Subfossil oaks from bogs in NW Europe as a
Transcription
Subfossil oaks from bogs in NW Europe as a
Subfossil oaks from bogs in NW Europe as a (dendro)archaeological archive HANNS HUBERT LEUSCHNER, UTE SASS-KLAASSEN Zusammenfassung: Über 2600 dendrochronologisch datierte Jahrringfolgen subfossiler Eichenstämme aus deutschen, niederländischen und irischen Mooren belegen den Zeitraum zwischen 6000 BC und 1700 AD. Die zeitliche Verteilung der Lebensspannen dieser Bäume weist im Vergleich innerhalb und zwischen Standorten mehr oder weniger deutliche Übereinstimmungen von Keimungs- und Absterbephasen (GDO-Phasen) auf. Beide Phänomene treten dabei tendenziell gleichzeitig und nicht zeitlich versetzt auf. Als Möglichkeit zur graphischen Erfassung solcher Phasen wurde daher das durchschnittliche Lebensalter aller Bäume in jedem Kalenderjahr berechnet (mean age-Chronologie). Ungestörtes Wachstum der Bäume/Wälder führt zu einem kontinuierlichen Anstieg des Durchschnittsalters, Verjüngungsphasen (Generationswechsel) bedingen einen Abfall der Kurve. Regionale mean-age-Kurven zeigen im europäischen Vergleich weitgehende Übereinstimmungen. Demnach ist Populationsdynamik der ehemaligen Moorwälder überwiegend durch überregionale klimatische Veränderungen mit GDO-Ereignissen in Feuchtphasen bedingt. Dendrochronologisch datierte Bohlenwege um 720 BC und 130 BC stimmen exakt mit markanten Abfällen in den mean-age-Chronologien überein – ein Hinweis auf den Einfluss des Klimas auch auf menschliche Aktivitäten. Abstract: The dendrochronological data set of absolutely dated subfossil oak trunks from Irish, Dutch and German bogs consists of some 2600 series. They cover the period from 6000 BC to 1700 AD. The distribution of the trees in time shows distinct changes in the frequency, germination and dying-off. The comparison of germination and dyingoff (GDO) phases within and between bog-oak sites shows that there is a tendency for both phenomena to occur not in separated phases but synchronously. One way to graphically represent germination and dying-off phases is to calculate the ‘mean age’ of all trees at every calendar year. Where trees are uniformly ageing the mean age chronology rises; recruitment of juvenile trees and dying-off of old trees causes the chronology to drop. Regional mean-age chronologies of the bog oaks contain similar elements, sometimes over long periods. This observation indicates common large-scale climate forcing. A strong link could be found between dendrochronologically dated bog trackways from the period around 720 BC and 130 BC and dendroclimatological results that indicate a change towards wetter conditions at the same periods. Keywords: bog oaks; climate; dendrochronology; population dynamics Introduction From 1970 onwards, European tree-ring laboratories have studied subfossil oak trunks preserved in bogs, river gravels and marine/brackish sediments. Tree-ring series of these trees have been used to compile ultra-long absolutely dated treering chronologies which extend back to 8400 BC (SPURK et al. 1998, LEUSCHNER 1992; PILCHER et al. 1984 JANSMA 1996). This paper considers these oaks that grew under difficult ecological conditions on the surface and margins of the peat. Fluctuations of the generally high ground-water table – partly triggered by climate – are assumed to have a major impact on population dynamics and tree growth on these sites. Obviously dendrochronologists wish to understand if there is a connection between climatic conditions and the distribution of the bog oaks through time. In the beginning, when only a few sites with subfossil oak findings were available there was evidence for synchronous dying-off phases of oaks in different bogs (LEUSCHNER et al. 1987). Such agreements in population dynamic would suggest climate control. And indeed, LEUSCHNER et al. (1987) found some indication that periods of simultane- ous dying-off episodes at more than one site were related to an increasing wetness of the climate. Overall, however, with information from more and more sites available, the picture tended to become blurred. Thus, with increasing information it becomes difficult in all but a few cases to see climate control; it is obvious that there are other, often local, factors involved. However, the huge amount of data that has been collected during the last 20 years in Germany, Ireland and the Netherlands enables to make comparisons on a bigger scale. In order to handle the large amounts of data involved in an inter-regional study some method has to be devised to integrate the information from different areas. The new variable ‘mean age’ is introduced because it combines both germination and dying-off events in a single parameter. It describes the variation in tree age through time and is used to detect common abrupt changes in population dynamics (germination and dying-off) of subfossil oaks from different regions in NW Europe. Regional tree-ring chronologies are used to detect (contemporary) changes in tree growth. This paper provides a general outline of the results, a detailed description can be found in LEUSCHNER et al. (2002). Subfossil oaks from bogs in NW Europe as a (dendro)archaeological archive 211 NORTH SEA IRELAND ENGLAND NETHERLD. GERMANY Fig. 1 Subfossil oak finds in Germany, The Netherlands and Ireland -6000 100 -5000 -4000 -3000 -2000 -1000 0 1000 -6000 -5000 -4000 -3000 -2000 -1000 0 1000 GERMANY 50 REPLICATION Beside such large scale comparisons, in situ finds of subfossil oaks provide the possibility to reconstruct local hydrological changes and bog developments. We demonstrate this potential for the site Hammah near Stade (LEUSCHNER & DELORME 1986). For the archaeologists it is interesting if periods where subfossil oaks indicate past climatic changes are also reflected in the record of human activities. German trackways that are dated by dendrochronology prove this link between the “natural” finds and archeological finds of oaks. 0 50 0 50 “Coastal” subset IRELAND NETHERLANDS 0 YEAR BC/AD Material Dendrochronologically dated tree-ring series of 2600 subfossil oaks from 197 sites located in Northern Germany (n = 1561, 101 sites), The Netherlands (n = 301, 43 sites) and Ireland (n = 741, 51 sites) were used in this study. The tree-ring series run from 6069 BC (Germany) to AD 1596 (Ireland). The German and Irish series are continuously replicated throughout most of this period, whereas the smaller collection of Dutch subfossil oaks shows a gap between 4700 BC and 3500 BC. Fig. 2 Replication of German (total and subset of low elevation sites), Irish and Dutch subfossil oaks The two oldest collections of Dutch subfossil oaks are dated against the German oak chronology (JANSMA 1996). All subfossil oak chronologies are dated against regional chronologies of archaeological and modern oak. The German material is further categorized into two subsets of “inland” (high elevation sites >2 m a.s.l and “coastal” (low elevation sites <2 a.s.l) material. Fig. 1 shows the locations of European subfossil oaks, Fig. 2 the temporal replication of the material. 212 Method The basic approach of the large-scale regional and temporal comparisons of population dynamics is the calculation of regional mean-age chronologies. The mean-age value for each given year is calculated as the arithmetic mean of the age of all single trees in this specific year. The value of a group of trees decreases when the older trees die while younger ones live on, or when young trees replace old trees. It increases when no population changes occur and the existing trees live on. In more general terms, this could be thought of as disturbed (mean age decreasing) and undisturbed (mean age increasing) populations. This is a robust way of conveying a lot of complex information and is useful for comparisons between regions. Mean age is therefore a tool to grasp the chronologically and spatially fuzzy distribution of generation changes in forests on a large scale in an objective way. Results and discussion Population dynamics of subfossil oaks The temporal distribution of the life spans of the North German subfossil oaks and the corresponding mean-age (MA) chronology is shown in Fig. 3. It is obvious that germination- as well as dying-off (GDO) phases on different sites occur more gradual instead of being abrupt changes in the population dynamics. However, the MA chronology clearly mirrors common changes in population dynamics at different sites. Even less distinct GDO phases are clearly reflected as declining phases in the MA chronology. As Fig. 4a shows, there is evidence for considerable similarity in the German and Dutch oak population, especially if the German material is reduced to the subset of low-elevation sites. This similarity led to the conclusion that it is valid to combine the German and Dutch oaks and to calculate a so-called “continental” MA chronology. This continental MA chronology is subsequently compared with the Irish MA chronology (Fig. 4b). The result is surprising: despite a distance of about 800 km between the two groups of subfossil oak a remarkable agreement can be seen between the continental and the Irish MA chronology in the early part of the record, from about 5500 BC to 2000 BC. There is not only a good match between the large scale GDO phases, e.g. between 4000-3900 BC, around 2500 BC, and in 2000 BC, but also a very good agreement in the high frequency variation between the two MA chronologies. It is very likely that phases with reduced mean Hanns Hubert Leuschner, Ute Sass-Klaassen age throughout this early period are the result of increasing wetness, whereas phases of increasing mean age may point to relatively dry conditions, with more young oaks establishing. After 2000 BC until 700 BC there are only episodic periods of agreement which may be no more than random. However, the two continental, the German and the Dutch MA chronology, show a good agreement until about 1300 BC (Fig. 4a). The Dutch and part of the German oaks grew on low-elevation sites whereas the Irish oaks generally grew on higher sites. This suggests that marine factors, i.e. sea-level changes and/or stewing back of river systems might have had influenced the hydrology on the low elevation sites in Germany and the Netherlands considerably. This could explain the obvious changes in the population dynamics of oaks at these low elevated sites. Population dynamics and human activity It is obvious from the comparison of the continental and Irish MA chronologies (Fig. 4b) that a large-scale event took place around 2000 BC that resulted in different population dynamics of oaks in continental and Irish mire woodlands. In this context it is interesting that there is a widespread indication of climate change in the later third millennium BC (DALFES et al. 1997). Moreover, there are suggestions that vegetation changes around and after 2000 BC are not related to changes in climate but are due to increasing human activity albeit probably driven by environmental conditions. As an example of the difficulty of separating human and natural factors, the notable drop in the Irish mean-age chronology at 950 BC can be taken: BAILLIE & BROWN (1996), who were exclusively looking at Irish evidence have noted that this decrease in the frequency of naturally preserved subfossil oaks in Ireland coincides with a major building phase involving oak trackways and settlements in different bog areas. Something they argue may have been due to overall drier conditions between ca. 1000 BC to 880 BC. However, German subfossil oaks show severe growth depression in the middle of the 10th century BC (LEUSCHNER et al. 2002). This wider context implies an environmental component, which could involve wetter conditions, something that could equally have contributed to the demise of Irish subfossil oaks at that time. Indications for another sudden change towards wetter climatic conditions in the first millennium BC can be found in dendrochronologically dated wood from continental archaeological excavations. Subfossil oaks from bogs in NW Europe as a (dendro)archaeological archive -6000 -5500 -5000 -4500 -4000 -3500 -3000 213 -2500 -2000 -1500 -1000 - 500 1 500 1000 -2500 -2000 -1500 -1000 - 500 1 500 1000 Sites > 2 m asl 200 150 YEARS Sites < 2 m asl MEAN AGE CURVE OF BOG SAMPLES 100 -6000 -5500 -5000 -4500 -4000 -3500 -3000 YEAR BC/AD Fig. 3 The life spans of subfossil oaks from Germany clustered according to their site provenance. The mean-age chronologies are given at the bottom. For reference, decrease of mean age (determined optically) is marked by lines 300 200 a 150 100 50 250 MEAN AGE [YR] The Netherlands / L. Sax (low elevation sites) 200 150 Continent / Ireland b 100 50 0 -6000 COINCIDENT MEAN AGE PHASES -5000 -4000 -3000 -2000 YEAR BC/AD -1000 1 1000 Fig. 4 Comparisons of macro-scale mean-age chronologies in common periods between a) German and Dutch and b) continental (all German and Dutch) and Irish subfossil oaks. Phases of good agreement are marked by green boxes 214 Hanns Hubert Leuschner, Ute Sass-Klaassen 100 200 300 G 500 400 1 2 3 4 600 5 700 6 oak pre 175 AD alder Index ring-width series, growth depressions coloured “events” 100 Starting with a clear germination phase (G), oaks grow together with the alders in a dry phase. 175 - 300 AD A rapidly expanding and rapidly growing raised bog “suffocates” the oaks and preserves the trunks. Boggrowth steps (1-6) are mirrored by growth depressions of “surviving” trees 350 - 700 AD after 700 AD Circles mark preserved pith/sapwood, arrows mark unknown number of rings/years in missing (rotten) wood 200 300 oak trunks and stumps 400 500 600 700 The site is as moist that almost no oaks grow in the alder carr. Only one oak with a strong growth depressions grows before 175 AD The last of the oaks, standing on sandy knolls, are covered by the rising bog Sphagnum peat Fen peat Mineral soil (sand) YEAR AD Fig. 5 Reconstruction of a the growth dynamic of the bog Kehdinger Moor near Hammah. Growth depressions of the tree-ring series and population dynamics of subfossil oaks hint at a stepwise growth of the raised bog One example involves two distant sites, an iron-age settlement in Biskupin, Poland, constructed around 720 BC (WAZNY 1994), which was abandoned most likely because the conditions became too wet, and trackways (e.g. 9 Le, 21 Le, 12 Ip) in Northwest Germany (SCHMIDT 1992; METZLER 1993), which were constructed between 720 BC and 710 BC. In the last case it seems to have been a rapid bog growth which implies that it was the change towards wetter climatic conditions that influenced the human construction activities. Another example relates to trackway constructions in Ireland in 148 BC (BAILLIE & BROWN 1996) and in Germany ca. 180 BC (FANSA & SCHNEIDER 1990) and 130 BC (FANSA 1992) which coincide with a major dying-off phase in the German subfossil oaks (DELORME et al. 1981; DELORME et al. 1983, see Fig. 3). We are looking foreward to results of the ongoing dendrochronological investigation of pine and oak samples from trackway 32 Pr in the Campemoor area which are found in context with naturally preserved subfossil pines at the same location. (BAUEROCHSE & METZLER 2001, this vol.). The average radiocarbon age of ca. 2930 cal BC is very close to a two-step rapid decrease of the German MA chronology in 2900 BC and 2800 BC, indicating dying-off phases in subfossil oaks. Here again it seems possible that large-scale climatic changes have influenced human activities. Reconstruction of local changes in hydrology According to LEUSCHNER et al. (2002) most changes in population dynamics of subfossil oaks correspond with contemporary long-term (up to decades) growth depressions in the affiliated regional treering chronologies. Especially at sites with in situ finds a combined evaluation of ring-width patterns and population dynamics allow an exact reconstruction of local hydrological changes. Such a reconstruction is demonstrated for the site Hammah in the Kehdinger Moor (LEUSCHNER & DELORME 1996). 32 subfossil oaks were dated by using dendrochronology, most of them lying on alder-carr peat at the base of raised-bog peat which covered the trunks. Other oaks were directly lying on the mineral soil at places where it exceeded the level of the carr peat. Fig. 5 shows the tree-ring series of the oaks and the reconstruction of the bog- and forest history of the site. There is clear evidence that both the sharp germination phase at 175 AD as well as the stepwise dying-off of the oaks between 300 AD and 700 AD correspond with phases of growth depression in the surviving oaks. There is an interesting tendency that especially these oaks which clearly mark a step of dying-off by showing growth depressions are the next to die. This can be interpreted as the effect of an increasing ground-water level, which induces the horizontal and/or vertical expansion of the bog. This Subfossil oaks from bogs in NW Europe as a (dendro)archaeological archive makes that oaks on higher locations and higher elavated stands are later reached by the expanding peat/bog. Such detailed analysis can lead to a better understanding of the dynamics of bog extension, which is mainly triggered by changes in hydrology and climate. However, to get a better understanding of this detailed information more research is needed about (1) the development, structure and ecology of different types of former mire woodlands and (2) aspects of conservation of tree remains. First attempts have been made by DELORME et al. (1983), LEUSCHNER et al. (1986), LEUSCHNER et al. (1987) and by LAGEARD et al. (1995). The latter excavated a mire woodland in Cheshire, Great Britain. More results will soon be available from two excavations in Ypenburg and Zwolle, the Netherlands. Oak supporting mire woodlands were very complex ecosystems with many sub-types that occurred due to differences in geographical location, topography, geology, soil characteristics and hydrology. However, all these sub-types can be sensitive indicators for changes in hydrology. Acknowledgements The research was supported by the “Environment and Climate Programme” under contract ENV4CT95-0127; by the Netherlands Organisation of Scientific Research (NWO/GW; 250-51-072 and NWO/AWL 750.700.04). We thank Mike Baillie, Queens University, Belfast, Northern Ireland and Esther Jansma, The Netherlands Centre for Dendrochronology, RING foundation for providing tree-ring data. References BAUEROCHSE, A. & METZLER, A. (2001): Landschaftswandel und Moorwegebau im Neolithikum in der südwestlichen Dümmer-Region. Telma 31: 105-133 BAILLIE, M.G.L. & BROWN, D.M. (1996): Dendrochronology of Irish Bog-Trackways. In RAFTERY, B. (Ed.): Trackway Excavations in the Mountdillon Bogs, Co. Longford. Irish Archaeological Wetland Unit, Transactions Vol. 3, Dept. of Archaeology, University College, Dublin: 395-402 DALFES, N.H., KUKLA, G. & WEISS, H. (1997): Third Millennium BC Climate Change and Old World Collapse. Springer, Berlin, 728 p. DELORME, A., LEUSCHNER, H.-H., HÖFLE, H.-CH. & TÜXEN, J. (1981): Über die Anwendung der Dendrochronologie in der Moorforschung am Beispiel subfossiler Eichenstämme aus niedersächsischen Mooren. Eiszeitalter und Gegenwart 31: 135-158 DELORME, A., LEUSCHNER, H.-H., TÜXEN, J. & HÖFLE, H.CH. (1983): Der subatlantische Torfeichenhorizont 215 “Sieden”, erneut belegt im Toten Moor am Steinhuder Meer. Telma 13: 33-51 FANSA, M. & SCHNEIDER, R. (1990): Neue Erkenntnisse über den Bohlenweg XXV (Pr) und den Pfahlsteg XXX (Pr) zwischen Damme und Hunteburg. Archäologische Mitteilungen aus Nordwestdeutschland 13, 17-26 FANSA, M. (1992): Moorarchäologie in Niedersachsen. Eine Einführung. Archäologische Mitteilungen aus Nordwestdeutschland 15: 5-22 JANSMA, E. (1996): An 1100-Year Tree-Ring Chronology of Oak for the Dutch Coastal Region. In DEAN, J.S., MEKO, D.M., SWETNAM, T.S. (Eds.): Tree-Rings, Environment and Humanity; Proceedings of the International Conference, Tucson, Arizona, 17-21 May 1994. Radiocarbon, Tucson: 769-778 LEUSCHNER, H.-H & DELORME, A. (1986): Dendrochronologische Befunde zu Torfeichen aus dem Kehdinger Moor bei Hammah, Kreis Stade.- In LANDKREIS STADE (Ed.): Landschaftsentwicklung und Besiedlungsgeschichte im Stader Raum, Stade: 183189 LEUSCHNER, H.-H., DELORME, A., TÜXEN, J. & HÖFLE, H.C. (1986): Über Eichenwaldhorizonte in küstennahen Mooren Ostfrieslands. Telma 16: 61-82 LEUSCHNER, H.H., DELORME A. & HÖFLE H.-C. (1987): Dendrochronological Study of Oak Trunks Found in Bogs in Northwest Germany. Proceedings of the International Symposium on ecological aspects of tree ring analysis, New York: 298-318 LEUSCHNER H.H. (1992): Subfossil Trees. In BARTHOLIN. T. (Ed.): Tree-Rings and Environment. Proceedings of the International Dendrochronological Symposium, Ystad, South Sweden. LUNDQUA Report 34: 193197 LEUSCHNER H.H., SASS-KLAASSEN, U., JANSMA, E., BAILLIE, M.G.L., & SPURK, M. (2002): Subfossil European bog oaks: population dynamics and long-term growth depressions as indicators of changes in the Holocene hydro-regime and climate. The Holocene 12 (6): 695706 METZLER, A. (1993): Zwei urgeschichtliche Wege im Campemoor, Ldkr. Vechta. Berichte zur Denkmalpflege in Niedersachsen 3: 114-116 PILCHER, J.R, BAILLIE, M.G.L., SCHMIDT, B. & BECKER, B. (1984): A 7272-Year Tree-Ring Chronology for Western Europe. Nature 312: 150-52 SCHMIDT, B. (1992): Hölzerne Moorwege als Untersuchungsobjekte für die Dendrochronologie. Archäologische Mitteilungen aus Nordwestdeutschland 15: 147-160 SPURK, M., FRIEDRICH, M., HOFMANN, J., REMMELE, S., FRENZEL, B., LEUSCHNER, H.H. & KROMER, B. (1998): Revisions and Extensions of the Hohenheim Oak and Pine Chronologies – New Evidence about the Timing of the younger Dryas/Preboreal -Transition. Radiocarbon 40(3): 1-10 WAZNY, T. (1994): Dendrochronology of Biskupin – absolute dating of the early Iron Age settlement. Bulletin of the Polish Academy of Sciences, Biological Sciences 42(3): 283-289 216 Hanns Hubert Leuschner, Ute Sass-Klaassen Addresses Dr. Hanns Hubert Leuschner Albrecht von Haller Institut für Pflanzenwissenschaften Abteilung Palynologie und Quartärwissenschaften – Labor für Dendrochronologie und Dendroklimatologie – Von-Siebold-Strasse 3a 37075 Göttingen Germany e-mail: hleusch@gwdg.de Dr. Ute Sass-Klaassen The Netherlands Centre for Dendrochronology RING Foundation Kerkstraat 1 3811 CV Amersfoort The Netherlands e-mail: ute.sassklaassen@wur.nl