ARTICLE IN PRESS

Dendrochronologia 24 (2007) 5360 www.elsevier.de/dendro

INVITED REVIEW PAPER

Human time in tree rings

Dieter Eckstein
University of Hamburg, Department of Wood Science, Division of Wood Biology, Germany Received 10 March 2006; accepted 7 May 2006

Abstract
Human time means time in the context of human culture and activities. It is shown how human time is biologically archived in and dendrochronologically extracted from tree rings. For illustration, examples in the late medieval Hanseatic city of Lu beck are selected. r 2006 Elsevier GmbH. All rights reserved.
Keywords: Dendrochronology; Cultural heritage; Human behavior; Socio-economic environment

Introduction
Time passes continuously and invariably. Only by dividing it into intervals does time become perceptible for man. Most trees divide their lifetimes into annual intervals by forming distinct growth rings. In doing so, trees record and archive information about time and environment life-long like in an almanac. It is the task of dendrochronologists to translate this information into an understandable human language. A single tree-ring series has, without any doubt, a value in itself. But by means of high-tech equipment and powerful computers, it is nowadays easily possible to collect huge amounts of dendrochronological data and to achieve a novel quality of results with them that is to say, information about past human cultures and behavior as well as on the socio-economic environment. In the following, I will substantiate some aspects of this topic within a real context but before that, few incipient considerations about tree growth may be helpful.
Department of Wood Science, Leuschnerstr. 91, D-21031 Hamburg, Germany. Tel.: +49 40 73962 452; fax: +49 40 42891 2835. E-mail address: d.eckstein@holz.uni-hamburg.de.

Incipient considerations
For trees, the pulse generator of the time is the annual alternation between a growing and a dormant season; in our temperate climate this changeover is normally forced by temperature. In other areas of our globe, like in the subtropics, this is caused by the seasonally changing moisture availability. But for trees in ever wet tropical rainforests, such a natural external pulse generator is sometimes missing so that time has to be articially introduced in a tree, e.g., by wounding the cambium from time to time. These timestamps can be seen after the tree has been felled and opened. More difcult than to detect the time is to recognize the simultaneously recorded non-chronological information in tree rings, such as information on environmental changes. It may be encoded in the width of a tree ring, or in the latewood and earlywood width, or in wood density and wood structure, or in the chemical composition of the cell wall, and enters a tree through its leaves and roots because these organs are in close contact with the environment. In the end, the environmental input is transformed into new wood through the cambium and its activity, varying from year to year, site to site, and tree species to tree species (Fig. 1). It may be

1125-7865/$ - see front matter r 2006 Elsevier GmbH. All rights reserved. doi:10.1016/j.dendro.2006.10.001

ARTICLE IN PRESS
54 D. Eckstein / Dendrochronologia 24 (2007) 5360

Fig. 1. Interaction between external and internal inuences on wood formation, and electron-microscopic photo of cell divisions (arrow heads) in the cambium zone (inset).

worth mentioning that the annual character of tree rings was known and their varying width was interpreted as meteorological almanac long before the nature of the cambium was identied. The information about past human cultures, such as timber trade, forest management practices, or preferences in timber selection, is recorded in all living trees, but even more exciting, in all trees which lived at earlier times and are preserved either underground or on the sea bottom, or were used as construction timber, or in wooden objects of art (Fig. 2), such as paintings on panels, organs, and other wooden music instruments, furniture, sculptures and the like.

and the Baltic Sea was densely covered by mixed deciduous woodlands of mainly beech and oak and that a document from 1188 AD (enacted by the German Emperor Barbarossa) safeguarded the right of the citizens of Lu beck to cut and use trees (Holzungsrecht) within an area of 1500 km2, but prohibited its utilization for ship building and for export. With the subsequent demographic evolution, the forests were nevertheless increasingly used for producing fodder for animal husbandry, timber production for house building, rewood for heating homes and for daily cooking. Initially, all secular buildings were made of wood. But after two severe res in 1251 and 1276 AD, a law was enacted prohibiting the erection of solely wooden houses. Due to the fact that the country was poor in natural stones, the fabrication of brick stones became necessary, consuming an additional amount of wood for the generation of energy. As early as 1220 AD, a law (Sachsenspiegel) disclosed the principle of sustainability (wood land should remain wood land). From this year on, a certain extent of forest management could be assumed in the region. This background information about the availability of and the demand for timber must be kept in mind for the following considerations.

Acquisition of building timber from a market or from the forest?

During our dendrochronological studies in and around Lu beck, we were repeatedly confronted by the questions: Was the building timber selected directly among the standing trees in the forest or was there a timber market as early as in the 13th century? No answers to these questions were found in the archives available. Provided that the standing trees were selected in the forest and then felled and transported directly to the building site, all timbers used within one building are presumably cut in the same year. However, if the trees were rst transported to a storage place and only then sold to a building contractor, timbers from different felling campaigns would have been mixed, and different felling years should be expected within one and the same complex of buildings. We found examples for both alternatives (Fig. 4). In one of the houses, originally an ecclesiastic building, four out of seven timbers of the roof construction had bark preserved and their youngest ring was formed in 1317 AD. According to our assumption the building timber would have been selected in the forest, cut and brought to the building plot and built in without any storage time in-between. The alternative example is a merchants house of the Gothic brick stone architecture. The 15 dated timbers can be grouped into four felling operations, ranging

beck Study object the late medieval town of Lu and its hinterland
Lu beck was the leading town of the Hanseatic League, a network of traders and merchants from late medieval to early modern times around the Baltic Sea including England in the west and western Russia in the east. The UNESCO has declared Lu beck as a World Cultural Heritage site. Old Lu beck was founded as a Slavic stronghold (Fig. 3), dendrochronologically dated to around 800 AD, at the western edge of the Slavic settling activities. About 300 years later, the German King Henry the Lion occupied this area and in 1143 AD his liege, earl Adolf II, founded todays Lu beck only a few kilometers further south of Old Lu beck. Within the subsequent 100 years, a settlement evolved with an orderly system of streets and plots of land, ve churches, a monastery, a hospital, and a fortication. From forest history we know that the country between the North Sea

ARTICLE IN PRESS
D. Eckstein / Dendrochronologia 24 (2007) 5360 55

Fig. 2. A selection of trees from former times: (a) post-medieval woodland; (b) early Iron Age settlement Biskupin, Poland, dated to 738 BC; (c) carpenters at work on a late-medieval building site (Chroniques de Hainaut I (1448) in: Binding (1993)); (d) Rembrandt painting on an oak panel dated to around 1630 AD; (e) organ in the St. Jacobi church in Lu beck dated to 1500 AD; (f) oak chest in the monastery of Ebstorf, N. Germany, dated to around 1338 AD.

from June 1284 to the winter 1285/86 AD, which means over 33 months. These timbers were randomly distributed within the roof construction. In such a case of mixed felling dates, we assume that the timbers were provided from a local timber market. Around the end of the 13th century, the timbers for three upper-class houses were identied to have presumably been selected from standing trees in the forest. According to the records, the occupants of these houses were wealthy merchants and possibly owners of forest land or had good contact to forest owners. Conversely, at about the same time six building contractors, belonging to the lower social classes, apparently bought the building material at a timber market. This reality can be traced dendrochronologically up to the 17th century.

Summer or winter felling?

In our dating work in Lu beck, we often found as in the last, example that the youngest growth ring under the bark was not complete. Such trees were felled during the summer season, but when and why? In Fig. 5 the dynamics of oak growth during one season is illustrated for North Germany: The x-axis is the time-scale of one vegetation period and the y-axis shows the accumulated radial biweekly increments; some phenological key events are indicated by arrows. Oak in northern Germany starts its cambial activity around the end of April/beginning of May. Further growth is characterized by different intensities, as shown by the S-shaped line. The earlywood/latewood transition occurs around the end of June/beginning of July. The tree-ring width is

ARTICLE IN PRESS
56 D. Eckstein / Dendrochronologia 24 (2007) 5360

Fig. 3. Lu beck, Slavic rampart and todays Old Town (Degn and Muu, 1965, 1968): (a) circular wall (see arrow) from around 800 AD; (b) section through the wall; (c) Old Town; (d) cloaca made from timbers from around 1252 AD in secondary use.

Fig. 4. Lu beck, house Parade 1, uniformly dated to 1317 AD (top); house Hundestr. 94, felling dates between June 1284 and winter 1285/86 AD; B means bark ring existing.

completed between late August and mid-September; the lignication of the cell walls follows with some delay. On this basis, the cutting time within the growing season can be xed rather precisely, at best even to the month.

Between the 13th and 17th century, approximately 25% of the building timbers were cut during the summer season, although in summer the farm workers were most likely occupied with farm work, and felling and transport of trees were presumably difcult because

ARTICLE IN PRESS
D. Eckstein / Dendrochronologia 24 (2007) 5360
July 7 8

57

3.0

radial growth [mm]

2.0
5 1 swelling buds 4 3 2 bud-break 3 initiation of phloem formation 4 mature leaves 5 latewood formation 6 stop of phloem growth 7 partial autumn colouring of leaves 8 total colouring of leaves

1.0
2 1

Fig. 6. Lu beck, median and range of the ages of rafters within one construction unit in relation to time (50 roof constructions, sum of rafters 401).

28.4.12.5.26.5. 9.6. 23.6. 2.7. 21.7. 5.8. 18.8. 2.9. 15.9.29.9.13.10. date of sampling

Fig. 5. Wood formation of oak in northern Germany during one vegetation season and cross-section through a developing tree ring in July (insert); onset of earlywood formation around end of April/beginning of May; end of wood formation between late August and mid-September.

the trees were with foliage and the timber was moist and heavy and more susceptible to attacks by insects and fungi. A possible reason for summer felling could have been the need for bark required for tanning animal skins to produce leather. Bark can be peeled off much easier in summer than in winter. However, bark for tanning was reportedly obtained from young oaks grown up in coppice forests because it contains a higher content of tannin compared to old oaks. But the construction timbers were, of course, not obtained from coppice woodlands. Whatever reason may have been really valid, oak bark was a much-valued raw material for the numerous known medieval tanneries in Lu beck.

Between 1300 and 1500 AD, the range of the tree ages around the median was low compared with the time before and after. Up to 1300 and after 1500 AD, slowgrown oaks prevailed, whereas the oaks over the intervening years were fast-grown. Were the forests over-exploited until 1300 AD due to a high demand f or construction timbers after the big res in 1251 and 1276 AD and due to a general economic upswing? The young oaks cut in the mid-15th century germinated after 1350 AD and right at the same time as the oaks cut during the building boom in the second half of the 16th century. Because of a depression phase in building activities around 1500 AD, the second group of oaks was conceded a longer lifetime. From such a point of view, the occurrence of young oaks could be interpreted as a beginning recovery and regeneration of the forest after heavy devastation and possibly as evidence for a sustainable forest management practice.

Timber import or domestic wood? Timber properties

From various features of the building timbers, such as age, dimensions, or age trend reected in their tree-ring width, the age of the trees used for rafters will be singled out here (Fig. 6). The median as well as the maximum and minimum age of all dated rafters per construction unit are plotted against calendar years resulting in several clusters and in at least one long-lasting gap (13401430 AD) of building activities which was partly caused by consequences of the Black Death epidemic around 1350 AD. It is striking that in the 13th century, 130-year-old oaks were felled whereas later on the tree age dropped down to 60 before it increased again to 120 years in the 16th century. In a town like Lu beck, a Hanseatic and harbor town with far-reaching trade connections, one should not be surprised to nd imported timber. The rst cases of dendrochronologically proven long-distance timber transport in Central Europe are reported for Dorestad/ The Netherlands and Haithabu/North Germany. These settlements are dated to the 9th century and located at waterways. However, these timbers were not purposefully imported but were transported as barrels and served as containers for all kinds of goods. At their destination they were re-used as enclosures for wooden wells. The dating and the determination of the origin of oak wood in Lu beck will be illustrated with the Triumphal

ARTICLE IN PRESS
58 D. Eckstein / Dendrochronologia 24 (2007) 5360

Fig. 7. Lu beck, Cathedral, Triumphal Cross, and Wooden Screen, made by Bernt Notke (14401509 AD) and his workshop; gure of John, open back (bottom row left) and in head-down position (bottom row right) as it was carved from an oak trunk; bar diagram, showing the dating results for the Triumphal Cross (the results for the Wooden Screen are not shown here) (photos by U.v. Ulmann).

Cross and Screen in the Cathedral, made by Bernt Notke and his workshop (Fig. 7). The imposing dimensions of the individual gures (up to 2.30 m tall) as well as the large amount of wood raise the question of how a wood-working medieval workshop proceeded in the planning, material acquisition and treatment and what technological knowledge presumably was already available in the 15th century. An inscription on the carrying beam revealed that the Triumphal Cross was erected in 1477 AD. During restoration in the 1970s, writing was found on the inside of the sealing boards of Mary and John, stating that the rear opening of the hollow gures had been closed in 14711472 AD, respectively. The 170-year-old oak used for sculpturing the gure of John was felled, according to dendrochronology, in the winter 1470/71 AD; because this gure was sealed in 1472 AD, only 1 year passed between felling and completion of the carving work. Based on the direction of knot cavities of former branches, it was even possible to prove for the two upright gures Mary and

John that they were carved head-down out of the trunks (Fig. 7). The time remaining until 1477 AD was required for drying the wood and painting the surfaces (polychroming). For the Screen, all plank-shaped wood parts were radially cut out of the oak trunks. Panels made in this manner do not warp when the wood moisture changes depending on changing air humidity. None of these boards contained sapwood. Some of them originate from 200300 year-old, slow-grown oak trees. In Fig. 8, the southern coast area of the Baltic Sea, including the location of Lu beck and of ve regional oak chronologies, is mapped. The tree-ring mean curve, made of the tall gures, is compared with these ve reference chronologies; the darker the color of the circles, the higher the similarity between the tree-ring series compared. Hence all the wood for gure carving is of domestic origin. Also the tree-ring mean curve, made of 20 boards of the Screen, is compared with the reference chronologies. On the basis of evidence

ARTICLE IN PRESS
D. Eckstein / Dendrochronologia 24 (2007) 5360 59

provided, the plank-shaped oak wood of the Screen originated from Polish/Baltic forests. Also, the oak timbers of other 15th century objects in Lu beck, such as the organ in the St. Jacobi church and the wooden

screen in the Holy Ghost Hospital, were imported from that same area.

Know-how in woodworking
From our study of the Triumphal Cross and Screen we deduced that fresh, unseasoned domestic oak wood was used for this late medieval sculpture carving. Suitable-sized trunks of oaks were carefully selected and utilized optimally with regard to their conical shaft shape and diameter, including the entire sapwood, because timber was a high-valued resource and short raw material. The plank-shaped wood for the Screen originated from up to 300 years old, slowly and homogeneously grown oaks. The boards were without branches and radially cut; the easily perishable sapwood had been removed, very likely just after cleaving, indicating a well-developed experience in wood treatment. The planks were obtained from a distant source, because the domestic oak trees were presumably too knotty and crooked as well as of wider and irregular tree-ring width for producing the desired plank quality. The oak trees were felled in Polish/Baltic forests, cleaved, oated down the rivers to the harbors, such as Gdansk, Riga, or Ko nigsberg, where the boards were loaded onto sea-going ships and brought to the trading centers in western Europe (Fig. 9).

Fig. 8. Map showing the location of Lu beck () and the approximate areas of ve regional reference chronologies of oak (O); comparison of the reference chronologies with the mean curves assembled from the tree-ring series of the ve tall gures (top) and of 20 boards of the Screen (bottom).

Conclusions
The contribution of dendrochronology to the enlightenment of our past, beyond pure dating, has briey been

Fig. 9. Trade routes of Polish/Baltic oak to Western Europe (left); the Gdansk harbor (top, right) in the 18th century niez (Jakrzewska-S ko, 1995); rafting on a river near Dordrecht (bottom, right) in 1810 AD (Scheifele, 1996).

ARTICLE IN PRESS
60 D. Eckstein / Dendrochronologia 24 (2007) 5360

illustrated with a selection of examples in a narrow areal and temporal section. In any case, such an approach begins with the dating of human remains, such as houses or objects of art. Only then is it possible to attempt to discover expressions of human behavior. Once a sufciently large dendrochronological data set for a specic pool of human artifacts in a dened geographical area and time window has been compiled, one can start asking questions regarding general and abstract expressions of past human life: e.g., what was the total amount of building timber used? how large was the forest area needed for timber acquisition? which forest management practices were applied? was there a serious timber shortage? and the like. If this paper succeeds in encouraging and inspiring others to pursue this line of research, it has achieved its purpose.

present this paper. I am also whole-heartedly grateful to Sigrid Wrobel, Hamburg, and Tomasz Wazny, Torun, who both contributed substantially to this work. But I am also aware of many colleagues around the world who deserve to be mentioned; however, it is not possible to be exhaustive at this point.

References
Binding, G., 1993. Baubetrieb im Mittelalter. Wissenschaftliche Buchgesellschaft, Darmstadt, 530pp. Degn, C., Muu, U., 1965. Eine Landeskunde in 80 farbigen Luftaufnahmen. Wachholtz, Neumu nster, 77pp. Degn, C., Muu, U., 1968. Eine Landeskunde in 72 Luftaufnahmen. Wachholtz, Neumu nster, 82pp. niez sk w dawnych rycinach. Jakrzewska-S ko, Z., 1995. Gdan w-Gdan sk-odz Ossolineum, Wrocaw-Warszawa-Krako , 147pp. Scheifele, M., 1996. Als die Wa lder auf Reisen gingen. Braun, Karlsruhe, 226pp.

Acknowledgments
I am much obliged to Manuela Romagnoli, Viterbo, who organized EuroDendro2005 and invited me to