butthe response of marine species across the full range of thetemperature changes is not well-known The LGM is partic-ularly important to address this issue because it representsa situation of large perturbation of temperatures from thepresent condition and therefore affords the clearest pictureof how marine species respond to environmental tempera-ture change Archaeozoology reveals important informationabout past climate change and corroborates palaeoclimaticinformation from other sources Palaeoclimatic evidencesuggests that Scandinavia was covered by glacial ice sheetsand in northern Europe permafrost conditions extended to theMassif Centrale in France with tundra conditions extending

Published by Copernicus Publications on behalf of the European Geosciences Union

182 A J Kettle et al Refugia of marine fish in the northeast Atlantic during the last glacial maximum

into Spain (eg Butzer 1964 Bradley 1999) In these cli-matic conditions northern terrestrial animal species rangedover western and southwestern Europe where they were uti-lized by human cultures and their remains are found in thepalaeolithic archaeological sites of the region (eg Butzer1964) Animal and plant species survived in glacial refu-gia in the Iberian peninsula Italy and the Balkans and it isfrom these centers that Europe was recolonized during theHolocene Genetic evidence has complemented archaeozoo-logical evidence in elaborating on the paths of colonizationfor terrestrial animals freshwater fish and plants (eg He-witt 2000 Stewart and Lister 2001 Provan and Bennett2008) For the marine fish species there is much less in-formation mostly because the archaeozoological record hasbeen lost as sea levels have risen by approximately 120 msince the LGM (eg Provan and Bennett 2008) How-ever genetic studies (eg Maggs et al 2008) and the sparsearchaeozoological evidence indicate startling shifts in thegeographic ranges of some species in response to climatechange Here we present an overview of how the rangesof four marine and amphihaline species changed during theLGM and apply a simple ecological niche model (ENM) toexplain how these altered ranges are consistent with sea sur-face temperature (SST) changes during the glacial period

Identifications of fish remains from archaeological siteshave revealed large changes in the ranges of some marineand amphihaline species during the LGM The migratory al-lis and twaite shad (Alosa alosaandA fallax respectively)and European eel (Anguilla anguilla) disappeared from ar-chaeological sites on the Atlantic drainage basins of Franceand Cantabria Spain during the LGM and probably survivedin refugia further to the south (Le Gall 2008) For the eelthis refugium was probably on the Atlantic coasts of Portu-gal and Morocco (eg Kettle et al 2008) The glacial refugeof the shad is unclear but remains have been discovered onthe upper Tagus River system at Aridos-1 from the Mindel-Riss interglacial approximately 300 ka BP (Morales 1980Le Gall 2000) and it is likely to have survived the LGM onthe Iberian Peninsula and northwest Africa From southernSpain the boreal gadid species haddock (Melanogrammusaeglefinus) and pollock (Pollachius pollachius) have beenreported by Rodrigo (1994) for the Cueva de Nerja nearMalaga dating to the LGM These results were significant be-cause these species are currently found in northwest Europeand the southernmost range for haddock and pollock is cur-rently the Bay of Biscay and northern Portugal respectively(Whitehead et al 1986) The most recent analyses of theLGM deposits from the Cueva de Nerja have revealed thatthe northern gadids (including saithePollachius virens codGadus morhua and lingMolva molva) make up more than30 of the identified fish remains and thus represented a sig-nificant presence among the exploited fish (Cortes-Sanchezet al 2008 Morales-Muniz and Rosello-Izquierdo 2008)The exploitation of gadids continued for an extended periodof time after the LGM until the early Holocene as sea levels

were returning to present levels and sea surface temperaturesapproached present values

Among the most spectacular shifts in species ranges havebeen reports of the remains of Atlantic salmon (Salmo salar)in archaeological sites from the Mediterranean drainagebasins of France Spain and Italy dating from the LGM up tothe early Holocene Additional small art objects unambigu-ously depicting salmon have been located at the Grand Grottede Bize on a Mediterranean drainage basin (Le Gall 1994a)These reports from late Palaeolithicndashearly Mesolithic sitesin France Spain and Italy (Juan-Muns i Plans 1985 Juan-Muns i Plans et al 1991) suggest the presence of residentsalmon populations in the western Mediterranean Sea Thecurrent southernmost range of Atlantic salmon in northwestEurope is northern Portugal which indicates that immigrantpopulations would have had to pass through the Straits ofGibraltar when temperatures in southern Spain during theLGM were similar to northern Europe at present Howeverthe archaeozoological evidence is contested Many of thesites in southern France are in proximity to Atlantic drainagebasins and it is possible that the Atlantic salmon remainsin the Mediterranean watersheds may have been transportedas part of a seasonal migration of fishermen (Le Gall 19831984 1992 1994a 1994b) This has even been invokedto explain Atlantic salmon finds in the more distant Men-ton caves on the French-Italian border (Riviere 1886 Clark1948) As well many of the early identifications of At-lantic salmon at the Caune de Belvis and the Grotte Jean-Pierre I by Desse and Granier (1976) and at Grimaldi (BarmeGrande II) have been subsequently questioned and revised totrout (Salmo trutta) by Le Gall (1994a) Desse-Berset (1994)and Desse and Desse-Berset (2002) respectively It is diffi-cult to distinguish Atlantic salmon and trout from their skele-tal elements and many of the early salmon identificationswere made mainly on the basis of the unusual size of verte-bral elements The revision of the initial identifications mayhave been made on reflection of the zoogeographic implica-tions of there being endemic Mediterranean Atlantic salmonpopulations during the LGM Up until now the prevailingparadigm has been that the SST in the western Mediterraneanduring the LGM was too high to meet the oxygen require-ments of the species (Le Gall 1983 1984 1994b) On theother hand recent unambiguous determinations of Atlanticsalmon in Solutrean levels in the Cueva de Nerja indicatethat ocean temperatures were cool enough during the LGMto permit the passage of the species through the Straits ofGibraltar (Morales-Muniz and Rosello-Izquierdo 2008)

Part of the reason why these archaeozoological finds areimportant is that there are few studies that have investigatedhow the geographic distribution of marine species changedduring the LGM and most of this is based on molecularmarkers Hence there are few a priori hypotheses of whatfish species should be found in archaeozoological collec-tions This situation contrasts strikingly with the state ofknowledge for terrestrial plants and animals and freshwater

Clim Past 7 181ndash201 2011 wwwclim-pastnet71812011

A J Kettle et al Refugia of marine fish in the northeast Atlantic during the last glacial maximum 183

fish where there is much zooarchaeological and genetic ev-idence for LGM refugia in southern Europe and recoloniza-tion pathways to northern Europe during the Holocene (egHewitt 2000 Stewart and Lister 2001 Provan and Bennett2008) Most of the important economic marine fish specieshave a high gene flow and weak genetic structure that makeit difficult to link local population structure with the exis-tence of glacial refugia Reviewing the genetic structure andLGM history of cod Pampoulie et al (2008) speculated thatglacial refuge populations survived on the Rockall Plateausouthwest Ireland and Irminger Sea However these pre-dictions are confusing because the LGM SST map of Pauland Schafer-Neth (2003) suggests that the ocean tempera-tures and sea ice conditions of these areas may have beentoo severe to serve as refugia The issue of assessing theMediterranean presence of North Atlantic marine species isimportant not only for determining the glacial refugia of eco-nomically important fish but also for establishing an accu-rate paradigm for palaeoenvironmental reconstructions us-ing sediment core records There is currently some uncer-tainty about the provenance of some North Atlantic plank-tonic species found in sediment cores of the western Mediter-ranean and this represents a puzzle for palaeoenvironmentalreconstructions (Rohling et al 1998)

ENMs attempt to predict the spatial distribution of aspecies based on databases of climatic (eg temperaturerainfall) and other information and the most robust exam-ples might predict the spatial distribution based on thresh-old envelopes of temperature for example above and belowwhich no viable populations are observed to occur ENMshave been successfully applied to explain the present-daydistributions of some terrestrial species and then carefullycalibrated to predict the past distributions during the LGMusing palaeoclimatic model data (Waltari et al 2007) How-ever for marine species these models have only recently beenapplied and for some species there are sometimes errorsin the predictions The extension of ENMs to explain ma-rine fish distributions during the LGM has recently been per-formed by Bigg et al (2008) who used two algorithms topredict the glacial distribution of cod Although one algo-rithm based on a maximum entropy method predicted suit-able LGM habitat along the Atlantic coast of northwest Eu-rope and also large areas of the Mediterranean and the BlackSea the actual occupation of these areas in southern Eu-rope was judged unlikely because of geographic disjunctionA second algorithm based on ecophysiological constraintspredicted only limited southward displacement during glacialtimes and this clearly shows important disagreements withrecent fossil evidence from the Cueva de Nerja in southernSpain (eg Cortes-Sanchez et al 2008 Morales-Muniz andRosello-Izquierdo 2008) Hence there is still a crucial gapin the understanding of the influence of oceanic conditionsthat governed the spatial distribution of the marine speciesduring the LGM The issue is important because future cli-mate warming scenarios predict that the spatial ranges of

important species are expected to shift northward (eg Perryet al 2005) and important amphihaline species like salmonare predicted to become extinct in the southern parts of theirrange (Lassalle and Rochard 2009)

In this contribution we consider the changes in spatialdistribution of four species during the LGM haddock pol-lock the eastern Atlantic shad species and Atlantic salmonAlthough most marine species must have experienced rangechanges during the LGM these four species have been se-lected mainly because of the long-standing conundrum sug-gested by the unusual locations of their subfossil evidenceThey are also economically important Of the total Eu-ropean capture production ofsim44times 105 sim30times 105 and12times 103 tons for pollock haddock and salmon respec-tively with a combined value of approximately 26 ofthe total US$62 billion for European fisheries export prod-ucts (FAO Fishstat Plus v 232httpwwwfaoorgfisherystatisticssoftwarefishstaten) In the Sect 2 (Methods) wepresent the current understanding of the environmental re-quirements of the fish species together with the ENM algo-rithms used to determine their present distribution In Sect 3(Results) we present an alternative ENM approach to esti-mate LGM distributions and use palaeoclimatic SST recon-struction from the LGM to estimate how the species rangeschanged In Sect 4 (Discussion) we synthesize the informa-tion indicating consistencies between our results and emerg-ing genetic information from other marine species We pointout potentially promising lines of research and outline howthe ENM approach may be used with climate model predic-tions of future SST changes to assess how fish populationswill be displaced

2 Methods an Ecological Niche Model (ENM) based ontemperature and bathymetry

Our predictions of the LGM distribution of marine speciesare based on a simple ENM that is constructed fromthe intersection of two environmental conditions (SST andbathymetry) Based on the approximate correspondence be-tween the environmental fields and expert assessments of thespecieslsquo ranges we define the approximate present-day en-velope of threshold environmental conditions that circum-scribes the species niche Then we infer the LGM distri-butions based on published reconstructions of the SST andbathymetry during glacial times Important advantages ofthis method of bioclimatic envelopes are (1) its transparencyin the interpretation of a small dataset (Phillips et al 2006Ready et al 2010) and (2) proven track record in predict-ing modern-day ranges of fish species from imperfect sur-vey data (Ready et al 2010) Although resource assess-ments (Kaschner et al 2006 Lenoir et al 2010 Readyet al 2010) and palaeoclimatic reconstructions (Kucera etal 2005) have followed a statistical approach where largedatabases have been available sparse archaeozoological or

wwwclim-pastnet71812011 Clim Past 7 181ndash201 2011


37

1

Figure 1 Spatial distribution of (a) haddock (Melanogrammus aeglefinnus) (b) pollock 2

(Pollachius pollachius) (c) allis and twaite shad (Alosa alosa and Alosa fallax) and (d) 3

Atlantic salmon (Salmo salar) Information is from the Fishbase Aquamap (colour-scale) and 4

atlas of Whitehead et al (1986 denoted by lsquoWhiteheadrsquo and shown by hatching) The 5

Fishbase Aquamap is a metric of relative environmental suitability that is based on the 6

application of environmental envelopes to explain fish survey data in terms of gridded fields 7

of bathymetry temperature salinity ice cover and primary production (Kaschner et al 2006 8

Ready et al 2010) The bathymetry is from ETOPO-5 with sea level and the single 9

indicated contour (200 m or 800 m) plotted on the map The two shad species are plotted 10

together because their archaeozoological remains are often difficult to distinguish 11

12

Fig 1 Spatial distribution of(a) haddock (Melanogrammus aeglefinnus) (b) pollock (Pollachius pollachius) (c) allis and twaite shad (Alosaalosaand Alosa fallax) and(d) Atlantic salmon (Salmo salar) Information is from the Fishbase Aquamap (colour-scale) and atlas ofWhitehead et al (1986 denoted by ldquoWhiteheadrdquo and shown by hatching) The Fishbase Aquamap is a metric of relative environmentalsuitability that is based on the application of environmental envelopes to explain fish survey data in terms of gridded fields of bathymetrytemperature salinity ice cover and primary production (Kaschner et al 2006 Ready et al 2010) The bathymetry is from ETOPO-5 withsea level and the single indicated contour (200 m or 800 m) plotted on the map The two shad species are plotted together because theirarchaeozoological remains are often difficult to distinguish

palaeontological datasets have value if they contain a strik-ing species (ldquoindicator speciesrdquo or ldquoguide fossilsrdquo Peacock1989) that must indicate certain threshold conditions Thishas been exploited in palaeoenvironmental reconstructionsfurther back in time where the data record may not be asrich as for the LGM (eg Huber et al 2000) Following theconvention used to assess modern fish resource distributionswith large databases (Kaschner et al 2006 Lenoir et al2010 Ready et al 2010) we have assumed that the salientniche features for the species in our study are captured witheach environmental parameter acting independently

The present day distribution of the four marine and am-phihaline species is shown in Fig 1 Information is availablefrom several sources but we present expert assessments fromtwo recognized authorities Fishbase (httpwwwfishbaseussearchphp) and Whitehead et al (1986) Other refer-ences are broadly consistent with the chosen expert sum-maries in terms of the geographic distribution but may

add extra regional information for example about where aspecies may be particularly common (ICES-Fishmap Had-dock 2009) Presenting only an envelope of presence-absence information with minimal information about abun-dance for some species the information in Whitehead etal (1986) is the simplest and most robust The Fish-base Aquamaps present a graded zonation of distributionthat is based on sea surface temperature sea surface salin-ity (SSS) productivity bathymetry and an annual ice cover(see Kaschner et al 2006 and Ready et al 2008) Althoughthe extra information implied by the abundance appears use-ful the algorithm is calibrated based on survey informationand it may not reproduce expert assessments For examplefor Atlantic salmon in Fig 1d there is a predicted distribu-tion in the Mediterranean and Black Seas based on habitatconditions but the present distributions of wild populationsare limited to the Atlantic coasts of northwest Europe as in-dicated by Whitehead et al (1986)



38

1

Figure 2 Climatological sea surface temperature of the northeast Atlantic in (a) winter 2

(January February March) and (b) summer (July August September) from WOA05 3

(Levitus 2006) 4

5

6

Fig 2 Climatological sea surface temperature of the northeast Atlantic in(a) winter (January February March) and(b) summer (JulyAugust September) from WOA05 (Levitus 2006)

The minimum and maximum summertime (July-August-September JAS) SST envelopes that define the southern andnorthern extent of each species are shown in Fig 1 TheSSTs are taken from a present-day climatological atlas at 1

resolution that is shown in Fig 2 (from the World Ocean At-las of Levitus 2006 abbreviated WOA05) Temperature isrecognized as an important determinant of the spatial dis-tribution of fish (Lenoir et al 2010) and this is groundedon theoretical considerations of oxygen limitation and thetemperature-dependence of the metabolic processes (Portner2001 Portner and Knust 2007) Although species showtemperature-dependence at every stage in their life cycleBigg et al (2008) consider temperatures in late spring andearly summer as the primary factor affecting the spawningrange of cod and Lenoir et al (2010) confirm the impor-tance of temperature as a defining criterion for the youngestfish stages For their ENMs Bigg et al (2008) use SSTas the metric of the spatial distribution to generate distri-bution maps This does not at first seem like an obviousmetric for bottom temperature conditions which are impor-tant for a demersal species like cod On the other handthe 10C annual average SST isotherm provides an impor-tant functional definition of the southern limit of the spa-tial distribution of cod on both sides of the Atlantic Ocean(Brander 1996) As well Lenoir et al (2010) point outthat there is a high correlation between surface temperaturesand temperatures at 100 m This supports the widespreadpractice in resource studies and palaeoclimatic reconstruc-tions of using SST for the upper ocean conditions to de-termine species distributions For our study we selectedthe summertime SST (JAS) as the primary criteria definingthe spatial distribution of the species instead of the win-tertime (December-January-February DJF) or annual aver-age SST In addition to thermal preferences the species alsohave a preferred depth habitat (Lenoir et al 2010) whichdefines the seaward extent of the range and this has been

taken from Fishbase and Whitehead et al (1986) Along withbathymetry limits (ETOPO-5httpwwwngdcnoaagovmggglobaletopo5HTML) the isotherm envelopes form thebasis of the present day distribution ranges shown in Fig 3These maps were generated by matching the summertimeSST with the species distributions shown in Fig 1 and a1C uncertainty in the definition of the thermal envelopesdoes not change the conclusions of this analysis

The chosen temperature envelopes are broadly consistentwith the temperature thresholds of the four species observedduring the field and laboratory studies For haddock theICES-Fishmap (Haddock 2009) gives a minimum temper-ature threshold of 65C and Brodziak (2005) reports thejuvenile and adult fish have been caught during fisheries sur-veys between temperatures of 2ndash16C Peck et al (2003)give the temperature of maximum growth at 12C whichis approximately at the center of the summer climatolog-ical envelope of 6ndash18C that has been used to define itsrange in Fig 1 For pollock also a demersal species theempirically-defined summer climatological SST envelope of10ndash18C for the northern and southern range of the species(Fig 1) is close to the physiological range of temperaturesof 9ndash18C outside of which severe growth decreases areobserved (Person-Le Ruyet et al 2006) These tempera-tures thresholds for haddock and pollock are broadly con-sistent with Lenoir et al (2010) using a statistical analysis ofa larger database For the allis shad which spawns in fresh-water in springtime the observed migration from the oceantakes place when river temperatures are between 133ndash23Cas measured by Acolas et al (2006) and this defines theproper freshwater environment for spawning and early rear-ing of young This observed temperature range is very closeto the empirical climatological average summer SSTs thathave been used to define the spatial distribution (13ndash22Cshown in Fig 1c) For the twaite shad the lower temperatureenvelope may be slightly reduced compared to allis shad to



39

1

Figure 3 Simplified species ranges using simplified criteria based on the summer SST and 2

bathymetry envelopes shown in each diagram The archaeozoological finds are given by blue 3

dots with the number key given in Table 1 For (a) and (b) the archaeological site is 1 Cueva 4

de Nerja Spain For (c) the archaeological sites are 1 Peacutegourieacute Atlantic France 2 Sous 5

Balme Mediterranean France 3 Aridos-1 Spain and 4 Lapa dos Coelhos Portugal For 6

(d) the archaeological sites are 1 Grotte Jean-Pierre I Mediterranean France 11 Baousseacute-7

Rousseacute Grottes de Menton Italy 12 Barme Grande Grimaldi caves Italy 13 Cueva de 8

Nerja Spain and a cluster of different sites along the Mediterranean coast of France and 9

Catalonia with labels 2ndash10 The bathymetry is from ETOPO-5 with sea level and the single 10

additional bathymetric contour indicated on the maps 11

Fig 3 Simplified species ranges using simplified criteria based on the summer SST and bathymetry envelopes shown in each diagram Thearchaeozoological finds are given by blue dots with the number key given in Table 1 For(a) and(b) the archaeological site is 1 Cueva deNerja Spain For(c) the archaeological sites are 1 Pegourie Atlantic France 2 Sous Balme Mediterranean France 3 Aridos-1 Spainand 4 Lapa dos Coelhos Portugal For(d) the archaeological sites are 1 Grotte Jean-Pierre I Mediterranean France 11 Baousse-RousseGrottes de Menton Italy 12 Barme Grande Grimaldi caves Italy 13 Cueva de Nerja Spain and a cluster of different sites along theMediterranean coast of France and Catalonia with labels 2ndash10 The bathymetry is from ETOPO-5 with sea level and the single additionalbathymetric contour indicated on the maps

explain its more northerly spatial distribution of occasionaloccurrence in Norway and Iceland (but not spawning at thesenorthern locations)

For Atlantic salmon the northern and southern bound-aries of the range are empirically determined as the 6Cand 19C summer SST isotherms (Fig 1d) There are dif-ferent possible reasons for this temperature envelope Thefish is highly susceptible at the post-smolt stage just after itmigrates from the rivers to the ocean in springndashearly sum-mer Although the mechanism is unclear stock size is neg-atively correlated with June SST and post-smolts are nega-tively affected by the early arrival of warm ocean tempera-tures (Friedland et al 2003) The empirical summer SSTenvelope is similar to the observed temperature of the down-stream migration in Norway in June 25ndash155C (Hvidstenet al 1998) Another hypothesized mechanism to explainthe spatial range of adult salmon relates to its high dissolved

oxygen requirements (gt7 mg Lminus1) whose saturation valuedepends mainly on the temperature but also the salinity ofseawater (Le Gall 1983 1984) Le Gall (1994b) stated thatthe southern distribution of the species ndash and particularly itsabsence in the Mediterranean ndash is governed by this oxygenrequirement and the theoretical basis of the temperature de-pendence of aerobic activity has been more recently clarifiedby Portner and Knust (2007)

Figure 3 also shows the locations of the unusual archae-ological features for the time interval from the LGM to theearly Holocene (Table 1) For haddock pollock and Atlanticsalmon the LGM distribution is far outside of the presentranges suggesting that there were extreme changes in thepast distributions during the glacial periods

The probable spatial distribution of the fish species dur-ing the LGM were independently assessed from LGMSST and bathymetry criteria Although CLIMAP Project



Table 1 Location of presenceabsence of the remains (or artwork) of haddock pollock shad and Atlantic salmon in pre-Holocene archaeo-logical sites outside their present geographic range The indices in the table identify the location of the archaeological sites in Figs 3 and 6Allis shad (Alosa alosa) and twaite shad (Alosa fallax) are placed together in the table because their archaeozoological remains are oftendifficult to distinguish For the time period of the archaeological site uncal refers to uncalibrated radiocarbon dates (C Bonsall reports anerror in Table 1 of Pickard and Bonsall 2004 which shows the Mesolithic remains of allis shad at Advik Varanger fjord Norway but is notfound in the primary source of Renouf 1989)

Species Archaeological site Time period Reference

Haddock 1 Cueva de Nerja LGMGravettian and Rodrigo (1994) Aura(Melanogrammus Spain Solutrean (c 24ndash Tortosa et al (2002)aeglefinus) 175 ky BP) Cortez-Sanchez et al

(2008) Morales-Munizand Rosello-Izquierdo(2008)

Pollock 1 Cueva de Nerja LGMGravettian and Rodrigo (1994) Aura(Pollachius Spain Solutrean (c 24ndash Tortosa et al (2002)pollachius) 175 ky BP) Cortez-Sanchez et al


Shad 1 Pegourie Lot Absence during Le Gall (1993 1994b(Alosa alosaand France (A alosain LGM small 1995 2000) MartinA fallax) Atlantic drainage) percentage at level 7 and Le Gall (2000)

(12 250plusmn 350 uncalBP) majority of fishremains at 11 000 BPin levels 4ndash6

2 Sous-Balme lrsquoAin Sauveterriens Le Gall (1994b 2000)France (A alosaor anciens 9000 BPA alosarhodensiasinMediterraneandrainage)

3 Aridos-1 Tagus Mindel-Riss Morales (1980)River Spain Interglacial Le Gall (2000)(Atlantic drainage) (sim300 k yr BP)

4 Lapa dos Coelhos Magdalenian Rosello and MoralesAlmonda tributary (2011)of Tagus RiverPortugal

Atlantic salmon 1 Grotte Jean-Pierre I Magdalenien Desse and Granier(Salmo salar) Saint-Thibaud- superieura (1976) evidence

de-Couz Savoie Mesolithique ancien disputed byFrance Desse-Berset (1994)

2 LrsquoAbeurador Younger Dryas and Le Gall (1983 1984France Preboreal 1994a)

3 Caune de Belvis Levels 2 3 4 Desse and GranierAude France (sim12 270plusmn 280 uncal (1976) Juan-Muns i

BP) Plans et al (1991)evidence disputed byLe Gall (1994a)



Table 1Continued


4 La Grande Grotte Upper Magdalenien Le Gall (1994a)de Bize Aude (12 550plusmn 210 uncal Le Gall (2001)France (artwork) BP)

5 Canecaude 1 Middle Magdalenian Le Gall (1994a)Aude France (14 230plusmn 160 uncal

BP)

6 La Grotte Gazel Middle Costamagno andAude France Azilien Laroulandie (2004

Magdalenian- citing Desse-Bersetunpublished)

7 La Grotte de Upper Magdalenian Le Gall (1994a)lrsquoOeil Aude (13 800ndash1 300 uncal Costamagno andFrance BP) Laroulandie (2004)

8 LrsquoArbreda Gravettian Solutrean Juan-Muns i PlansSerinya Girona (1985) Juan-MunsSpain (1987) Juan-Muns i

Plans et al (1991)

9 Davant Pau Solutrean Juan-Muns i PlansSerinya Girona (1985) Juan-Muns iSpain Plans et al (1991)

10 Reclau Viver Solutrean Juan-Muns i PlansSerinya Girona (1985) Juan-Muns iSpain Plans et al (1991)

11 Baousse-Rousse Palaeolithic Rivieres (1886)Grottes de MentonItaly

12 Barme Grande Juan-Muns i Plans et alGrimaldi cave (1991) citing ClarkItaly (1948)

13 Cueva de Nerja Solutrean Morales-Muniz andSpain (18 420plusmn 530 Rosello-Izquierdo

17 940plusmn 200 (2008) (largeSalmosp15 990plusmn 260 identified asSalmo salar)uncal BP)

Members (1976) produced the first global view of LGM cli-mate conditions recent efforts have led to an updated com-munity consensus of LGM conditions These have beendownloaded from the Glacial Ocean Atlas (GOAhttpwwwglacialoceanatlasorg) the gridded surface data fieldsof Paul and Schafer-Neth (2003) and the synthesis dataset ofMARGO Project Members (2009) The background of thesedatasets highlights the unresolved challenges of this fieldwith the fields of Paul and Schafer-Neth (2003) representinga compilation of objective measurements and expert opinion

to achieve the best guess gridded field that is necessary for amodelling study By contrast the dataset of MARGO ProjectMembers (2009) employs a more conservative approach todevise statistics for 5 times 5 latitude-longitude boxes wheredata occur and this means that there are data gaps and reso-lution problems using the sparse record of deep sea cores

The spatial interpolation of point determinations ofpalaeo-SSTs at core locations is still an unresolved problem(Schafer-Neth et al 2005) It is important for ENM stud-ies such as ours which seek to trace a pathway of favorable



40

1

2

Figure 4 Climatological SST from the LGM as determined by Paul and Schaumlfer-Neth (2003) 3

for (a) winter and (b) summer and by MARGO Project Members (2009) for (c) winter and (d) 4

summer The maps of Paul and Schaumlfer-Neth (2003) are from the Global Ocean Atlas (GOA 5

httpwwwglacialoceanatlasorg) and are overplotted with a sea ice template if any of the 6

three months represented in the field have ice cover The MARGO-interpolated fields (also 7

from GOA) were calculated using a simple interpolation method based on an inverse 8

distance-squared weighting factor using the 5degtimes5deg summary data of MARGO Project 9

Members (2009) with an LGM land template is derived from Peltier (1994) and ETOPO-5 10

(httpwwwngdcnoaagovmggglobaletopo5HTML) No sea ice template was provided 11

with this this dataset 12

13

14

Fig 4 Climatological SST from the LGM as determined by Paul and Schafer-Neth (2003) for(a) winter and(b) summer and by MARGOProject Members (2009) for(c) winter and(d) summer The maps of Paul and Schafer-Neth (2003) are from the Global Ocean Atlas (GOAhttpwwwglacialoceanatlasorg) and are overplotted with a sea ice template if any of the three months represented in the field have icecover The MARGO-interpolated fields (also from GOA) were calculated using a simple interpolation method based on an summary dataof MARGO Project Members (2009) with an LGM land template is derived from Peltier (1994) and ETOPO-5 (httpwwwngdcnoaagovmggglobaletopo5HTML) No sea ice template was provided with this this dataset

environmental conditions between ocean basins The GOAwebsite actually has two sets of fields by Paul and Schafer-Neth (2003) one based on expert determinations of the SSTisolines and a second based on a kriging interpolation of SSTdata directly from deep-sea cores The second set of fieldsbased on primary data has been used here For the dataof MARGO Project Members (2009) we have developed aninterpolation scheme on a 1

times 1 grid (with an LGM landtemplate derived from Peltier 1994) using a weighting fac-tor based on inverse square distance that follows broadly onthe WOA05 approach We consider data within a thresh-old distance of 1800 km to ensure that information is propa-gated from at least one point of the original dataset and thisinterpolated product is hereafter referred to as ldquoMARGO-interpolationrdquo Although this simple interpolation schemedoes not take account of the preferred zonal projection ofinformation in the North Atlantic Ocean (ie as is implicitlyassumed by Paul and Schafer-Neth 2003 using expert deter-minations of isolines) the MARGO palaeotemperature infor-mation is dense enough that realistic SST fields are producedTaken together the fields of Paul and Schafer-Neth (2003)and MARGO-interpolation represent a range of palaeo-SST

conditions to test the predictions of our ENM and this fol-lows the recommended practice of Nogues-Bravo (2009) toassess the uncertainty of reconstructed palaeoclimate wherethis is available

3 Results fish distributions during the Last GlacialMaximum (LGM)

The two palaeo-SST datasets for the northeast Atlantic fromPaul and Schafer-Neth (2003) and MARGO-interpolation areshown in Fig 4 for winter (January-February-March) andsummer (July-August-September) The difference betweenthe LGM reconstructions and the present day climatology(Fig 5) illustrates the temperature changes in the north-east Atlantic during the LGM Most of the mid-latitude ar-eas exhibit some cooling during the LGM with the mostsignificant decreases around the UK in summer (exceeding12C for the Paul and Schafer-Neth 2003 dataset) In theMediterranean Sea temperature decreases are less severewith sim8ndash10C summertime decreases in the west andsim4ndash6C summertime decreases in the east Summer tempera-ture conditions at the Strait of Gibraltar during the LGM



41

1

2

3

4

5

Figure 5 Difference in climatological SST LGM-present for Paul and Schaumlfer-Neth (2003) 6

(a) winter and (b) summer and for MARGO Project Members (c) winter and (d) summer For 7

each palaeo-environmental reconstruction the present day SST climatology is from the World 8

Ocean Atlas of Levitus (2006) 9

10

Fig 5 Difference in climatological SST LGM-present for Paul and Schafer-Neth (2003)(a) winter and(b) summer and for MARGO ProjectMembers(c) winter and(d) summer For each palaeo-environmental reconstruction the present day SST climatology is from the WorldOcean Atlas of Levitus (2006)

would have been similar to the south coast of the UK atpresent Although the spatial features of the the two palaeo-SST reconstructions are broadly similar the newer informa-tion of MARGO Project Members (2009) suggests tempera-tures of the Nordic Seas were not as cold as depicted by Pauland Schafer-Neth (2003) while temperatures in the easternMediterranean were several degrees cooler than previouslybelieved especially in summer (Hayes et al 2005)

The LGM distributions of fish species (Figs 6 and 7)have been assessed using ENMs based on these two palaeo-SST fields and the palaeotopography fields of Peltier (1994)All species show a significant southward displacement awayfrom their present ranges but the cooler palaeo-SST predic-tions of Paul and Schafer-Neth (2003) have especially im-portant implications for the location of the LGM northernboundaries Both ENM results are consistent with the avail-able archaeozoological results The Cueva de Nerja in theStrait of Gibraltar was well within the estimated LGM rangeof haddock and pollock and the ENMs predict that thesespecies may have populated the western Mediterranean Seaas far east as the Adriatic Sea or even the Aegean Sea ac-cording to the MARGO-interpolation fields The predictedfundamental ecological niche in the Black Sea and Red Seawas probably not realized for the species of this investigationbecause of geographic disjunction

Atlantic salmon could likewise swim through the Strait ofGibraltar to form populations in the western MediterraneanSea consistent with its archaeozoological presence in theMediterranean watersheds of Spain France and Italy duringthe LGM This species is cold-adapted and is found at presentin northern Europe as far south as northern Portugal (al-though there are historical reports of the species as far southas the Guadalquivir River up to 20th century B Elvira per-sonl communication 2008) Both ENMs for the LGM pre-dict that it maintained a continuous presence on the Gironde-Dordogne river system through the LGM where its unin-terrupted archaeological presence across the LGM has beendocumented

The allis shad requires warmer SST conditions and bothENMs indicates that it was displaced from the Atlantic wa-tersheds of France during the LGM and this is consistentwith its apparent absence from archaeological sites in theDordogne region (Le Gall 2008) The recent identificationof the remains of (allis) shad in Magdalenian levels of theLapa dos Coelhos archaeological site on the Tagus water-shed of central Portugal gives support to the existence of aglacial refuge in the southern Iberian Peninsula during theLGM (Rosello and Morales 2010) The model prediction ofa theoretical niche extension into the eastern Mediterraneanis not much further than its current documented eastern limit



42

1

Figure 6 Predicted species ranges from the LGM reconstruction of Paul and Schaumlfer-Neth 2

(2003) for (a) haddock (b) pollock (c) allis shad and (d) Atlantic salmon The ranges are 3

based on the same criteria as for the present distribution in Fig 3 The archaeological site key 4

is the same as Fig 3 The LGM ranges within the frame of these maps are 14ndash33 of the 5

present day ranges and the LGM distributions are almost completely shifted away from 6

present-day ranges shown in Fig 3 The LGM bathymetry is derived from the 1deg sea level 7

data of Peltier (1994) at 21 ka before present interpolated to the 5rsquo fields of ETOPO-5 LGM 8

sea level and the single additional bathymetric contour are indicated on the map 9

10

Fig 6 Predicted species ranges from the LGM reconstruction of Paul and Schafer-Neth (2003) for(a) haddock(b) pollock (c) allis shadand(d) Atlantic salmon The ranges are based on the same criteria as for the present distribution in Fig 3 The archaeological site key isthe same as Fig 3 The LGM ranges within the frame of these maps are 14ndash33 of the present day ranges and the LGM distributions arealmost completely shifted away from present-day ranges shown in Fig 3 The LGM bathymetry is derived from the 1 sea level data ofPeltier (1994) at 21 ka before present interpolated to the 5prime fields of ETOPO-5 LGM sea level and the single additional bathymetric contourare indicated on the map

at Sicily but it is also unclear how much of this fundamen-tal niche would have been occupied with potential compe-tition with other shad species already in the Mediterranean(Nogues-Bravo 2009)

A second important message in Figs 6 and 7 is the ex-tent to which the species ranges were reduced during thesouthern displacement and confined to the continental sloperegions For the Paul and Schafer-Neth (2003) SST fieldsin Fig 6 the LGM ranges were decreased to approximately19 19 33 and 14 of the present range for haddock pol-lock allis shad and salmon respectively For the MARGO-interpolation fields in Fig 7 the corresponding LGM rangewas about 53 31 47 and 39 of the modern day range Theresults for the gadids haddock and pollock are broadly con-sistent with the reduction of the range of cod in the NorthAtlantic reported by Bigg et al (2008)sim20 of presentday extent Although the largest contiguous area of speciesis presently the North Sea the largest contiguous area for

all four species during the LGM may have been the shal-low shelf area between Tunisia and Sicily Significantly theMARGO-interpolation fields suggest that the cold-adaptedspecies haddock and salmon may have maintained a pres-ence in the Nordic seas although accuracy of the palaeo-SSTreconstructions in this region are still uncertain (MARGOProject Members 2009)

4 Discussion

The ENM results presented here give a very different pre-diction of LGM glacial refuge areas for some of the eco-nomically important fish species of northwest Europe com-pared with the few previous published studies The mainfinding is that the glacial refuge of four highly vagile fishspecies are significantly shifted away from their present bio-geographical range mostly in northwest Europe There is



43

1

Figure 7 Predicted species ranges from the LGM reconstruction using the interpolated 5degtimes5deg 2

fields of MARGO Project Members (2009) for (a) haddock (b) pollock (c) allis shad and (d) 3

Atlantic salmon The ranges are based on the same criteria as for the present distribution in 4

Fig 3 The archaeological site key is the same as Fig 3 The LGM ranges within the frame 5

of these maps are 31ndash53 of the present day ranges and the LGM distributions are displaced 6

southward from present-day ranges shown in Fig 3 The LGM bathymetry is derived from 7

the 1deg sea level data of Peltier (1994) at 21 ka before present interpolated to the 5rsquo fields of 8

ETOPO-5 LGM sea level and the single additional bathymetric contour are indicated on the 9

map 10

11

Fig 7 Predicted species ranges from the LGM reconstruction using the interpolated 5times 5 fields of MARGO Project Members (2009) for

(a) haddock(b) pollock (c) allis shad and(d) Atlantic salmon The ranges are based on the same criteria as for the present distributionin Fig 3 The archaeological site key is the same as Fig 3 The LGM ranges within the frame of these maps are 31ndash53 of the presentday ranges and the LGM distributions are displaced southward from present-day ranges shown in Fig 3 The LGM bathymetry is derivedfrom the 1 sea level data of Peltier (1994) at 21 ka before present interpolated to the 5prime fields of ETOPO-5 LGM sea level and the singleadditional bathymetric contour are indicated on the map

strong evidence from two palaeo-SST reconstructions and ar-chaeozoological identifications that cold-adapted gadids andAtlantic salmon invaded the western Mediterranean throughthe Strait of Gibraltar The results are startling because thepredicted LGM spatial ranges are different from previous as-sumptions and modelling studies For example Pampoulieet al (2008) speculate that cod survived the LGM on theRockall Plateau and the Irminger Sea in a reduced subareaof the present range (Fig 8) Also the ecophysiologicalmodel Bigg et al (2008) predicted that cod survived theLGM in continental shelf areas of northwest Europe and thatpopulations were displaced only slightly further south com-pared to the present locations The predictions of our ENMare supported by subfossil archaeological evidence of thefish species in archaeological sites far to the south of theirpresently-recognized southern boundaries

It is difficult to independently verify the predictions of ourENM with genetic information Molecular markers are not

well-suited to elucidating the glacial refugia of highly vagilefish species (Gysels et al 2004b) like the examples of ourstudy Highly mobile fish can follow optimal sea tempera-tures during changing climate conditions so that the locationof their glacial refugia is unclear This is particularly truefor haddock pollock and allis shad where the few molecu-lar marker studies have indicated weak population structureamong stocks which was effectively erased during the LGM(see Jamieson and Birley 1989 and Reiss et al 2009 forhaddock Charrier et al 2006 for pollock and Alexandinoand Boisneau 2000 for shad) For some species the pop-ulation age estimated from molecular markers may be inex-plicably older than the known length of habitat occupationsince the LGM (Francisco et al 2009) Fish of low dis-persal ability are valuable tracers of glacial refugia and thesand goby (Pomatoschistus minutus) is one such model ma-rine species whose present population structure is a legacy ofrange constrictions during the glacial period (Gysels et al



44

1

Figure 8 Marine glacial refugia of plants (green font) invertebrates (blue font) and teleosts 2

(red font) modified from reviews in Goacutemez et al (2007) Hoarau et al (2007) Maggs et al 3

(2008) Larmuseau et al (2009) and Olsen et al (2010) together with their cited primary 4

reports Thin black lines show the present and LGM coastline and the 200 m LGM isobaths 5

The red boxes present the possible marine refugia reviewed by Maggs et al (2008) These are 6

supported by a reanalysis of genetics results except for the Lofoten Islands (ie a recognized 7

terrestrial but not marine glacial refugium) and the Mediterranean (ie a controversial 8

marine refuge area for northeast Atlantic species (Larmuseau et al 2009) without supporting 9

species in the review of Maggs et al (2008)) The blue boxes and species marked with 10

question marks present other speculated marine refugia without firm support from 11

archaeozoological or genetic evidence (eg the glacial refugia for Gadus morhua as given in 12

Pampoulie et al (2008)) Marine glacial refugia have identified for plants (Ascophyllum 13

nodosum (Olsen et al 2010) Fucus serratus (Coyer et al 2003 Hoarau et al 2007) 14

Fig 8 Marine glacial refugia of plants (green font) invertebrates (blue font) and teleosts (red font) modified from reviews in Gomez etal (2007) Hoarau et al (2007) Maggs et al (2008) Larmuseau et al (2009) and Olsen et al (2010) together with their cited primaryreports Thin black lines show the present and LGM coastline and the 200 m LGM isobaths The red boxes present the possible marinerefugia reviewed by Maggs et al (2008) These are supported by a reanalysis of genetics results except for the Lofoten Islands (ie arecognized terrestrial but not marine glacial refugium) and the Mediterranean (ie a controversial marine refuge area for northeast Atlanticspecies Larmuseau et al 2009 without supporting species in the review of Maggs et al 2008) The blue boxes and species marked withquestion marks present other speculated marine refugia without firm support from archaeozoological or genetic evidence (eg the glacialrefugia forGadus morhuaas given in Pampoulie et al 2008) Marine glacial refugia have identified for plants (Ascophyllum nodosumOlsen et al 2010Fucus serratus Coyer et al 2003 Hoarau et al 2007Palmaria palmata Provan et al 2005Zostera marina Olsenet al 2004) invertebrates (Carcinus maenas Roman and Palumbi 2004Celleporella hyalina Gomez et al 2007Idotea balthica Wares2001 Pectinaria koreni Jolly et al 2005 2006Macoma balthica Luttikhuizen et al 2003 Nikula et al 2007Pollicipes pollicipesCampo et al 2010Semibalanus balanoides Wares and Cunningham 2001) and teleosts (Pomatoschistus microps Gysels et al 2004bPomatoschistus minutus Larmuseau et al 2009Raja clavata Chevolot et al 2006Salmo salar Consuegra et al 2002 Langefors 2005Scomber scombrus Nesboslash et al 2000) Species marked with an asterisklowast have been critically reviewed by Maggs et al (2008) Some of theunusual (and sometimes diverse) genetic signatures in the North Sea and Baltic are speculated to arise from recolonization from east of theFennoscandian ice sheet and possibly as far as the Pacific Ocean which was ice-free at high latitudes during the LGM (Olsen et al 2004)

2004a Larmuseau et al 2009) The population has a con-tiguous distribution in northern Europe but is fragmented inthe Mediterranean consistent with a species that expandedsouth during glaciations and retreated northwards during theinterglacials as climate conditions became warmer during theHolocene

For some species molecular marker studies reveal pat-terns of recolonization from glacial refugia Many of thehypothesized refugia in Fig 8 are located in different partsof southwestern Europe and northwestern Africa south ofthe ice sheets and sea ice cover (Fig 9) The Bay of Bis-cay Iberian Peninsula and Macaronesia (Madeiras Azoresand Canaries) have all been invoked as glacial refugia of



various marine species with a present distribution in north-west Europe This is consistent with the predictions of ourSST-based ENMs However smaller periglacial refugia ndash theHurd Deep near the mouth of the glacial Channel River andsouthwest Ireland ndash have also been inferred from genetic ev-idence for other species These regions may have been sub-ject to ice cover for part of the year (Fig 9) but this does notpreclude them as glacial refugia for marine species (Gomezet al 2007) Other refugia have been hypothesized furthernorth in northwest Scotland northern North Sea Faeroe Is-lands Iceland and Norway for species adapted to colder con-ditions and longer sea ice periods (Fig 9) The identifica-tion of some of these glacial refugia pose important ques-tions about the locations of the European ice sheets (Sejrupet al 2005) which has been identified as an important openquestion in a recent palaeoclimatological review (Mix et al2001)

Atlantic salmon is an important exception whose geneticstructure reveals a complex and interesting recolonizationhistory which introduces another dimension to our resultsOur study suggests that BrittanyHurd Deep was the north-ernmost range for this species and that the species was sim-ply displaced southward during the LGM The unusual ge-netic structure in the Baltic Sea may have resulted frompopulation isolation during the complex marine transgres-sion history resulting from the interplay of sea level changesand the isostatic rebound in the early Quaternary (Lepiksaar2001 p 40) However it has also been used to infer recol-onization from possible glacial refugia in the southern NorthSea or beyond the eastern edge of the Fennoscandian icesheet (eg Consuegra et al 2002 Langefors 2005 Kinget al 2007 but see also Makhrov et al 2005) A westwardrecolonization pathway into the Baltic Sea from the WhiteSea after the LGM was previously perceived as unlikely Onthe other hand the genetic signatures of the other marinespecies indicate a possible link between extant populations inthe Baltic Sea (and eastern Atlantic) with those further eastin the Arctic and Pacific Oceans (Luttikhuizen et al 2003Addison and Hart 2005 Nikula et al 2007)

The LGM ENM predictions of Bigg et al (2008) for codare valuable because the approach is among the first appli-cations of the technique for the marine domain and it isbeing used for other North Atlantic species (Provan et al2009) Bigg et al (2008) used two different ENMs for codthat were calibrated to present conditions and it is impor-tant to investigate the possible reasons for the partial discrep-ancy of their models with our results The maximum entropyENM used by Bigg et al (2008) was based on bathymetrySST and SSS This produced an LGM distribution for codthat stretched from northwest Europe into large areas of theMediterranean consistent with the report of archaeozoolog-ical remains from the Cueva de Nerja The ecophysiologicalENM incorporated a spawning temperature threshold of 0ndash9C between February and June This is different from thetemperature range that we have used for haddock (ie an

46

1

Figure 9 Months of ice cover from LGM (Paul and Schaumlfer-Neth 2003) with glacier template 2

at 21 ka before present indicated by hatching (Peltier 1994) The LGM bathymetry (sea level 3

and 200 m depth) is derived from the 1deg sea level data of Peltier (1994) at 21 ka before 4

present interpolated to the 5rsquo fields of ETOPO-5 The present-day coastline is indicated on 5

the map The small boxes give a more recent assessment (De Vernal et al 2005) of the 6

average number of months of sea ice (top left triangle gives the average number of months 7

Fig 9 Months of ice cover from LGM (Paul and Schafer-Neth2003) with glacier template at 21 ka before present indicated byhatching (Peltier 1994) The LGM bathymetry (sea level and 200 mdepth) is derived from the 1 sea level data of Peltier (1994) at21 ka before present interpolated to the 5prime fields of ETOPO-5 Thepresent-day coastline is indicated on the map The small boxes givea more recent assessment (De Vernal et al 2005) of the averagenumber of months of sea ice (top left triangle gives the average num-ber of months the lower right triangle gives the average-plus-one-standard-deviation) During the LGM the Nordic seas may havebeen mostly ice-free during the summer months

analogous demersal gadid species with a similar range) inFig 1a whose spatial range we have defined to lie betweenthe summertime isotherms of 6ndash18C (or the 3ndash12C win-tertime isotherms not shown) The ENM temperature enve-lope used by Bigg et al (2008) was slightly lower than ourchoice It may have resulted in northern offsets of both thenorthern and southern present-day species boundaries whenevaluated against field surveys as noted in the original pub-lication When projected back to the LGM these thermalenvelopes have implications for the predicted glacial refugiaThe direct comparison of the different temperature criteriais difficult between the ecophysiological model of Bigg etal (2008) and the temperature envelopes of our study partlybecause strong seasonal variation of the North Atlantic near-surface ocean temperature An important message in Bigg etal (2008) is that the LGM distributions of the marine speciescan be predicted with just a few parameters (bathymetry andSST) and this is an important feature of robust simplicitythat is maintained in our study

The assumption of ldquoniche stability through timerdquo forms thefoundation of all ENM approaches (Nogues-Bravo 2009)but it has not been addressed by Bigg et al (2008) or in ourstudy and it is difficult to prove without a large database of



independent information to delimit past distributions Forour work it translates into the assumption that the fish re-sponded to ocean temperature and depth during the LGM inthe same way as at present On the other hand some evi-dence indicates that the fish may have changed their behav-ior in the past for unknown reasons possibly associated withhuman exploitation or climate change For example Beeren-hout (1994) pointed out the remains of full-grown haddockin early Neolithic sites in the Netherlands where they werecaught in brackish water estuarine environments The re-port is startling because the species is currently subject tocommercial exploitation at the northern edge of the NorthSea at approximately 200 m depth (ICES-Fishmap Haddock2009) The Neolithic human populations of the Netherlandsdid not have the technological means to secure haddock fromits 20th century habitat and there is an important implica-tion that the fish range may have changed to a very differ-ent stable state in response to human exploitation Fromthe late-Palaeolithic period a similar message is presentedby the haddock finds from the Epipalaeolithic levels of theCueva de Nerja and these fish were also probably obtainedas part of an inshore shallow-water fishery Changes in thetrophic structure of marine ecosystems have been inferredfrom changes in species assemblages and sizes based on evi-dence from Stone Age archaeological sites and there is animplication that it may have been due to human exploita-tion (Desse and Desse-Berset 1993 Morales and Rosello2004) This may be linked with alterations in preferred habi-tats of fish populations In assessing the possibility of aMesolithic deep-sea fishery of northern Europe Pickard andBonsall (2004) have had to address the provenance of certainspecies in archaeological sites that presently occur offshorein deep water and specifically identify bluefin tuna (Thun-nus thynnus ie an offshore species with observed spawningmigrations inshore) golden redfish (Sebastes marinus) hake(Merluccius merluccius) halibut (Hippoglossus hippoglos-sus) tusk (Brosme brosme) and wolffish (Anarhichas lupus)Although the authors assume that ldquothe habitat preferences ofprehistoric fish populations were broadly similar to those ofmodern speciesrdquo the presence of significant numbers of re-mains of so many deep water species in Mesolithic archae-ological sites forces consideration that the fish may havechanged their habitat preferences This is not mere specu-lation since strong fishing pressure has been repeatedly doc-umented to provoke changes in the behaviour of fish popu-lations (Perry et al 2010 Planque et al 2010) Alterna-tively such shifts may have been provoked by the systematicfishing of inshore populations that may have led to local ex-tinctions of those living in the shallowest and most accessi-ble waters Indeed the aforementioned studies (Perry et al2010 Planque et al 2010) reveal that vital demographic pa-rameters of fish populations such as a shorter life cycle canbe a response to a systematic fishing of the largest specimensmaking certain stocks more vulnerable to extinction Thiscould have been most easily achieved on the most accessible

biotopes (eg estuaries shallow waters etc) where evenwithout a particularly high fishing pressure removal of suchfishes could bring about the cascade of changes leading tothe extinction of a not-too-large population

The results of our study represent an evolving picture ofthe marine species distributions that will be revised with fu-ture work on palaeoclimatic environmental reconstructionsand modelling activities Palaeoclimatic reconstructions ofSST have evolved since the initial CLIMAP Project Mem-bers (1976) assessments but there are still ocean areas wherethe LGM assessments of SST and SSS are unclear and wheredifferent proxy methods may disagree (MARGO ProjectMembers 2009) On the other hand from the archaeozoo-logical perspective there appears to be a convergence of evi-dence for the climatic conditions in the Mediterranean basinduring the LGM The LGM SST reconstruction in Figs 4and 5 indicates that whereas western Mediterranean SSTconditions decreased drastically and were low enough to sup-port populations of northern gadids and Atlantic salmon an-nual average SSTs in the eastern Mediterranean were onlyabout 15ndash3C cooler than present (Hayes et al 2005) Ma-rine ecosystems in the eastern Mediterranean did not expe-rience the same intense cooling conditions as in the westernbasin and SSTs may have been similar to modern conditionsThe new LGM temperature reconstructions force a departurefrom an earlier paradigm that SSTs were too warm to permitthe migration of these species through the Strait of Gibral-tar (eg Le Gall 1994b) Rather than a barrier at the Straitof Gibraltar our ENM indicates that there may have beenan important thermal boundary in the middle of the Mediter-ranean at Sicily and this is consistent with the genetic stud-ies of other marine species that show an important divisionat Tunisia-Sicily saddle (Domingues et al 2005) Likewisethe predicted LGM range of the species of our study indi-cate that there may have been a geographic disjunction be-tween the western Mediterranean and the Adriatic Sea Thishighlights that our ENM can only predict the fundamentalecological niche and the realized niche in the Adriatic Seaand eastern Mediterranean remains less certain On the otherhand the significance of the barrier represented by the ItalianPeninsula on the free migration between populations in thewestern Mediterranean and the Adriatic Sea may help to ex-plain the distinct genetic signatures of certain marine speciesin these two basins (Gysels et al 2004a Debes et al 2008Maggio et al 2009)

Future experimental work may aim to verify the predic-tions of the ENM that have been presented here For exam-ple genetic studies might aim to determine the structure ofthe vanished populations in the Mediterranean (Nielsen andHansen 2008) For the archaeological LGM Atlantic salmonpopulation in northern Spain Consuegra et al (2002) founda dominant haplotype that has almost vanished from mod-ern populations in the region However the question of thegenetic structure of the hypothesized LGM salmon popula-tions in the Mediterranean is open The amount of exchange



through the Straits of Gibraltar during the LGM is unknownand it is not clear if the Mediterranean population at the ex-treme of the speciesrsquo range had the opportunity to developits own genetic signature

Our ENMs make predictions about the extent of the ex-pansion of haddock pollock allis shad and Atlantic salmoninto the central Mediterranean that can only be unequivo-cally resolved with further archaeozoological studies TheENMs effectively explain observations of LGM fish speciesdistributions at the Strait of Gibraltar and in southern FranceHowever there is some ambiguity about the ENM predic-tions at the extremes of the ranges in the eastern Mediter-ranean along west Africa and in the Nordic Seas and thisis due partly to the uncertainty in the SST reconstructions inthese areas Where the palaeoclimatic information in deepsea cores is equivocal archaeozoological information mayhelp to reduce the large uncertainty in LGM SST reconstruc-tions especially in the Nordic Seas whose summertime LGMSST is between the minimum temperature envelopes delim-iting the northern range of the different species consideredin our study For example the large differences in the ENMrange predictions for haddock in the Nordic Seas during theLGM are due to the different SST reconstructions and a sin-gle archaeozoological find of haddock in the contested re-gion would help resolve the differences between the maps ofPaul and Schafer-Neth (2003) and MARGO Project Mem-bers (2009) Although LGM archaeological sites with the re-mains of marine fish are uncommon an alternative approachis the quantification of fish otoliths from shallow sedimentcores on the continental shelf The concept has been exploredby Elder et al (1996) who used the otoliths of a small fishspeciesCeratoscopelus maderensisoff the eastern seaboardof the United States to demonstrate significant changes inits range between glacial and postglacial times Cod (ie alarger fish with a lower population density) otoliths were alsoidentified in their bottom samples and highlight possibilityof using this species to discriminate between different SSTreconstructions There is also potential evidence from themollusc remains in shallow marine deposits and archaeolog-ical sites Malatesta and Zarlenga (1986) cite many examplesof indicator molluscs (ldquoNorthern Guestsrdquo) as evidence of re-peated species invasions (and retreats) of the MediterraneanSea from northwest Europe during the coldest intervals ofthe Pleistocene glaciations For example the range of theAtlantic quahog (Arctica islandica) extended into the west-ern Mediterranean Sea during the LGM but the species be-came extinct in its southern European domain approximately9800 years BP and currently does not survive south of Brit-tany (Froget et al 1972 Dahlgren et al 2000) Along withthe invasion of Atlantic fish through the Strait of Gibraltarduring the LGM the presence of these molluscs may force are-interpretation of LGM planktonic ecosystem assemblagesfrom sediment cores in the western Mediterranean (Rohlinget al 1998) Although ENMs and genetic evidence pro-vide indications of past speciesrsquo ranges physical evidence

provided by fossil or subfossil material provides the most un-ambiguous identification of actual glacial refugia

Ultimately the importance of understanding the past dis-tributions of economic species during the LGM is to as-sess their thermal niche and predict how their ranges mightchange with future climate warming Analyzing fish surveytime series from the North Sea Perry et al (2005) calculatedthat the ranges of demersal species have shifted north approx-imately 170 km between 1962ndash2001 Other studies presentevidence of rapid range shifts on the order of several thou-sand kilometers over a few years for such pelagic species assardines and anchovies due to their nonterritorial behaviourand higher dispersal capabilities (McFarlane et al 2000Beare et al 2004) Predictions of future temperature in-creases have been used to predict the extinction of migratoryspecies from the southern parts of their ranges (eg Lassalleand Rochard 2009) but the calibration of these models de-pends on how species responded to climatic temperature per-turbations in the past The results presented here capture thesouthern displacement of certain marine species during theLGM The warm conditions during the mid-Holocene warmperiod offer a potential proxy for future climate warming Atthis time warm water species like the European sea bass (Di-centrarchus labrax) were displaced from their historically-recognized range south of the British Isles to become eco-nomically important for the Mesolithic cultures of Denmark(eg Enghoff et al 2007) and the German Baltic Sea coast(Heinrich 2001 Schmolcke et al 2006) Likewise the At-lantic quahog whose present northern distribution stops atthe White Sea ranged to Spitsbergen western Greenland andfar across the northern coast of Russia during the warmer cli-mate conditions of the mid-Holocene (Dahlgren et al 2000)The archaeozoological evidence thus provides critical data topredict how the spatial ranges of marine species may changein the coming decades and centuries

AcknowledgementsWe thank K Uehara for supplying highresolution palaeotopography fields We appreciate the helpfulcomments of M Kucera A Paul and one anonymous reviewerthat have improved this work

Edited by Andre Paul

References

Acolas M L Veron V Jourdan H Begout M L SabatieM R and Bagliniere J L Upstream migration and repro-ductive patterns of a population of allis shad in a small river(LrsquoAulne Brittany France) ICES J Mar Sci 63 476ndash484doi101016jicesjms200407023 2006

Addison J A and Hart M W Colonization dispersal andhybridization influence phylogeography of North Atlantic seaurchins (Strongylocentrotus droebachiensis) Evolution 59532ndash543doi10155404-238 2005



Alexandrino P and Boisneau P 8 Diversite genetique in LesAloses (Alosa alosaet Alosa fallaxspp) Ecobiologie et Vari-abilite des Populations edited by Bagliniere J-L and Elie PInstitut National de la Recherche Agronomique (INRA) Paris179ndash198 2000

Aura Tortosa J E Jorda Pardo J F Perez Ripoll M Ro-drigo Garcıa M J Badal Garcıa E and Guillem CalatayudP The far south the Pleistocene-Holocene transition inNerja Cave (Andalucıa Spain) Quatern Int 93ndash94 19ndash30doi101016S1040-6182(02)00004-6 2002

Beare D J Burns F Grieg A Jones E G Peach K Kien-zle M McKenzie E and Reid D G Long-term increasesin prevalence of North Sea fishes having southern biogeographicaffinities Mar Ecol-Prog Ser 284 269ndash278 2004

Beaugrand G and Reid P C Long-term changes in phytoplank-ton zooplankton and salmon related to climate Global ChangeBiol 9 801ndash817doi101046j1365-2486200300632x 2003

Beerenhout B What conclusions can be drawn from mature had-dock bones in a Neolithic site in the Netherlands Offa 51 341ndash347 1994

Belyea L R Revealing the Emperorrsquos new clothes niche-based palaeoenvironmental reconstruction in the lightof recent ecological theory Holocene 17 683ndash688doi1011770959683607079002 2007

Bigg G R Cunningham C W Ottersen G Pogson G HWadley M R and Williamson P Ice-age survival of Atlanticcod agreement between palaeoecology and genetics P RoySoc B-Biol Sci 275 163ndash172doi101098rspb200711532008

Bradley R S Paleoclimatology Reconstructing Climates of theQuaternary Academic Press San Diego 1999

Brander K Effects of climate change on cod (Gadus morhua)stocks in Global Warming Implications for freshwater and ma-rine fish edited by Wood C M and McDonald D G Societyfor Experimental Biology Seminar Series 61 Cambridge Univer-sity Press 255ndash278 1996

Brodziak J K T HaddockMelanogrammus aeglefinnus LifeHistory and Habitat Characteristics NOAA Technical Memoran-dum NMFS-NE-196 US Department of Commerce December2005

Butzer K W Environment and Archaeology Aldine PublishingCompany Chicago 1964

Campo D Molares J Garcia L Fernandez-Rueda PGarcia-Gonzalez C and Garcia-Vazquez E Phylogeogra-phy of the European stalked barnacle (Pollicipes pollicipes)identification of glacial refugia Mol Biol 157 147ndash156doi101007s00227-009-1305-z 2010

Charrier G Durand J-D Quiniou L and Laroche J An inves-tigation of the population structure of pollack (Pollachius pol-lachius) based on microsatellite markers ICES J Mar Sci 631705ndash1709doi101016jicesjms200607006 2006

Chevolot M Hoarau G Rijnsdorp A D and Stam WT Phylogeography and population structure of thornbackrays (Raja clavataL Rajidae) Mol Ecol 15 3693ndash3705doi101111j1365-294X200603043x 2006

Clark J G D The development of fishing in prehistoric EuropeAntiq J 28 45ndash85 1948

CLIMAP Project Members The surface of the ice-age EarthScience 191 1131ndash1137doi101126science191423211311976

Consuegra S Garcıa de Leaniz C Serdio A Gonzalez MoralesM Straus L G Knox D and Verspoor E Mitochon-drial DNA variation in Pleistocene and modern Atlantic salmonfrom the Iberian glacial refugium Mol Ecol 11 2037ndash2048doi101046j1365-294X200201592x 2002

Cortes-Sanchez M Morales-Muniz A Simon-Vallejo M DBergada-Zapata M M Delgado-Huertas A Lopez-Garcıa PLopez-Saez J A Lozano-Francisco M C Riquelme-CantalJ A Rosello-Izquierdo E Sanchez-Marco A and Vera-Pelaez J L Palaeoenvironmental and cultural dynamics ofthe coast of Malaga (Andalusia Spain) during the Upper Pleis-tocene and early Holocene Quaternary Sci Rev 27 2176ndash2193doi101016jquascirev200803010 2008

Costamagno S and Laroulandie V Lrsquoexploitation des pe-tits vertebres dans les Pyrenees francaises du Paleolithique auMesolithique un inventaire taphonomique et archeozoologiquein Petits Animaux et Societes Humaines du Complement Al-imentaire aux Ressources Utilitaires edited by Brugal J-Pand Desse J XXIVe rencontres internationales drsquoarcheologie etdrsquohistoire drsquoAntibesEditions APDCA Antibes 403ndash416 2004

Coyer J A Peters A F Stam W T and Olsen J L Post-ice agerecolonization and differentiation ofFucus serratusL (Phaeo-phyceae Fucuceae) populations in Northern Europe Mol Ecol12 1817ndash1829doi101046j1365-294X200301850x 2003

Dahlgren T G Weinberg J R and Halanych K M Phylo-geography of the ocean quahog (Arctica islandica) influences ofpaleoclimate on genetic diversity and species range Mar Biol137 487ndash495doi101007s002270000342 2000

De Vernal A Eynaud F Henry M Hillaire-Marcel C Lon-deix L Mangin S Matteissen J Marret F Radi T Ro-chon A Solignac S and Turon J-L Reconstruction of sea-surface conditions at middle to high latitudes of the NorthernHemisphere during the Last Glacial Maximum (LGM) based ondinoflagellate cyst assemblages Quaternary Sci Rev 24 897ndash924 2005

Debes P V Zachos F E and Hanel R Mitochondrialphylogeography of the European sprat (Sprattus sprattusLClupeidae) reveals isolated climatically vulnerable populationsin the Mediterranean Sea and range expansion in the north-east Atlantic Mol Ecol 17 3873ndash3888doi101111j1365-294X200803872x 2008

Desse G and Granier J Les poissons in La Prehistoire francaiseI 1 CNRS Paris 437ndash441 1976

Desse J and Desse-Berset N Peche et surpeche en Mediterraneele temoignage des os in Exploitation des animaux sauvagesa travers le temps actes des XIIIe Rencontres internationalsdrsquoarcheologie et drsquohistoire drsquoAntibes Octobre 1992 AntibesEditions APDCA 327-339 1993

Desse J and Desse-Berset N Le cortege de Neptune les pois-sons de la Mediterranee durant lrsquoHolocene in Mouvements oudeplacements de populations animales en Mediterranee au coursde lrsquoHolocene edited by Gardeisen A Seminaire France29 Septembre 2000 BAR International Series 1017 83ndash962002



Desse-Berset N Les poissons in Les grottes Jean-Pierre I et IIa Saint-Thibaud-de-Couz (Savoie) edited by Bintz P GalliaPrehistoire 36 218ndash224 1994

Domingues V S Bucciarelli G Almada V C and BernardiG Historical colonization and demography of the Mediter-ranean damselfishChromis chromis Mol Ecol 14 4051ndash4063doi101111j1365-294X200502723x 2005

Elder K L Jones G A and Bolz G Distribution of otolithsin surficial sediments of the US Atlantic continental shelf andslope and potential for reconstructing Holocene climate Paleo-ceanography 11 359ndash367doi10102996PA00042 1996

Enghoff I B MacKenzie B R and Nielsen E E TheDanish fish fauna during the warm Atlantic period (ca 7000ndash3900 BC) Forerunner of future changes Fish Res 87 167ndash180doi101016jfishres200703004 2007

Francisco S M Castilho R Soares M Congiu L Brito AVieira M N and Almada V C Phylogeography and demo-graphic history ofAtherina presbyter(Pisces Atherinidae) in theNorth-eastern Atlantic based on mitochondrial DNA Mar Biol156 1421ndash1432doi101007s00227-009-1182-5 2009

Friedland K D Reddin D G and Castonguay M Oceanicthermal conditions in the post-smolt nursery of North Atlanticsalmon ICES J Mar Sci 60 343ndash355doi101016S1054-3139(03)00022-5 2003

Froget C Thommeret J and Thommeret Y Mollusques septen-trionaux en Mediterranee occidentale datation par le14CPalaeogeogr Palaeocl 12 285ndash293 1972

Gomez A Hughes R N Wright P J Carvalho G R andLunt D H Mitochondrial DNA phylogeography and matingcompatibility reveal marked genetic structuring and speciationin the NE Atlantic bryozoansCelleporella hyaline Mol Ecol16 2173ndash2188doi101111j1365-294X200703308x 2007

Gysels E S Hellemans B Patarnello T and Volckaert F A MCurrent and historic gene flow of the sand gobyPomastoschistusminutuson the European Continental Shelf and in the Mediter-ranean Sea Biol J Linn Soc 83 561ndash576doi101111j1095-8312200400411x2004a

Gysels E S Hellemans B Pampoulie C and Volckaert FA M Phylogeography of the common gobyPomatoschistusmicrops with particular emphasis on the colonization of theMediterranean and the North Sea Mol Ecol 13 403ndash417doi101046j1365-294X200302087x 2004b

Hayes A Kucera M Kallel N Sbaffi L and Rohling EJGlacial Mediterranean sea surface temperatures based on plank-tonic formaminiferal assemblages Quaternary Sci Rev 24999ndash1016doi101016jquascirev200402018 2005

Heinrich D Bemerkungen zu den Tierknochenfunden von densubmarinen Siedlungsplatzen der Erteboslashlle-Kultur bei Tim-mendorfPoel und bei Neustadt (Marienbad) ndash ein VorberichtBeitrage zur Archaozoologie und Prahistorischen Anthropologie3 39ndash43 2001

Hewitt G The genetic legacy of the Quaternary ice ages Nature405 907ndash913doi10103835016000 2000

Hoarau G Coyer JA Veldsink JH Stam WT and OlsenJL Glacial refugia and recolonization pathways in thebrown seaweedFucus serratus Mol Ecol 16 3606-3616doi101111j1365-294X200703408x 2007

Huber B T MacLeod K G and Wing S L Warm Climates inEarth History Cambridge University Press Cambridge 2000

Hvidsten N A Heggberget T G and Jensen A J Sea watertemperatures at Atlantic salmon smolt enterance Nord J Fresh-water Res 74 79ndash86 1998

ICES-FishMap Haddock (WWW document)httpwwwicesdkmarineworldfishmapicespdfhaddockpdf(last access1 March 2011) 2009

Jamieson A and Birley A J The distribution of transferrin alle-les in haddock stocks J Cons Int Explor Mer 45 248ndash262doi101093icesjms453248 1989

Jolly M T Jollivet D Gentil F Thiebaut E and Viard FSharp genetic break between Atlantic and English Channel popu-lations of the polychaetePectinaria koreni along the North coastof France Heredity 94 23ndash32doi101038sjhdy68005432005

Jolly M T Viard F Gentil F Thiebaut E and Jollivet DComparative phylogeography of two coastal polychaete tube-worms in the North East Atlantic supports shared history andvicariant events Mol Ecol 15 1841ndash1855doi101111j1365-294X200602910x 2006

Juan-Muns i Plans N La ictiofauna dels jaciments arqueologicsCatalans Cypsela Girona V 21ndash33 1985

Juan-Muns N La ictiofauna de la cova de lrsquoArbreda (SerinyaGirona) Cypsela (Girona) VI 97ndash100 1987

Juan-Muns i Plans N Rodrigo Garcıa M J and Rodrıguez San-tana C G La ictiofauna de los yacimientos arqueologicosSus posilidades en la reconstruccion paleoecologica y de inter-pretacion paleoeconomica Arqueologıa Madrid 83ndash99 1991

Kaschner K Watson R Trites A W and Pauly D Mappingworld-wide distributions of marine mammal species using a rel-ative environmental suitability (RES) model Mar Ecol-ProgSer 316 285ndash310doi103354meps316285 2006

Kettle A J Heinrich D Barrett J H Benecke N and LockerA Past distributions of the European freshwater eel from ar-chaeologicaldoi101016jquascirev200803005 2008

King T L Verspoor E Spidle A P Gross R Phillips R BKoljonen M-L Sanchez J A and Morrison C L 5 Biodi-versity and population structure in The Atlantic salmon Genet-ics conser J L Blackwell Publishing Oxford 117ndash166 2007

Kucera M Weinelt M Kiefer T Pflaumann U Hayes AWeinelt M Chen M-T Mix A C Barrows T T Cor-tijo E Duprat J Juggins S and Waelbroeck C Recontruc-tion of sea-surface temperatures from assemblages of planktonicforaminifera multi-technique approach based on geographicallyconstrained calibration data sets and its application to glacial At-lantic and Pacific Oceans Quaternary Sci Rev 24 951ndash998doi101016jquascirev200407014 2005

Langefors A H Adaptive and neutral genetic variation and colo-nization history of Atlantic salmonSalmo salar Environ BiolFish 74 297ndash308doi101007s10641-005-0501-z 2005

Larmuseau M H D Van Houdt J K J Guelinckx J HellemansB and Volckaert F A M Distributional and demographic con-sequences of Pleistocene climate fluctuations for a marine dem-ersal fish in the north-eastern Atlantic J Biogeogr 36 1138ndash1151doi101111j1365-2699200802072x 2009

Lassalle G and Rochard E Impact of twenty-first century cli-mate change on diadromous fish spread over Europe NorthAfrica and the Middle East Glob Change Biol 15 1072ndash1089doi101111j1365-2486200801794x 2009



Le Gall O Presence de vestiges de saumon Atlantique dans unsite Epipaleolithique et Mesolithique du bassin versant Mediter-ranean Bulletin du Societe drsquoAnthropologie de Sud-OuestXVIII 53ndash55 1983

Le Gall O LrsquoExploitation de lrsquoichthyofaune par les paleolithiquesquelques exemples in N Desse-Berset (Ed) 2nd Fish Os-teoarchaeology Meeting Table Rond Sophia Antipolis ndash Val-bonne 14ndash16 Octobre 1983 CNRS Centres de RecherchesArcheologiques Notes et Monographies Techniques No 16 89ndash112 1984

Le Gall O Poissons et peches au Paleolithique (Quelques donneesde lrsquoEurope occidentale) LrsquoAnthopologie Paris 96 121ndash1341992

Le Gall O Evolution des peches de lrsquoEpipaleolithique auNeolithique Ancien in Prehistoire Anthropologie Mediter-raneenes LAPMO Universite de Provence CNRS 2 135ndash1421993

Le Gall O Elements de reflexion sur la peche dans le bassinmediterraneen nord-occidental pendant le developpement desfacies leptolithiques XXIVe Congres Prehistorique de France ndashCarcassonne 26ndash30 Septembre 1994 ndash Les facies leptolithiquesdu nord-ouest mediterraneen milieux naturels et culturels 251ndash265 1994a

Le Gall O Quelques remarques sur lrsquoadaptationa court eta longtermes chez les poissons drsquoeau douce du sud de la France inFish Exploitation in the Past edited by Van Neer W Proceed-ings of the 7th meetings of the ICAZ Fish Remains WorkingGroup Annales du Musee Royal de lrsquoAfrique Centrale SciencesZoologiques no 274 Turvuren 91ndash98 1994b

Le Gall O Etudes des Poissons in La Grotte de PegourieCaniac-du-Causse (Lot) edited by Seronie-Vivien M-RPrehistoire Quercinoise Supplement no 2 149ndash154 1995

Le Gall O 6 Origine et histoire des aloses in Les Aloses (Alosaalosaet Alosa fallaxspp) edited by Bagliniere J L and ElieP Ecobiologie et variabilite des populations INRA Paris 127ndash136 2000

Le Gall O Les representations de poissons dans lrsquoart mobiliermagdalenien Une expression de lrsquoimportance culturelle de lapeche Bulletin Prehistoire du Sud-Ouest no 8 55ndash69 2001

Le Gall O Les poissons des eaux douces pleistocenes sont-ils desindicateurs paleoclimatiques Une approcheelargiea lrsquoEuropein Archeologie du Poisson 30 Ans drsquoArcheo-Ichtyologie auCNRS edited by Bearez P Grouard S and Clavel B Hom-mage aux travaux de Jean Desse et Nathalie Desse-Berset XIV-th ICAZ Fish remains working group meetingEditions APDCAAntibes 311ndash326 2008

Lenoir S Beaugrand G and LecuyerE Modelled spa-tial distribution of marine fish and projected modifications inthe North Atlantic Ocean Global Change Biol 17 115ndash129doi101111j1365-2486201002229x 2011

Lepiksaar J Die spat- und postglaziale Faunengeschichte derSuszligwasserfische Schwedens Oetker-Voges Verlag Kiel 2001

Levitus S World Ocean Atlas 2005 NOAA Atlas NES-DIS US Government Printing Office Washington DChttpwwwnodcnoaagovOC5WOA05prwoa05html(last access1 March 2011) 2006

Luttikhuizen P C Drent J and Baker A J Disjunct dis-tribution of highly diverged mitochondrial lineage clade andpopulation subdivision in a marine bivalve with pelagic larval

dispersal Mol Ecol 12 2215ndash2229doi101046j1365-294X200301872x 2003

Maggio T Brutto S L Garoia F Tinti F and Arculeo MMicrosatellite analysis of red mulletMullus barbatus(Perci-formes Mullidae) reveals the isolation of the Adriatic Basin inthe Mediterranean Sea ICES J Mar Sci 66 1883ndash1891 2009

Maggs C A Castilho R Foltz D Henzler C Jolly M TKelly J Olsen J Perez K E Stam W Vainola R Viard Fand Wares J Evaluating signatures of glacial refugia for NorthAtlantic benthic marine taxa Ecology 89 Supplement S108ndashS122doi10189008-02571 2008

Makhrov A A Verspoor E Artamonova V S and OrsquoSullivanM Atlantic salmon colonization of the Russian Arctic coastpioneers from North America J Fish Biol 67 (Supplement A)68ndash79doi101111j0022-1112200500840x 2005

Malatesta A and Zarlenga F Northern guests in the PleistoceneMediterranean Sea Geologica Romana 25 91ndash154 1986

MARGO Project Members Constraints on the magnitude andpatterns of ocean cooling at the Last Glacial Maximum NatGeosci 2 127ndash132doi101038ngeo411 2009

Martin H and Le Gall O La faune un indicateur des comporte-ments humains La chasse de cerf au Post-Paleolithique deuxexemples Anthropologie et Prehistoire 111 364ndash369 2000

McFarlane G A King J R and Beamish R J Have there beenrecent changes in climate Ask the fish Prog Oceanogr 47147ndash169doi101016S0079-6611(00)00034-3 2000

Mix A C Bard E and Schneider R Environmental processesof the ice age land oceans glaciers (EPILOG) Quaternary SciRev 20 627ndash657 2001

Morales A Los peces fosiles del yacimiento achelense de Aridos-1 (Arganda Madrid) in Ocupaciones achelenses en el Valle delJarama edited by Santonja M and Lopez N Madrid 93ndash1041980

Morales A and Rosello E Fishing down the food web in Iberianprehistory A new look at the fishes from Cueva de Nerja(Malaga Spain) in Petits Animaux et Societes Humaines DuComplement Alimentaire aux Ressources Utilitaires edited byBrugal J-P and Desse J XXIVe rencontres internationalsdrsquoarcheologie et drsquohistoire drsquoAntibesEditions APDCA Antibes111ndash123 2004

Morales-Muniz A and Rosello-Izquierdo E Twenty thousandyears of fishing in the Strait in Human Impacts on Ancient Ma-rine Ecosystems A Global Perspective edited by Rick T Cand Erlandson J M University of California Press Berkeley243ndash277 2008

Nesboslash C L Rueness E K Iversen S A Skagen D Wand Jakobsen K S Phylogeography and population his-tory of Atlantic mackerel (Scomber scombrusL) a genealog-ical approach reveals genetic structuring among the easternAtlantic stocks P Roy Soc B-Biol Sci 267 281ndash292doi101098rspb20000998 2000

Nielsen E E and Hansen M M Waking the dead the value ofpopulation genetic analyses of historical samples Fish Fish 9450ndash461doi101111j1467-2979200800304x 2008

Nikula R Strelkov P and Vainola R Diversity and trans-Arcticinvasion history of mitochondrial lineages in the North AtlanticMacoma balthicacomplex (bivalvia tellinidae) Evolution 61928ndash941doi101111j1558-5646200700066x 2007



Nogues-Bravo D Predicting the past distribution of speciesclimatic niches Global Ecol Biogeogr 18 521ndash531doi101111j1466-8238200900476x 2009

Olsen J L Stam W T Coyer J A Reusch T B H Billing-ham M Bostrom C Calvert E Christie H Granger SLa Lumiere R Milchakova N Oudot-Le Secq M-P Pro-caccini G Sanjabi B Serrao E Veldsink J WiddicombeS and Wyllie-Echverria S North Atlantic phylogeographyand large-scale population differentiation of the seagrassZosteramarina L Mol Ecol 13 1923ndash1941doi101111j1365-294X200402205x 2004

Olsen J L Zechman F W Hoarau G Coyer J A Stam W TValero M andAberg P The phylogeographic architecture ofthe fucoid seaweedAscophyllum nodosum an intertidal ldquomarinetreerdquo and survivor of more than one glacial-interglacial cycle JBiogeogr 37 842ndash856doi101111j1365-2699200902262x2010

Pampoulie C Stefansson MO Jorundsdottir T D Danilow-icz B S and Danıelsdottir A K Recolonization history andlarge-scale dispersal in the open sea the case study of the NorthAtlantic codGadus morhuaL Biol J Linn Soc 94 315ndash329doi101111j1095-8312200800995x 2008

Paul A and Schafer-Neth C Modeling the water masses of theAtlantic Ocean at the Last Glacial Maximum Paleoceanography18 1058doi1010292002PA000783 2003

Peacock J D Marine molluscs and late Quaternary environmen-tal studies with particular reference to the late-glacial period innorthwest Europe a review Quaternary Sci Rev 8 179ndash192doi1010160277-3791(89)90006-1 1989

Peck M A Buckley L J Caldarone E M and Bengtson D AEffects of food consumption and temperature on growth rate andbiochemical-based indicators of growth in early juvenile Atlanticcod Gadus morhuaand haddockMelanogrammus aeglefinusMar Ecol-Prog Ser 251 233ndash243doi103354meps2512332003

Peltier W R Ice-age paleotopography Science 265 195ndash201doi101126science2655169195 1994

Perry A L Low P J Ellis J R and Reynolds J D Climatechange and distribution shifts in marine fishes Science 3081912ndash1915doi101126science1111322 2005

Perry R I Cury P Brander K Jennings S Mollmann C andPlanque B Sensitivity of marine systems to climate and fishingConcepts issues and management responses J Mar Syst 79427ndash435doi101016jjmarsys200812017 2010

Person-Le Ruyet J Buchet V Vincent B Le Delliou H andQuemener L Effects of temperature on the growth of pollack(Pollachius pollachius) juveniles Aquaculture 251 340ndash345doi101016jaquaculture200506029 2006

Phillips S J Anderson R P and Schapire R E Maximum en-tropy modeling of species geographic distributions Ecol Mod-ell 190 231ndash259doi101016jecolmodel200503026 2006

Pickard C and Bonsall C Deep-sea fishing in the EuropeanMesolithic fact for fantasy Eur J Archaeol 7 273ndash290doi1011771461957104056504 2004

Planque B Fromentin J-M Cury P Drinkwater K F JenningsS Perry R I and Kifani S How does fishing alter marinepopulations and ecosystems sensitivity to climate J Mar Syst79 408ndash417doi101016jjmarsys200812018 2010

Portner H O Climate change and temperature-dependentbiogeography oxygen limitation of thermal toler-ance in animals Naturwissenschaften 88 137ndash146doi101007s001140100216 2001

Portner H O and Knust R Climate change affects marine fishesthrough the oxygen limitation of thermal tolerance Science 31595ndash97doi101126science1135471 2007

Provan J and Bennett K D Phylogeographic insights intocryptic glacial refugia Trends Ecol Evol 23 564ndash571doi101016jtree200806010 2008

Provan J Wattier R A and Maggs C A Phylogeographic anal-ysis of the red seaweedPalmaria palmatareveals a Pleistocenemarine glacial refugium in the English Channel Mol Ecol 14793ndash803doi101111j1365-294X200502447x 2005

Provan J Beatty G E Keating S L Maggs C Aand Savidge G High dispersal potential has maintainedlong-term population stability in the North Atlantic copepodCalanus finmarchicus P Roy Soc B-Biol Soc 276 301ndash307doi101098rspb20081062 2009

Ready J Kaschner K South A B Eastwood P D Rees TRius J Agbayani E Kullander S and Froese R Predictingthe distributions of marine organisms at the global scale EcolModell 221 467ndash478doi101016jecolmodel2009100252010

Reiss H Hoarau G Dickey-Collas M and Wolff W J Geneticpopulation structure of marine fish mismatch between biologicaland fisheries management units Fish and Fisheries 10 361ndash395doi101111j1467-2979200800324x 2009

Renouf M A P Prehistoric Hunter-Fishers of VarangerfjordNortheastern Norway Brit Archaeol Rep In 478 Oxford1989

Riviere E Faune des oiseaux des reptiles et des poissons trouvesdans les cavernes des Baousse-Rousse (Italie) dites Grottes deMenton Materiaux pour lrsquoHistoire Primitive et Naturelle delrsquoHomme 20 525ndash535 1886

Rodrigo M J Remains ofMelanogrammus aeglefinus(Linnaeus1758) in the Pleistocene ndash Holocene passage in the cave of NerjaMalagaSpain Offa 51 348ndash351 1994

Rohling E J Hayes A De Rijk S Kroon D Zachari-asse W J and Eisma D Abrupt cold spells in thenorthwest Mediterranean Paleoceanography 13 316ndash322doi10102998PA00671 1998

Roman J and Palumbi S R A global invader at home populationstructure of the green crabCarcinus maenus in Europe MolEcol 13 2891ndash2898doi101111j1365-294X200402255x2004

Rosello E and Morales A Lapa dos Coelhos informe sobre losrestos de peces in Homenaje a Francisco Giles Pacheco editedby Mata E Publicaciones Junta de Andalucia Sevilla 261ndash276 2010

Sarnthein M Gersonde R Niebler S Pflaumann U Spiel-hagen R Thiede J Wefer G and Weinelt M Overviewof Glacial Atlantic Ocean Mapping (GLAMAP 2000) Paleo-ceanography 18 1030doi1010292002PA000769 2003

Schafer-Neth C Paul A and Mulitza S Perspectives on map-ping the MARGO reconstructions by variogram analysiskrigingand objective analysis Quaternary Sci Rev 24 1083ndash1093doi101016jquascirev200406017 2005



Schmolcke U Endtmann E Kloos S Meyer M MichaelisD Rickert B-H and Rossler D Changes of sea level land-scape and culture a review of the south-western Baltic area be-tween 8800 and 4000 BC Palaeogeogr Palaeocl 240 423ndash438doi101016jpalaeo200602009 2006

Sejrup H P Hjelstuen B O Dahlgren K I T HaflidasonH Kuijpers A Nygard A Praeg D Stoker M S andVorren T O Pleistocene glacial history of the NW Euro-pean continental margin Mar Petrol Geol 22 1111ndash1129doi101016jmarpetgeo200409007 2005

Stewart J R and Lister A M Cryptic northern refugia and theorigins of the modern biota Trends Ecol Evol 16 608ndash613doi101016S0169-5347(01)02338-2 2001

Waltari E Hijmans R J Peterson A T Nyari A S PerkinsS L and Guralnick R P Locating Pleistocene refugia com-peting phylogeographic and ecological niche model predictionsPLoS ONE 2 e563doi101371journalpone0000563 2007

Wares J P Intraspecific variation and geographic iso-lation in Idotea balthica (Isopoda valvifera) JCrustacean Biol 21 1007ndash1013doi1016510278-0372(2001)021[1007IVAGII]20CO2 2001

Wares J P and Cunningham C W Phylogeogra-phy and historical ecology of the North Atlantic in-tertidal Evolution 55 2455ndash2469 doi1015540014-3820(2001)055[2455PAHEOT]20CO2 2001

Whitehead P J P Bauchot M L Hureau J C Nielsen Jand Tortonese E Fishes of the North-eastern Atlantic and theMediterranean UNESCO Paris 1986



into Spain (eg Butzer 1964 Bradley 1999) In these cli-matic conditions northern terrestrial animal species rangedover western and southwestern Europe where they were uti-lized by human cultures and their remains are found in thepalaeolithic archaeological sites of the region (eg Butzer1964) Animal and plant species survived in glacial refu-gia in the Iberian peninsula Italy and the Balkans and it isfrom these centers that Europe was recolonized during theHolocene Genetic evidence has complemented archaeozoo-logical evidence in elaborating on the paths of colonizationfor terrestrial animals freshwater fish and plants (eg He-witt 2000 Stewart and Lister 2001 Provan and Bennett2008) For the marine fish species there is much less in-formation mostly because the archaeozoological record hasbeen lost as sea levels have risen by approximately 120 msince the LGM (eg Provan and Bennett 2008) How-ever genetic studies (eg Maggs et al 2008) and the sparsearchaeozoological evidence indicate startling shifts in thegeographic ranges of some species in response to climatechange Here we present an overview of how the rangesof four marine and amphihaline species changed during theLGM and apply a simple ecological niche model (ENM) toexplain how these altered ranges are consistent with sea sur-face temperature (SST) changes during the glacial period

Identifications of fish remains from archaeological siteshave revealed large changes in the ranges of some marineand amphihaline species during the LGM The migratory al-lis and twaite shad (Alosa alosaandA fallax respectively)and European eel (Anguilla anguilla) disappeared from ar-chaeological sites on the Atlantic drainage basins of Franceand Cantabria Spain during the LGM and probably survivedin refugia further to the south (Le Gall 2008) For the eelthis refugium was probably on the Atlantic coasts of Portu-gal and Morocco (eg Kettle et al 2008) The glacial refugeof the shad is unclear but remains have been discovered onthe upper Tagus River system at Aridos-1 from the Mindel-Riss interglacial approximately 300 ka BP (Morales 1980Le Gall 2000) and it is likely to have survived the LGM onthe Iberian Peninsula and northwest Africa From southernSpain the boreal gadid species haddock (Melanogrammusaeglefinus) and pollock (Pollachius pollachius) have beenreported by Rodrigo (1994) for the Cueva de Nerja nearMalaga dating to the LGM These results were significant be-cause these species are currently found in northwest Europeand the southernmost range for haddock and pollock is cur-rently the Bay of Biscay and northern Portugal respectively(Whitehead et al 1986) The most recent analyses of theLGM deposits from the Cueva de Nerja have revealed thatthe northern gadids (including saithePollachius virens codGadus morhua and lingMolva molva) make up more than30 of the identified fish remains and thus represented a sig-nificant presence among the exploited fish (Cortes-Sanchezet al 2008 Morales-Muniz and Rosello-Izquierdo 2008)The exploitation of gadids continued for an extended periodof time after the LGM until the early Holocene as sea levels

were returning to present levels and sea surface temperaturesapproached present values

Among the most spectacular shifts in species ranges havebeen reports of the remains of Atlantic salmon (Salmo salar)in archaeological sites from the Mediterranean drainagebasins of France Spain and Italy dating from the LGM up tothe early Holocene Additional small art objects unambigu-ously depicting salmon have been located at the Grand Grottede Bize on a Mediterranean drainage basin (Le Gall 1994a)These reports from late Palaeolithicndashearly Mesolithic sitesin France Spain and Italy (Juan-Muns i Plans 1985 Juan-Muns i Plans et al 1991) suggest the presence of residentsalmon populations in the western Mediterranean Sea Thecurrent southernmost range of Atlantic salmon in northwestEurope is northern Portugal which indicates that immigrantpopulations would have had to pass through the Straits ofGibraltar when temperatures in southern Spain during theLGM were similar to northern Europe at present Howeverthe archaeozoological evidence is contested Many of thesites in southern France are in proximity to Atlantic drainagebasins and it is possible that the Atlantic salmon remainsin the Mediterranean watersheds may have been transportedas part of a seasonal migration of fishermen (Le Gall 19831984 1992 1994a 1994b) This has even been invokedto explain Atlantic salmon finds in the more distant Men-ton caves on the French-Italian border (Riviere 1886 Clark1948) As well many of the early identifications of At-lantic salmon at the Caune de Belvis and the Grotte Jean-Pierre I by Desse and Granier (1976) and at Grimaldi (BarmeGrande II) have been subsequently questioned and revised totrout (Salmo trutta) by Le Gall (1994a) Desse-Berset (1994)and Desse and Desse-Berset (2002) respectively It is diffi-cult to distinguish Atlantic salmon and trout from their skele-tal elements and many of the early salmon identificationswere made mainly on the basis of the unusual size of verte-bral elements The revision of the initial identifications mayhave been made on reflection of the zoogeographic implica-tions of there being endemic Mediterranean Atlantic salmonpopulations during the LGM Up until now the prevailingparadigm has been that the SST in the western Mediterraneanduring the LGM was too high to meet the oxygen require-ments of the species (Le Gall 1983 1984 1994b) On theother hand recent unambiguous determinations of Atlanticsalmon in Solutrean levels in the Cueva de Nerja indicatethat ocean temperatures were cool enough during the LGMto permit the passage of the species through the Straits ofGibraltar (Morales-Muniz and Rosello-Izquierdo 2008)

Part of the reason why these archaeozoological finds areimportant is that there are few studies that have investigatedhow the geographic distribution of marine species changedduring the LGM and most of this is based on molecularmarkers Hence there are few a priori hypotheses of whatfish species should be found in archaeozoological collec-tions This situation contrasts strikingly with the state ofknowledge for terrestrial plants and animals and freshwater











37

1











12







38

1



(Levitus 2006) 4

5

6








39

1





































Table 1Continued




BP)





Plans et al (1991)












40

1

2











13

14









41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References
















































































































































37

1











12







38

1



(Levitus 2006) 4

5

6








39

1





































Table 1Continued




BP)





Plans et al (1991)












40

1

2











13

14









41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References








































































































































38

1



(Levitus 2006) 4

5

6








39

1





































Table 1Continued




BP)





Plans et al (1991)












40

1

2











13

14









41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References








































































































































39

1





































Table 1Continued




BP)





Plans et al (1991)












40

1

2











13

14









41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References


























































































































































Table 1Continued




BP)





Plans et al (1991)












40

1

2











13

14









41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References








































































































































40

1

2











13

14









41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References








































































































































41

1

2

3

4

5





10








42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References








































































































































42

1









10





4 Discussion




43

1









map 10

11








44

1






















46

1
























References








































































































































43

1









map 10

11








44

1






















46

1
























References








































































































































44

1






















46

1
























References











































































































































46

1
























References




















































































































































References







































































































































Refugia of marine ﬁsh in the northeast Atlantic during the ...A. J. Kettle et al.: Refugia of...

Documents

Transcript of Refugia of marine ﬁsh in the northeast Atlantic during the ...A. J. Kettle et al.: Refugia of...