Weichselian landscape development and human settlement in Mecklenburg-Vorpommern ( NE Germany )

This papet presents the results of recent studies on the Weichselian Lateglacial in Mecklen­ burg-Vorpommern (NE Germany). The develop­ ment of basins and valleys is illustrated with some examples. Soil formation could be demonstrated for the transition Pleniglacial-Lateglacial, the Allerod, and the Early Younget Dtyas. Aeolian processes occurred predominantly during the Younger Dr­ yas, whereas formation of cryogenic soil features is inferred for the colder periods of the Lateglacial. Detailed palynological analyses cover the complete Lateglacial and allow a reconsttuction of the vegeta­ tion in relation to climate. Archaeological findings give information on the development of the fauna and on human settlement. Finally an integrated model of landscape development is presented. 'Anschrift der Verfasser: PD Dr. THOMAS TERBERGER, Lehrstuhl für Urund Frühgeschichte, Ernst-MoritzArndt-Universität, Hans-Fallada-Straße 1, D-17487 Greifswald, Germany, terberge@uni-greifswald.de; Dr. PIM DE KLERK, Botanisches Institut, ErnstMoritz-Arndt-Universität, Grimmerstraße 88, D17487 Greifswald, Germany; Dr. HENRIK HELBIG, Landesamt für Geologie und Bergwesen SachsenAnhalt, Köthener Straße 34, D-06118 Halle (Saale), Germany; Dr. KNUT KAISER, Fachbereich Geographie, Universität Marburg, Deutschhausstraße 10, D35032 Marburg; Dr. PETER KÜHN, Institut für Bodenkunde und Bodenerhaltung, Justus-LiebigUniversität, Heinrich-Buff-Ring 26 (IFZ), D-35392 Giessen, Germany

provide a basis for geomorphological and pedological research of hitherto unknown extent.INGER 1985INGER , 1997;;LITT & STEBICH 1999;LITT et al. 2001;DE KLERK 2002, 2004 b;ERIKSEN 2002) (cf.Table 1 and the section "vegetation history and palaeoclimate").The present paper is neither suitable for a lengthy discussion on the terminological problems nor appropriate for the proposal of a solution out of the confusion.
In this text, therefore, the 'traditional' termi nology is used in a loose informal sense.The term "Boiling" is used according to the 'Dutch stratigraphy of HOEK (1997; cf.VAN GEEL et al. 1989) and thus can be correlated with the later part of the "Meiendorf" (cf.Table 1).In many cases both terms can be combined as "Boiling/ Meiendorf".The term "Allerod" is used in its 'classical' sense and thus encompasses both the "Bölling-Alleröd Komplex" of USINGER (1985) and the sequence of "Boiling Interstadial -Mit tlere Dryaszeit -Allerod Interstadial" or "Boiling -Oldest Dryas -Allerod" of BOKELMANN et al. (1983), BOCK et al. (1985) and LITT & STEBICH (1999) (cf.Table 1).The terms "Earliest Dryas" and "Earlier Dryas" are used according to the 'Dutch' stratigraphy of HOEK (1997) and VAN GEEL et al. (1989) (though the present authors do not completely agree on the dates of these phases as given in Table 1).The "Earliest Dryas" in this sense corresponds with the first part of the "Meiendorf" (cf.Table 1).Finally the con cept of "Younger Dryas" (though "Late Dryas" would be more appropriate) is used in its 'tradi tional' sense (cf.ISARIN 1997 and Table 1).
Other scientific methods used in the various disciplines discussed in this paper were de scribed in great detail in previous publications (e.g. HELBIG 1999a;KAISER 2001;DE KLERK 2002;KÜHN 2003a).1995; Fig. 1).Unfortunately, a reliable abso lute chronology of the Weichselian ice decay is still absent.The extensive deglaciation after the Mecklenburg Glacial Advance dates around 14000 14 C years BP according to radiocarbon data from the southern Baltic area (GÖRSDORF & KAISER 2001).
In recent years basins of all sizes as well as small valleys in Mecklenburg-Vorpommern were geomorphologically and palaeoecologically investi gated.Together with material from older stud ies there are now around 100 informative and mainly palynologically dated lacustrine/fluvial sequences available for stratigraphic statements (KAISER 2002(KAISER , 2004)).The recent investigations, in contrast to the older studies, are characterised by a higher number of analysed profiles per ba sin ("transect studies") and a higher resolution of individual profiles.
The investigated basins can be classified as large glaciolacustrine basins (former proglacial lakes, >100 km 2 ), medium-sized lakes (0.03-100 km 2 ), and kettle holes (<0.03 km 2 ).The development of the Lateglacial large lakes in the southern Baltic basin (e.g.LEMKE 1998) will not be discussed here.
During the last deglaciation around 14000 14 C years BP large proglacial lakes existed in the coastal region of NE Germany (cf.Fig. 1) in which at first layers of clay and silt sedimentated with thicknesses up to 5 m, followed by a deposition of fine-and medium-grained sand up to 15 m thickness (Fig. 2A).Locally shallowwater gyttja and aeolian sand occur that date in the early Lateglacial, indicating that the final phases of large lakes occurred in the Pleniglacial (KAISER 2001).Particular varied and welldated findings from the late Lateglacial and early Holocene were derived from the basins of the Ueckermünder Heide (the area of the so-called "Haffstausee"), the Altdarss, and the Barther Heide area (e.g.KAISER 2001;BOGEN et al. 2003;KÜHN 2003a;KAISER et al. 2003, submitted).These findings include soils, peat beds and shallow-water gyttja which indicate that during the Allerod and the early Younger Dryas more or less dry conditions persisted in the large glaciolacustrine basins; lakes were only locally present (often as the result of dead-ice melting).During the late Younger Dryas ba sin sands were redeposited by wind over large areas.In the early Holocene these large basins were terrestrial -though locally also lacustrinetelmatic -in character.In the Atlantic period parts of the glaciolacustrine basins came under marine influence and became integrated in the Baltic Sea or in coastal lagoons.
The development of medium-sized lake basins within the time-frame under consideration can be illustrated with the Endinger Bruch area (approx.12 km 2 ) within the nearly flat till-plains in northern Vorpommern (Figs. 1,2B).This area is with over 1000 geological profiles and a large quantity of chronological data one of the most intensively studied basins of this size in NE The sedimentation processes of medium-sized lakes and kettle holes in NE Germany are sum marised by KAISER (2001).About 90 % of these basins had a beginning of sedimentation in the Lateglacial, 38 % in the Allerod.In general basin-forming dead-ice melting processes can be dated from the Pleniglacial up to the Early Holocene.An earlier phase of intensified deadice melting can be placed in the Earliest Dryas and Bölling/Meiendorf and a main concentra tion in the late Allerod and the early Younger Dryas.Over a third of the profiles include basal peat mainly from the Allerod, which ended regularly in a secondary position due to settling as the result of dead-ice melting.Characteristic for the Lateglacial is deposition of silicate gyt tja, with the exception of peat and of gyttja deposits rich in carbonates and organic mat ter (mainly originating from the Allerod).The dominant silicate input during the Lateglacial is due to an only thin vegetation cover, disorder in the fluvial regime (braiding), and an unstable overall relief (abluation, gully erosion, dead-ice melting).Sedimentation of organic gyttja, cal careous gyttja and peat characterises the early Holocene.This is mainly due to a reduction of silicate input and an increase in primary production.Many basins show a prominent increase in lake-levels during the Younger Dryas (cf.e.g.KAISER 1996KAISER , 2001, in press b;, in press b;HELBIG 1999a;LORENZ 2003).
The fluvial structures in Mecklenburg-Vor pommern can be divided into a grid of large NE-SW-and NW-SE-oriented valleys in the east (i.e.following the direction of the glaciers, or following a direction along the glacier front) and into a 'chaotic network of small valleys' in the Mecklenburg Lake District in the west and south (cf.Fig. 1).Situated along the eastern and Often they are covered by a layer of silty coversand ("'Geschiebedecksand'") (Fig. 3): this is the upper part of the till that was influenced by former periglacial processes and by Lateglacial and Holocene pedogenesis (HELBIG 1999a(HELBIG , 1999b;;KÜHN 2003aKÜHN , 2003b)).
The grain size distribution of the fissure fillings as well as quartz grain surface analyses clearly demonstrate that the filling material neither is eroded till material nor silty coversand, but that it is of aeolian origin.The filling sand shows of ten features of lessivation (i.e.brownish, vertical transferred clay is visible).The upper part of the sand structures is often disturbed.This is prob ably the result of solimixtion and of disturbance of the active layer by former permafrost or deep seasonally frozen ground (cf.KOZARSKI 1974;HOFFMANN & BLUME 1977;MACKAY 1993).
Thickness of the disturbed zone and/or the silty coversand (including tilled soil) ranges between 30-70 cm; mean thickness is 50 cm.Many sand structures are more or less deformed.MURTON & FRENCH (1993) describe a downslope bend ing as a phenomenon of thawing sand-wedge sand that moved in downslope direction with the overlying material.
At the moment it is not possible to say whether the forms in Vorpommern are sand wedges or seasonally frozen ground soil wedges.Large, well developed sand wedges are known to in dicate permafrost conditions (MURTON 1996).Smaller sand-filled wedges, on the other hand, also may occur outside the range of permafrost (DIJKMANS 1989;MURTON 1996;MURTON et al. 2000).Direct dating of the filling sands by means of radiocarbon dating or thermoluminescence was not succesful (HELBIG 1999a).
Considering the present knowledge on the Lateglacial palaeoenvironment of Vorpommern the most favourable conditions of formation were during the period between the deglacia tion and the beginning of the Lateglacial and during the Younger Dryas.Because most of the features in Vorpommern are not large and well developed, there are three possibilities for their genesis: 1: Formation might have been under condi tions of continuous permafrost during the late Pleniglacial.In that case large forms could not develop because too little time remained until temperatures rose at the beginning of the Late glacial.
2: Formation was under conditions of discon tinuous permafrost during the late Pleniglacial and early Lateglacial or Younger Dryas.In that case the higher temperatures (compared to con tinuous permafrost conditions) as well as the too short time-span limited the growth.3: The structures were formed under conditions of strong seasonally frozen grounds during the late Pleniglacial and/or early Lateglacial or the Younger Dryas: then large forms could not de velop because temperatures were too high.
A main or second phase of sand wedge deve lopment during the Younger Dryas cannot be excluded (KÜHN 2003b).A prerequisite for the formation of primary mineral-filled frost wedges is aeolian sand transport on a snowfree surface in winter (DIJKMANS 1989)  Vorpommern during the formation of the fis sures.However, sand-wedge formation might have occurred at wind-exposed and snow-free sites, which therefore might only indicate dry microclimatic conditions.

Cryoturbations
Most cryoturbations are located in sediments of basins and valleys because of their high amount of soil water and a layering of sediments of different texture (HELBIG 1999a).Typical are both diapirs and drop-like forms.On higher till plains deformations in sand-filled fissures occur.
The fissures show bump-like deformations due to thaw and freeze cycles.
In two cases cryoturbations could be positively dated in the Younger Dryas: at the locations "Bansin" (HELBIG 1999a) and "Grimmen 4" (HELBIG & DE KLERK 2002).At some other lo calities cryoturbations occur in sand that also is likely to date from the Younger Dryas (HELBIG During recent years dunes and sand sheets were identified in the large glaciolacustrine basins of the Ueckermünder Heide, the Rostocker Heide, the Barther Heide and the Altdarss area that could be attributed to the Younger Dryas with various dating methods (KAISER 2001;KÜHN 2003a;LUDWIG 2002;BOGEN et al. 2003;KAISER et al. submitted;cf. Fig. 4).Furthermore there are indications of some local aeolian activities during the Older Dryas.As a rule, the aeolian sands of the Younger Dryas cover terrestrial soils ("Finow-Soil", "Usselo-Soil"), peat, or gyttja.The various dates allow the assumption that stable land surfaces still persisted in the early Younger Dryas: the aeolian processes therefore, must be placed in the later Younger Dryas and the Younger Dryas-Prebo- The lithological features allow the reconstruc tion of aeolian episodes/events.The aeolian sediments can be parallelised with the "younger cover sands" as found in NW Germany, NW Denmark and in The Netherlands (KOLSTRUP 1991;KASSE 1999).Until now there is no evi dence found in Mecklenburg-Vorpommern of aeolian activity during the early Holocene as is known from Brandenburg and W Poland (BUSSEMER et al. 1998;SCHIRMER 1999) with the exception of the Elbe river valley (CASPERS & SCHWARZ 1998).
A prominent morphological differentiation of aeolian formations is observable: in the northwestern and southern regions of Meck lenburg-Vorpommern there is a dominance of aeolian sand covers up to 4 m in thickness with weakly developed relief and low dunes, whereas some prominent dune complexes of up to about 25 m in height exist in the southwestern and the eastern regions.It is uncertain whether this re flects a palaeoclimatic differentiation ("oceanic/ continental", cf.BÖSE 1991) or whether a more or less random cause should be assumed, i.e. dependent of the form and consistency of the aeolian drift areas.The position of aeolian sand sheets and dune fields in the medium-scaled relief, the orientation of dated longitudinal and parabolic dunes, and measurements of the direction of deposition suggest that sand was mainly transported from west to east during the Younger Dryas (KAISER 2001).

Soil formation
Research into Lateglacial soil-genetic processes was twofold: on the one hand studies concen trated on soils influenced by (ground)water, on the other hand relic and buried soils of ter restrial sites were investigated.During the first vegetation phase in the Lategla cial, the Open vegetation phase I (Earliest Dr yas or earlier phase of the Meiendorf; cf.Fig. 5), the upland was inhabited by a vegetation of pre dominantly herbs and (dwarf)shrubs.Inciden tally, also Betula trees must have been present, as could be demonstrated in the Reinberg researches (DE KLERK et al. 2001).This phase starts probably with an only minor temperature rise compared to the preceding Pleniglacial (DE KLERK et al. 2001) and, thus, must be inter preted as still relatively cold compared to the following Lateglacial vegetation phases.
During the Hippophae phase (Bölling/Meiendorf) Hippophae thickets expanded prominently in this open landscape.As Hippophae needs re lative high summer temperatures for flowering (KOLSTRUP 1979), this phase is interpreted to have been relatively warm (DE KLERK 2002;DE KLERK et al. 2001).The fact that Hippophae shrubs were not outshadowed by expanding Betula trees suggests that climate was too dry for the latter taxon to expand considerably (cf.USINGER 1997), or that soil conditions were still immature (cf.VAN GEEL 1996).& STOLZE (2002) shows that this vegetation phase is not or only very slight discernible in the southern pollen diagrams.This is interpreted to be the result of higher summer temperatures in the southern regions due to which Hippophae was able to maintain itself for a longer period, or in these regions climate and/or environment was more favourable for Betula trees which con sequently expanded earlier than in the northern regions and suppressed Hippophae.Sediment data from the Reinberg basin indicate that during the Lateglacial precipitation had gradually increased (DE KLERK et al. 2001).As a result of this, and/or as the result of the devel opment of more mature soil conditions, Betula trees were able to expand in northern Vorpom mern at the beginning of the Lateglacial Betula/ Pinus forest phase (Allerod).The upland vegeta tion, however, remained predominantly open for some time (cf.Fig. 5).A sharp decrease in NAP-values around 1340 cm depth marks the sudden closing of the birch forests: these now dominated the upland.The low values of pol len attributable to Pinus indicate an only minor presence of pine trees during the Lateglacial Betula/Pinus forest phase.This is in contrast with pollen diagrams from S Mecklenburg, Brandenburg and the Berlin area, which show a prominent Pinus phase in the later period of the Allerod.This difference is attributed to a domi nance of pine on the sandy soils of the promi nent outwash-plains in the south, whereas the northern till plains carried a forest of predomi nantly birch trees (THEUERKAUF 2003).The sharp decrease in CaC0 3 -content in core Hoher Birkengraben around 1285 cm depth (cf.Fig. 5) might reflect a sedimentary reaction on a short-time acidification of the environ ment after the Laacher See eruption: the 'Re-inberg C pollen diagram includes the Laacher See tephra at a palynostratigraphic level similar to the decline of CaC0 3 -content in Fig. 5.An analysis of many pollen diagrams from NE Ger many indicates that the vegetation probably did not directly react on a decrease of temperatures induced by the volcanic eruption; increased precipitation and consequent temporal water level increases, as well as extremely intensive lightning activity and consequent increased fire might have temporarily disturbed the vegeta tion on a small scale (cf.THEUERKAUF 2003; DE KLERK et al. submitted).

Open vegetation phase II (Earlier Dryas
The pollen diagram of Fig. 5 shows that during Open vegetation phase III (Younger Dryas) an open upland vegetation consisting of Artemisia, Ericales, several shrubs such as Juniperus, Hip pophae, dwarf birches, and probably shrubs of Salix, persisted.Trees of Betula and Pinus most likely were present as single specimens, not forming closed stands.NAP-values show that maximum openness occurred at the end of this phase.As far as openness is related to cold, this contradicts studies that place maximum cold ness of the Younger Dryas at its beginning (cf.e.g.ISARIN 1997;ISARIN & BOHNCKE 1999).
Openness, however, may also be the result of drought.
It is remarkable that in pollen diagrams from N Vorpommern the Younger Dryas is represented without any exception with thicker sediments than those of the Allerod, while in general (but not always) a change from organic sediments to clastic sediments occurs.In the southern regions, on the contrary, the sediments from the Allerod always are the thicker ones (as far as these are completely preserved), and a change to clastic sediments normally does not occur.It is interpreted that these differences are caused by a denser vegetation (at least along the basin slopes) in the southern regions due to which soil erosion was severely hampered and conse quently thick clastic sediments were not washed into the basins.This denser vegetation also is attributed to slightly warmer conditions in the southern regions than in the north (cf.DE KLERK 2002;DE KLERK & STOLZE 2002).
The sharp decrease of NAP-values at the transi tion to the Early Holocene Betula/Pinus forest phase (Preboreal; cf.Fig. 5) shows that the up land vegetation suddenly closed: dense forests of birch and pine trees developed rapidly.The veg etation probably reacted on the rise in tempera ture without major time-lag, as the forests could develop quickly from the incidental stands of birch and pine already present in the previous open phase (cf. HOEK 2001).At the begin ning of the Early Holocene Betula/Pinus forest phase several Lateglacial 'relic' plants still must have been present, as can be concluded from the values of pollen attributable to Juniperus, Betula shrubs, Artemisia, and Ericales (cf.Fig. 5).Along the basin margins a lake-shore veg etation consisting ofTypha latifolia, Equisetum, Filipendula and later Poaceae (possibly Phrag mites) seem to have developed.A short, open vegetation phase during the Early Holocene as reconstructed for other regions in Europe (BEHRE 1978; VAN GEEL et al. 1981) is not registered in the diagram Hoher Birkengraben: the conspicuous peak of pollen attributable to wild grasses is interpreted as a pollen signal of the wetland vegetation and not as an expansion of upland grasses (cf.DE KLERK 2002).The observed differences between the northern and the southern regions indicate that during the Lateglacial a sharp climate gradient existed in NE Germany.A present gradient is explai nable by the influence of the Baltic Sea, but it is more difficult to explain such phenomenon for a period where the Baltic Sea did not yet exist.Steep climate gradients in Europe during the Lateglacial, however, might be due to the existence of steep temperature gradients between the ice cover in Scandinavia and the prominent continental climate on the ice-free continent (cf.

Development of the fauna
As evidence on the Lateglacial and early Holocene faunal record of NE Germany is limited, it is necessary to integrate information from neighbouring areas as well.Recent studies demonstrated that earlier dates of important faunal finds based on stratigraphic or palyno logical observations alone are sometimes unre liable.An AMS-date of the aurochs skeleton from Potsdam-Schlaatz (Brandenburg), tradi tionally dated to the Younger Dryas, now as cribes the find to the early Preboreal (BENECKE 2002).Although direct dating makes progress (cf.e.g.LILJEGREN & EKSTRÖM 1996), there is still unreliable information incorporated in the Lateglacial faunal record of N Germany and S Scandinavia (Fig. 6).
The large mammal fauna of the early open vege tation phase (Bölling/Meiendorf) was probably dominated by reindeer (Rangifer tarandus).This is demonstrated by numerous reindeer remains form the well known Meiendorf site north of Hamburg (KRAUSE 1937).Also in Denmark there is evidence of this species for that period since ca.12500 14 C years BP (AARIS-SORENSEN 1992;HOLM 1996).In Scania reindeer is not present before 12000 BP because there was probably previously no landbridge (continu ously) established for the seasonal migrations of the species at that time.At Meiendorf reindeer remains from the Bölling/Meiendorf are accompanied by few horse (Equus ferus) remains, whereas two elk (Alces alces) bones detected above the find layer cannot be reliabely ascribed to the Boiling/ Meiendorf (cf.KRAUSE 1937).An isolated AMS-date of elk remains from Scania suggests presence of the species in northern regions around ca. 12400 BP (LILJEGREN & EKSTRÖM 1996).It is also possible that wolf (Cams lupus) already lived in the Danish territory at the end of the Bölling/Meiendorf (AARIS-SORENSEN 1992, 146).Smaller faunal elements like hare, grouse, some birds and fish also contrib uted to the faunal record in Schleswig-Holstein (KRAUSE 1937).

Remains of mammoth (Mammuthus primigenius)
were not found at archaeological sites from the Bölling/Meiendorf in the north.The youngest mammoth finds from South Scandinavia are dated at ca. 13000 14 C years BP and it is argued that the species closely followed the retreating The faunal record in Vorpommern during the Allerod is demonstrated at the site Endingen VI where some elk bones prove hunting of the spe cies.Elk is also represented at Klein-Nordende (Schleswig-Holstein) dated to ca. 12000 1 4 C years BP (BOKELMANN et al. 1983).Because elk is also present at sites in Denmark it can be as sumed that this species was an important hunt ing game in the early Allerod (Fig. 6).A worked giant deer (Megaloceros giganteus) antler piece from Endingen VI and another isolated find of a worked skull of giant deer from Lüdersdorf suggest that this large herbivore also lived in the Allerod landscape of NE Germany.Some giant deer remains from Denmark and Scania dem onstrate that this species also existed in these re gions around 11600 14 C years BP (cf.KAISER et al. 1999).The presence of a herbivore with a pos sible antler width of more than 3.5 m indicates that the forest cannot have been very dense and that open environments still must have existed in the Allerod.This interpretation is supported There is increasing evidence that at the start of the Preboreal the fauna in N Germany and S Scandinavia became more variable with elk, aurochs, wild pig and red deer (Fig. 6; BE NECKE 2004).An isolated find of an aurochs from Potsdam-Schlaatz demonstrates successful hunting of that species (BENECKE 2002;2004;GRAMSCH 2004).The most impressive example of successful hunting in the early Preboreal is represented by the deposits of elk bones from a kettle hole near Lundby bog on Zealand (MOLLER HANSEN et al. 2004).Some faunal remains from aurochs, red deer and wild pig at this site might be somewhat younger.There are no sites in N Germany that prove the existence of reindeer in the faunal record of the Preboreal, although some individuals have to be expected: systematic radicarbon dating demonstrates that this species was still present in Scania in the early Preboreal (LILJEGREN & EKSTRÖM 1996).

Human settlement
The earliest settlement of N Germany and S Scandinavia is connected with the Hamburgian dated in the Bölling/Meiendorf.Basic knowl edge on this hunter-gatherer technocomplex was obtained by excavations of the Meiendorf site in the 1930s (RUST 1937).Nowadays the Hamburgian can be divided into two groups: the early classical Hamburgian -characterised by shouldered points and known from sev eral sites in Schleswig-Holstein -and a younger phase with Havelte points (CLAUSEN 1997;ERIKSEN 2002).Recent discoveries support the idea that the early Hamburgian was possibly also present further to the north in Jutland (HOLM 1996(HOLM , 2003)), which might indicate that hunters moved northward more rapidly than hitherto expected following the reindeers with seasonal camp moves only.But until now no site of the classical Hamburgian was found in the area glaciated during the Weichselian in NE Germany (Fig. 7a).
It is known for a long time that the Hamburgian with shouldered points is also present east of the  (TERBERGER 1996a;1997).Isolated organic finds from Brandenburg like a worked reindeer antler from Großwuster witz (nowadays lost) probably are elements of that phase.At the same time a Zinken from a site close to the Nebel valley north of Güstrow probably is a first indication for the Ham burgian in the Mecklenburg area (Fig. 7a, 13) (TERBERGER 1997;TERBERGER & LÜBKE 2004).
If this assumption is correct a broad zone of (early) Hamburgian from the eastern to the western lowlands has to be expected.But it can be imagined that people avoided the Weichse lian till plains in the early Hamburgian phase as is suggested by evidence from neighbouring areas.Then the (early) Hamburgian is restricted to more southern areas of Mecklenburg and Brandenburg.
New excavations at the Ahrenshöft site in northern Schleswig-Holstein for the first time gave clear evidence for a sequence of the classical Hamburgian and the Havelte phase (CLAUSEN 1997).Palynological analysis dates the Havelte phase at Ahrenshöft in the later Bölling/Meiendorf around ca 12300 -12000 14 C years BP (USINGER 1997).Due to detections of the 1980s and 1990s Havelte sites are known from Jutland (HOLM 1996(HOLM , 2003) ) and Zealand (VANG PETERSEN & JOHANSEN 1996;JOHANSEN 2000;ERIKSEN 2002) and so they show a dis tribution from the Netherlands to the Danish islands.In contrast to the classical Hamburgian the Havelte sites have a distribution more in NW Europe and no evidence is found east of the river Oder (cf. CLAUSEN 1997).Against this background it can be expected that Havelte sites occur at the eastern boarder of this technocom plex in NE Germany.
Although animal remains are only available from few sites these indicate dominance of rein- deer hunting for the Hamburgians.The used hunting weapons are still a matter of discussion: it is possible that the Hamburgian were the first people using bow and arrow for their hunt, as the slender Havelte point looks suitable as an ar row tip and as bones with embedded fragments of projectile points indicate shots with remark able energy (BRATLUND 1990;HOLM 2003).It is proposed that Hamburgian sites are often connected with the migration routes of rein deers (VANG PETERESEN & JOHANSEN 1996).
Seasonal information from sites is so limited that it is not possible to develop a reliable set tlement model for that time.It is discussed that the Hamburgian sites in Denmark represent only short visits and do not reflect a systematic use for a longer time (cf. ERIKSEN 2002).
At the beginning of the Allerod a general change took place in the environment and in people's way of life.The development of a birch-pine forest was connected with a fauna of woodland and open landscape elements.In general the raw material for the stone artefacts of the Federmesser technocomplex is of limited quality and the assemblages are characterised by a simpler blade technology with shorter and irregular blades.Short scrapers and Federmesser are typical elements and the backed points were probably used for arrow tips.Roe deer antler was also usually taken as organic raw material.Information on the Federmesser complex is main ly known from other areas (e.g.BAALES 2000; PELEGRIN 2000; FAGNART & COUDRET 2000), but is limited in N Germany and S Scandinavia (Fig. 7b).Transfering information from other areas should be carried out with great care (JO HANSEN 2000;ERIKSEN 2002).
Two investigated elk hunter camps at the site Klein-Nordende CR in Schleswig-Holstein (BOKELMANN et al. 1983) date the beginning of the Federmesser technocomplex around 12000 14 C year BP/12000 cal.year BC which corresponds with the end of the Boiling/ Meiendorf and the transition to the Allerod, although an age of the humous find layer prior to the Allerod was discussed (BOKELMANN et al. 1983).Possibly there is an overlap with the Havelte phase of the Hamburgian as Havelte points and Federmesser were found together in Ahrenshöft LA 58 (CLAUSEN 1997).Site Alt Duvenstedt LA 120B (Schleswig-Holstein) contains a Federmesser inventory with faunal elements like beaver and fox; charcoal from a hearth is dated slightly younger around 11800 14 C years BP (CLAUSEN 2004).
The elk hunter site of Endingen VI (Vorpom mern) was detected at the beginning of the 20 th Century (cf.TERBERGER 1996a; STREET 1996).
The faunal assemblage was AMS-dated to the early Allerod (TERBERGER 1996a;KAISER et al. 1999).Of special interest are a worked horse rib and a piece of giant deer antler.This find and a giant deer skull with traces of processing from W Mecklenburg suggest that giant deer was of some importance for hunting and as source of raw material (KAISER et al. 1999).Unfortu nately no lithics were collected at Endingen VI.Though some surface sites with backed points exist (e.g.BERLEKAMP 1957), a more important Federmesser site was detected in Brandenburg (GRAMSCH 1969): a typical relatively small as semblage with backed points and a sand stone smoother was found here, organic remains were not preserved.
the Allerod (CLAUSEN 2004).The find dem onstrates that care is necessary in the schematic connection of "typical" implements with technocomplexes as the cultural context of the find is unclear at the moment.
In the later Allerod the early tanged point com plex raises with the Brommean in S Scandina via.Numerous sites with the typical large tan ged points, scrapers and burins are known from Denmark and Scania that are interpreted as evi dence of a year round exploitation (JOHANSEN 2000;ERIKSEN 2002).Few animal remains suggest that elk was the main hunting game.In N Germany comparable assemblages are known for a longer time that are assigned to the Segebro-Bromme-Gruppe and Tolk-Sprenge-Gruppe by TAU TE (1968).Geostratigraphical and palynological data date a find layer near Hamburg with Feder messer and Lyngby/ Bromme points to the later Allerod or slightly younger (SCHWABEDISSEN 1957;TAUTE 1968).This date was confirmed by an early radiocarbon date of around 11450 14 C years BP of charcoal from the find layer.More recently discovered well-stratified Brommean sites are missing in N Germany.One should be cautious not to equate all finds of larger tanged points with the Brommean (Fig. 7b).
The Ahrensburgian is traditionally connected with reindeer hunters of the Younger Dryas.A closer view at the material shows that there are several tanged point finds (Fig. 7c) but only very few dated sites with rare faunal information.It is likely that species like elk were typical faunal elements at that time.So the idea of reindeer hunting people with Ahrensburgian points emerging with the start of the Younger Dryas has to be questioned.On the background of these facts it is more probable that elk hunt ers of the late Allerod developed smaller tanged arrow points and formed the basis for the Ah rensburgian of the Younger Dryas.
There are other observations that show that care has to be taken in the connection of so called type fossils with special time periods, e.g. the reindeer antler axe from Klappholz mentioned above.Also results from direct AMS-dating of bone points from the northern lowlands show some differences compared to the expected ages: e.g. a point of Maglemosian character was dated to the Younger Dryas (CZIESLA 2004).
Despite this new information it is in general correct to connect Ahrensburgian sites with the Younger Dryas.Probably most of the isolated reindeer remains of NE Germany date from this period.Several tanged point sites were found in the Saaler Bodden area, near Lake Kummerower See, near Stralsund, and in the Ueck ermünde Heide area (Fig. 7c) (TAUTE 1968;KAISER & TERBERGER 1996;TERBERGER 1996b;BOGEN et al. 2003).The topography of such sites is best illustrated with the situation near Verchen.The outflow of the river Peene from Lake Kummerow See nearly forms a peninsula where some sites are positioned close to the wa ter on both sides of the river.The frequency of tanged points indicates intensive hunting activities, which probably took place during the migration seasons of reindeer in spring and/or autumn.Unfortunantely there are no organic finds available for AMS-dating.A small tanged point inventory was excavated in the 1990s at the site Nienhagen (KAISER & TER BERGER 1996; cf.Fig. 1).The assemblage was characterised by a heavy retouched large tanged point accompanied by small projectiles like a simple retouched point and a backed piece.The stratigraphic position of the find layer argues for a dating to the Younger Dryas.On typological evidence, however, some kind of affiliation with the Brommian industry cannot be ruled out.1996;JOHANSEN & STAPERT 2000), but adds a more dynamic view on the ca.1200 years of the Younger Dryas.The idea is supported by the fact that only very few tanged point finds are known from Dennmark (ERIKSEN 2002), whereas also the British peninsula saw only very restricted human activities (BARTON 2002;BAR TON et al. 2003).Recently some faunal remains in Scania were identified that are connected with the presence of hunter groups (LARSSON et al. 2002) and -though finds with clear traces of human activities is limited -indicate human presence in the late Allerod/ early Younger Dr yas, the middle Younger Dryas, and in the late Younger Dryas/ early Preboreal.This is a further indication for episodes of presence in the north ern areas during more favourable conditions (LARSSON et al. 2002).
Some assemblages with a long blade industry from N France and S Britain are assigned to the late Younger Dryas and/or the transition to the early Preboreal (cf.e.g.BARTON 2002).These are characterised by very long blades that  1996b;GRAMSCH 1987;CZIESLA 2001b).
often show signs of heavy use along the blade edges.These assemblages show similarities with the Eggstedt-Stellmoor-Gruppe of TAUTE (1968;cf BAALES 1996;BARTON 2002).Recently a lithic assemblage from Jutland was recognised with typical elements of a long blade industry (So-RENSEN & STERNKE 2004).Although absolute dating of the site was not possible a connection with the Eggstedt-Stellmoor-Gruppe and long blade sites in NW Europe is very probable.Therefore it seems plausible to propose a comparable de velopment of the assemblages in N Germany and S Scandinavia at the transition from the late Younger Dryas to the Holocene.The sites are normally connected with high quality flint sources.It still has to be cleared in more detail whether the long blade assemblages reflect only specialised flint workshops or whether they can also be interpreted in a chronological sense (cf.BARTON 2002;SORENSEN & STERNKE 2004).
Unfortunateley there is very limited informa tion available to characterise the subsistence strategy of these sites.There is an ongoing debate whether there was a major population shift towards the north at the beginning of the Preboreal.It has been assumed that tanged point technocomplexes in Scan dinavia that emerge at the very late Younger Dryas/ early Holocene originate from more southern areas (cf. e.g. TAUTE 1968;KINDGREN 1996).Recently SCHILD ( 2001) and KOBUSIE-WICZ (2004) presented arguments for the idea that new people were immigrating the Polish lowland at the beginning of the Mesolithic.
There is, however, increasing archaeological evidence from the transition to the Holocene and the early Preboreal.Sites such as the long blade assemblage mentioned above, the aurochs find from Potsdam-Schlaatz (BENECKE 2002) and the elk bone deposits from Lundby bog with 14 C-dates of ca.9950 -9860 BP/ ca.9600 -9250 cal.BC (MOLLER HANSEN et al. 2004) suggest that people lived in N Germany and 5 Scandinavia at that time (BENECKE 2002;2004;MOLLER HANSEN et al. 2004;SORENSEN 6 STERNKE 2004;TERBERGER 2004;cf. Fig. 8).
Of high relevance is the find assemblage from Lundby bog with an early Maglemosian charac ter (MOLLER HANSEN et al. 2004).An elk antler adze finds its parallels at the early Mesolithic site Star Carr and in the lower Mesolithc level of Friesack.This evidence is in some contradic tion to the ideas of surviving "Epi-Ahrensburgian" (reindeer) hunter groups "during the first 500 years or so of the Holocene" (JOHANSEN & STAPERT2000, p. 84).
Relying on the new information presented by MOLLER HANSEN et al. (2004) CLAUSEN 1998;2004;TERBERGER 1996b)  correlation of Lateglacial events and processes in Mecklenburg-Vorpommern with the oxygen isotope curve of the Greenland GPJP ice-core and the sequence of Greenland Stadiais (GS) and Greenland Interstadials (GI) as proposed by BJÖRCK et al. (1998) is given.This is based on a preliminary correlation of the Greenland ice-core stratigraphy with the pollen diagram 'Hoher Birkengraben' (Fig. 5) presented by BILLWITZ et al. (2000).The chronology in 14 C years as presented in Fig. 9

Fig. 1 :
Fig. 1: Landscape structures in Mecklenburg-Vorpommern and the location of the study areas discussed in the text.
till beds and four ma jor Weichselian terminal moraines are present in Mecklenburg-Vorpommern (MÜLLER et al.

Fig 3 :
Fig 3: Genesis of relic sand fissures under influence of the active layer and of pedogenetic processes.

Fig. 4 :
Fig. 4: Dated dunes and cover sand profiles in Northeast Germany.

Fig. 5 :
Fig. 5: Pollen diagram 'Hoher Birkengraben from the Endinger Bruch, selected pollen types (modified aftet DE KLERK 2002).The curves are displayed with an additional 5-times exaggeration (open curves with depth bars).Solid curves: types included in pollen sum; dotted curves: types excluded from pollen sum (for pollen frequency calculation and pollen type nomenclature see DE KLERK 2002).For details on pollen type nomenclature (f, m, p, *) see DE KLERK (2002).
) was again dominated by herbs and (dwarf)shrubs.Especially some minor diagrams from the End inger Bruch area(DE KLERK 2002) indicate prominent stands of Juniperus.This indicates that this vegetation phase represents a vegeta tion regression, which is probably the result of a decrease in summer temperatures(DE KLERK 2002).A comparison of pollen diagrams from northern Vorpommern with those of S Meck lenburg, N Brandenburg and the Berlin area carried out by DEKLERK (2002) and DE KLERK
LITT,T.& STEBICH, M. (1999): Bio-and chronostratigraphy of the lateglacial in the Eifel region, Germany.-Quaternary Interna- Afterwards a new water body did not develop in the Reinberg basin: only slowly ris ing water levels induced the development of a small peatland in which sedge/brownmoss peat was formed.
The Earlier Dryas is characterised by deposi tion of predominantly clastic gyttja and fluvial sand that indicate a local sedimentary differen tiation.The sediments of the Allerod (organic gyttja, calcareous gyttja, and peat) allow the a maximum at the end of the Younger Dryas.A sudden lowering of water levels at the beginning of the Holocene is responsible for a sedimentary hiatus, which is a widespread phenomenon in small basins in NE Germany (cf.DE KLERK 2004 a).

Relic cryogenic macrofeatures 2.2.1 Sand-filled fissures and sand wedges
(KAISER et al. 2002te glacial aeolian sediments are within the large glaciolacustrine basins in the coastal region of today's Baltic Sea and in the distal parts of the large outwash plains of the Mecklenburg Lake District (cf.Fig.1).After an initial formation during the Late Pleistocene/Early Holocene, the aeolian landscapes were strong influenced by man-made erosion and accumulation during the Late Holocene(KAISER et al. 2002).An exceptional feature in the Weichselian till plains of Vorpommern is the occurrence of some loess deposits; there are, however, no detailed sedimentological analyses of these sedi VANDENBERGHE (1988)9BRAMER 1961)ons described in this paper and byKLIEWE & SCHULTZ (1970)belong to the smaller forms (amplitude <0.6 m) sensuVANDENBERGHE (1988), who states that these are an indication for a Mean Annual Air Tem perature (MAAT) 5 -1°C.With the due caution because of the only few hard datings, this value may count as indication for the temperature regime in Vorpommern during the Younger Dryas.2.3 Aeolian processesAeolian sand sheets and sand dune complexes in Mecklenburg-Vorpommern cover a total area of approximately 450 km 2 (SEELER 1962).Dune fields mainly occur in the SW part of the Federal State (extension ca.300 km 2 ) on the Weichselian outwash plains and on the adja cent Saalian till plains.ments(cf.SCHÖNHALS 1944;BRAMER 1961).A study from NW Poland demonstrates that on a local scale sedimentation of loess might have occurred in the late Pleniglacial north of the Pomeranian terminal moraines (ISSMER 1999).The accumulation of silt in ca.30 % of the topsoil-profiles (silty coversand: see above) in dicates that some aeolian activity also occurred on a broader scale.Similar finds can be found in buried soils and lacustrine sediments.Ice wedg es filled with fine sands indicate aeolian activity during the Pleniglacial and early Lateglacial or Younger Dryas on the till plains.

Paak 9700 ealBC -9240 ealBC 460 68« Paak «573 calBC -9290 ealBC 283 50X Paak 9527 calBC -9113 ealBC 214
P [r-.l]- DE KLERK (2002)ation with the program Calpal (see www.calpal.de)andslightlydeviatesfrom the (preliminary) chronology of the vegetation phases of Vorpom mern as proposed byDE KLERK (2002).The knowledge on the environment and proc esses of the late Pleniglacial and early Lateglacial is still limited: especially the deglaciation proc esses of NE Germany between ca.20 000 and 14000 14 C years BP is still badly dated.There is, however, an increasing amount of available data on the Lateglacial.Two pollen diagrams cover the earliest Lateglacial vegetation phases, whereas with the discovery of the 'Reinberg ho rizon' now the first known soil formation proc esses date at the transition from the Pleniglacial to the Lateglacial.Unfortunately too little data are available to say whether this is an exception or whether similar palaeosols also developed at other locations.A first phase of intensified dead-ice melting is placed in the Earliest Dryas and Bölling/Meiendorf.Soil erosion is high as consequence of the open vegetation; indication on Aaeolian activity, however, is limited.Also the knowledge on the dominant fauna in the Early Lateglacial still is limited.Though there is increasing evidence for the classical Ham burgian and the Havelte phase after ca.12600 cal.yearsBC in NW Germany, Denmark and Poland, sites still are missing in Mecklenburg-Vorpommern.More data are available on the Allerod.Lake levels lowered and fluvial activity decreased.As consequence of the expansion of birch/ pine forests soil erosion decreased and mainly organic sediments deposited in basins.The de velopment of peat layers was a consequence of increasing bioproduction under wet conditions, often in connection with a lowering of the lake level.A second phase of intensified dead-ice melting is proposed for this period.There are several observations of palaeosols that date in this phase.The earliest reliable evidence of hu man settlement is related with this period: elk shows to be the typical hunting game.Remains of giant deer and horse indicate the existence of open parts in the landscape.Up to the beginning of the Younger Dryas Gleysols and Histosols have been prevalent in depressions.Arenosols (Regosols), weakly de veloped Podzols and Cambisols on sandy parent material as well as weakly developed Luvisols on till plains are common on sites not influenced by groundwater.In the open landscape of the Younger Dryas soil erosion increased again, whereas fluvial dynam ics enlarged.Lake levels predominantly rose and mainly clastic sediments were deposited.There are clear indications for increasing aeolian activ ity mainly in the second part of the Younger Dryas. A alaeosol in the Altdarss area was formed in an early phase of the Younger Dryas.Reindeer seems to be the predominant faunal element, but horse was also present.New infor mation from N Germany suggests that the early Ahrensburgian industry developed in the late Allerod, but most Ahrensburgian sites belong to the Younger Dryas.The transition to the Holocene shows a clos ing of forests, a lowering of water levels and the sedimentation of organic sediments as well as formation of peat.Soil erosion decreased.Probably the final melting of dead ice in NE Germany took place in the Preboreal.Remains of aurochs and elk indicate these animals as prominent faunal elements.New archaeological finds suggest a change of hunting species, mate rial culture, and the way of living at that time.restrial Late Glacial Fauna in South Swe den.-In:LARSSON, L. (ed.),The Earliest Settlement of Scandinavia and its relation ship with neighbouring areas.Acta Archaeologica Lundensia Series, 8/24: 135-139; Stockholm.LITT, T. (2003): Environmental response to cli mate and human impact in central Europe during the last 15,000 years -a German contribution to PAGES-PEPIII.-Quater nary Science Reviews, 22: 1-4; Amsterdam.