Session_7.Rmd

# Session 7: Novel methods for dating and provenance analysis {.unnumbered}

## S7-O-001 {.unnumbered}

**Advancing Elm dendrochronology in the United Kingdom**

Neil J. Loader^1^, Danny McCarroll^1^, Daniel Miles^2^, Martin Bridge^3^, Darren Davies^1^, Christopher Bronk Ramsey^2^, Cathy Tyers^4^

*^1^ Department of Geography, Swansea University, Swansea, United Kingdom. ^2^ Department of Archaeology, University of Oxford, Oxford, United Kingdom. ^3^ Institute of Archaeology, University College London, London, United Kingdom. ^4^ Scientific Dating Team, Historic England, Swindon, United Kingdom.*

<n.j.loader@Swansea.ac.uk>

The application of stable isotopes for dendrochronology provides additional possibilities for secure dating of historic timbers and wooden artefacts. This paper explores the wider application of the technique, initially developed for oak, to date elm (Ulmus spp). Elm is common in the UK historic buildings but it is challenging to date.

In contrast to the many thousand oak tree-ring series dendrochronologically-dated over the last 40 years, Bridge (2020) reported only 4 instances of successful dating of elm in the UK. A subsequent survey of more than 70 buildings found that whilst elm site sequences could be derived, dendrochronological matches against oak reference chronologies were either not strong enough to be considered secure or shown to be unreliable by radiocarbon dating.

These problems arise in the UK due to a lack of elm reference chronologies, an absence of long-lived extant trees and the tendency of elm to exhibit highly variable (disturbance) growth patterns. The result of these limitations is that elm often remains unsampled or fails to date. Consequently a large part of the UK historic buildings archive remains unexplored.

Early application of stable isotope dendrochronology to samples of elm indicate that the strong common signal preserved in oak tree-ring stable isotopes can be used to securely date elm and may also provide an approach for developing ringwidth chronologies for this genus. New opportunities now exist to explore the chronology, distribution and use of elm in the UK.

Bridge, M. 2020. *Vernacular Architecture* 51 (1) DOI: <https://doi.org/10.1080/03055477.2020.1794245>.


## S7-O-002 {.unnumbered}

**Dendroprovenancing historical oak timber at a small geographic scale using a multi-variable approach: a preliminary study on living oak trees**

Roberta D’Andrea^1^, Christophe Corona^2,\ 3^, Anne Poszwa^4^, Alan Crivellaro^5,\ 6,\ 7^, Christelle Belingard^1^, Fabien Cerbelaud^1^, Rémi Crouzevialle^1^, Guy Costa^8^, Sandrine Paradis-Grenouillet^1,\ 7^

*^1^ Laboratoire de Géographie Physique et Environnementale, Université de Limoges, Limoges, France. ^2^  Laboratoire de Géographie Physique et Environnementale, Université Clermont Auvergne, Clermont-Ferrand, France. ^3^ French National Centre for Scientific Research – CNRS, France. ^4^ Laboratoire Interdisciplinaire des Environnements Continentaux, Université de Lorraine, Nancy, France. ^5^ Department of Geography, University of Cambridge, United Kingdom. ^6^ Forest Biometrics Laboratory, Faculty of Forestry, University of Suceava, Ukraine. ^7^ Éveha, Bureau d’étude archéologique, Limoges, France. ^8^ Laboratoire PEIRENE, Université de Limoges, Limoges, France.*

<roberta.dandrea@etu.unilim.fr>

A study focusing on timber-framed houses in Limoges (France) is currently being carried out to pinpoint the exact source areas of construction timbers, and consequently to understand how local forests were exploited and managed for housing needs between the 15th and the 19th century. To do so, four sub-regional reference chronologies are being created using living oak trees and old wooden structures of the French department Haute-Vienne. Following the example of several dendroprovenancing studies carried out in the last years, we developed a multi-proxy method which we expect to help locate the origin of timber with more precision. This method includes wood anatomical variables, stable strontium isotopes (^87^Sr/^86^Sr) signatures and DNA and it is tested for the first time at such a small geographic scale. Before conducting these analyses on archaeological timber, we used the living oak trees included in the four sub-regional chronologies to test the potential of each parameter. We present the results of the quantitative wood anatomical (QWA) study and the preliminary results of the isotopic analyses. As regards the QWA study, time series of anatomical variables have been produced, and the principal component gradient analysis (PCGA) approach (Buras *et al.* 2016) has been used to cluster the series based on ecological gradients. Moreover, stable strontium isotopes ratios have been measured in a subsample from each site. The results show that combining TRW with vessel-related variables and ^87^Sr/^86^Sr ratios allows to assign trees to their origin more efficiently than using TRW alone.


## S7-O-003 {.unnumbered}

**Fingerprinting the provenance of large wood in rivers**

Javier del Hoyo^1^, Torsten Vennemann^1^, Marceline Vauridel^1^, Virginia Ruiz-Villanueva^1^

*^1^ University of Lausanne, Institute of Earth Surface Dynamics, Quartier Mouline, Lausanne, Switserland.*

<Javier.delhoyo@unil.ch>

Methods to decipher wood origin have been developed for decades in many fields (e.g., dendroacheology, wood commercialization, oceanic driftwood). Still, inferring the origin of wood in rivers (i.e., instream large wood, LW) is frequently overlooked. LW's presence in fluvial ecosystems enhances its geomorphology and biodiversity but also increases potential risk during floods. Hence, knowledge about its source is essential for understanding LW dynamics and optimizing river and riparian forest management. This project aims at developing a fingerprinting technique to decipher LW origin in mid-size river catchments (i.e., 1000 -- 5000 km2).

Therefore, we tested stable isotopes coming from the water molecule: D/H and ^18^O/^16^O. They show spatial variations due to fractionation during evaporation-precipitation processes. A tree absorbs water and stores a specific isotopic signal in the cellulose associated with its location; consequently, this signal can be used to infer its provenance once recruited and transported through the river network.

We selected a 50 km reach of the Rhone river, between the Lake Geneva and the Genissiat dam, where all incoming wood is retained.

Preliminary results revealed significant differences between the two main wood supply areas in the basin, the Arve and the Valserine tributaries; these differences were stronger in the most recent tree rings.

Other tracers related to geology (i.e., minor and trace elements) will be analyzed and combined with the isotopic analyses, to reduce uncertainty and more robustly infer the origin of LW.

This method could be easily extrapolated to several application fields of dendroprovenance dealing with similar spatial scales


## S7-O-004 {.unnumbered}

**Oxygen Isotopes assist dating buildings in Sussex (UK)**

Martin Bridge^1,\ 2^, Neil Loader^3^

*^1^ UCL Institute of Archaeology, London, United Kingdom. ^2^ Oxford Dendrochronology Laboratory, Mapledurham, Oxfordshire, United Kingdom. ^3^ Geography Department, University of Swansea, Swansea, United Kingdom.*

<martin.bridge@ucl.ac.uk>

Oxygen isotope assisted dendrochronology is becoming a useful additional tool in dating both 'complacent' and shorter ring width series. A study focussing on reconstructing climate over the last millennium in a relatively small area required historical information from building timbers, many of which could not be dated using conventional dendrochronology. This study is looking at buildings at the early and late parts of the millennium, where not many buildings have been dated, and using the available timber to date buildings that could not otherwise be dated, while providing climatic evidence for the wider study.

Other examples from southern England have already provided dates not available through conventional ring width techniques, and species other than oak are also being assessed.


## S7-O-005 {.unnumbered}

**Exact dating of the first Europeans in the Americas**

Michael Dee^1^, Margot Kuitems^1^

*^1^ Centre for Isotope Research, University of Groningen, Groningen, the Netherlands.*

<m.w.dee@rug.nl>

Dendrochronology is a precise dating technique, but it requires long sequences of tree rings and a master record for both the species and region in question. Groningen have been pioneering a new approach to dating that combines the precision of dendrochronology with the versatility of radiocarbon dating. It is based on spikes in the annual radiocarbon record, thought to be caused by enormous solar storms. Several such spikes have already been found in archives of known-age wood. Thus by finding one in an archaeological sample, it becomes possible to date the item exactly. The method is applicable to many species and may only require a handful of growth rings. We recently used it to date the earliest evidence for Europeans in the Americas. Three samples were obtained from the Norse layers at L'Anse aux Meadows, Canada, all of which exhibited cut-marks made by metal blades, a material not manufactured by the local indigenous people. In each case, the 993 CE spike was found 28 rings from the waney edge, implying the trees were cut down in 1021 CE. Here, the likelihood of driftwood can almost be completely discounted, given the objectives of the Norse, and the diminutive probability of obtaining three samples with the same final growth year. However, 1021 CE does not necessarily represent the first or last year of Norse presence, only one specific year in which they were active on the continent. It is expected this method will allow many other chronological questions to be resolved.


## S7-O-006 {.unnumbered}

**The potential of tree-ring drought atlases for dating and provenancing archaeological timbers**

Mukund Palat Rao^1,\ 2,\ 3^, Marta Domínguez-Delmás^4,\ 5,\ 6^, Aoife Daly^7^, Sjoerd van Daalen^8^, Pascale Fraiture^9^, Edward R. Cook^1,\ 2^

*^1^ University Corporation for Atmospheric Research, USA. ^2^ University of California Davis, USA. ^3^ Lamont-Doherty Earth Observatory of Columbia University, USA. ^4^ Amsterdam School for Heritage, Memory and Material Culture, Faculty of Humanities, University of Amsterdam, Amsterdam, the Netherlands. ^5^ Department of Conservation & Science, Rijksmuseum, Amsterdam, the Netherlands. ^6^ DendroResearch, Wageningen, the Netherlands. ^7^ SAXO-Institute, University of Copenhagen, Copenhagen, Denmark. ^8^ Van Daalen Dendrochronologie, Deventer, the Netherlands. ^9^ Laboratories department (Dendrochronology Lab), Royal Institute for Cultural Heritage (KIK-IRPA), Brussels, Belgium.*

<mukund@ldeo.columbia.edu> - <mukund24rao@gmail.com>

Humans have relied on wooden timber material for the construction of buildings, trade, ships, artifacts, and works of art for many millennia. Using dendrochronological methods, it is often possible to develop internally cross-dated tree-ring chronologies using this historic wooden material. However, the dynamic nature of human mobility and long-distance transport of timber can often make it difficult to determine the provenance of historic timbers. Accurate dendroprovenancing can be challenging due to matches with multiple reference chronologies, which are typically also made out historic material of uncertain provenance. Here we will showcase the potential of the tree-ring drought atlas network as a novel tool to determine the provenance historical timbers. Case studies will include 16th century Norwegian oak (*Quercus* sp.) timbers, which were massively exported into the northern Netherlands for construction purposes; 16th century Swedish oak used at the Evangelistas altarpiece in Seville cathedral (Spain); and 16th and early 17th century Baltic oak from different geographical sources, used in art works in the northern and southern Netherlands. Our results suggests that tree-ring drought atlases can be a powerful tool for dendroarchaeology to determine the source area of wooden material, particularly when it is possible to build a robust chronology using multiple historical timbers.


## S7-O-007 {.unnumbered}

**Multi-proxy provenance studies of carbonized wood: elemental and isotopic signatures**

Anna Stulcova^1^, Stéphane Ponton^2^, Anne Poszwa^3^, Jean-Luc Dupouey^2^, Julien Bouchez^4^, Frédéric Delarue^5^ , Sylvie Coubray^6^, Michel Lemoine^1^, Christophe Rose^2^, Julien Ruelle^2^, Maximilien Beuret^3^, Thanh Thuy Nguyen Tu^5^, Alexa Dufraisse^1^

*^1^ AASPE UMR 7209 (CNRS-MNHN), Muséum national d'Histoire naturelle, Paris, France. ^2^ SILVA UMR 1434 (INRAE) Université de Lorraine, AgroParisTech, INRAE, Nancy, France. ^3^ Université de Lorraine, CNRS, LIEC, Nancy, France. ^4^ Université de Paris, Institut de physique du globe de Paris, CNRS, Paris, France. ^5^ Sorbonne Université, METIS UMR 7619 (CNRS, EPHE, PSL), Paris, France. ^6^ INRAP, Centre -- Île-de-France, Pantin, France.*

<anna.stulcova@mnhn.fr>

Wood chemical composition is expected to reflect the availability of nutrients in the soil in which trees grow. Combined with isotopic tracers like ^87^Sr/^86^Sr ratios, elemental markers potentially constitute powerful tools to assess wood provenance. Dendrochemical analysis is of interest for carbonized archaeological wood remains, circumventing some of the limitations of dendrochronology linked to tree ring loss. However, thermal degradation process might introduce significant bias in wood chemical and isotope analyses.

This experimental study focuses on the effects of carbonization temperature on oak wood elemental and isotopic signatures. Wood cores from different trees and stand locations were pyrolyzed at five temperatures up to 800 °C, their trace elemental content was measured with XRF and ICP-QMS, while their strontium and neodymium isotope composition was determined by MC-ICP-MS. The concentration of trace elements generally increases with temperature, which improves the discrimination of woods of different origins. However, the magnitude of the enrichment depends on the element, the wood component (sapwood vs. heartwood) and the geological substrate. Rubidium, strontium, manganese, magnesium, potassium and to a lesser extent calcium and phosphorus were identified as non-volatile elements. Ratios between the concentrations of these elements and ^87^Sr/^86^Sr ratios were stable over the whole temperature range, both in sapwood and heartwood, although intra-individual variance increased with temperature. Our multi-tracer approach therefore brings promising new information to determine the provenance of charred archaeological woods. Its potential will be tested for the first time on burnt timber of the Notre-Dame cathedral, as part of ongoing restoration and research projects.


## S7-O-008 {.unnumbered}

**Old Wood in a New Light -- A dendrochronological database**

Johannes Edvardsson^1^, Philip Buckland^2^, Anton Hansson^1^, Mattias Sjölander^2^, Johan von Boer^2^, Hans Linderson^1^, Björn Gunnarson^3^, Hans W Linderholm^4^, Igor Drobyshev^5^, Björn Nilsson^6^, Dan Hammarlund^1^

^*1*^ *Laboratory for Wood Anatomy and Dendrochronology, Department of Geology, Lund University, Lund, Sweden. ^2^ The Environmental Archaeology Lab and the Strategic Environmental Archaeology Database (SEAD), Umeå University, Umeå, Sweden. ^3^ Department of Physical Geography, Stockholm University, Stockholm, Sweden. ^4^ Department of Earth Sciences, University of Gothenburg, Gothenburg, Sweden. ^5^ Southern Swedish Forest Research Centre, The Swedish University of Agricultural Sciences (SLU), Alnarp, Sweden. ^6^ Department of Archaeology and Ancient History, Lund University, Lund, Sweden.*

[johannes.edvardsson\@geol.lu.se](mailto:johannes.edvardsson@geol.lu.se){.email}

The Old Wood in a New Light project focusses on digitisation and accessibility of the results of dendrochronological samples analysed and archived at the four Swedish university-based tree-ring laboratories located at Lund University, Stockholm University, University of Gothenburg, and the Swedish University of Agricultural Sciences. Collaboration with the Environmental Archaeology Laboratory and Humlab at Umeå University will enable long-term open access to data and metadata. In this project, we will (1) systematically undertake large-scale entry and open access publication of results from scientifically analysed and archived wood samples in Sweden, and their associated metadata, into the SEAD research data infrastructure (Strategic Environmental Archaeology Database; www.sead.se), and (2) actively promote the database as a resource for new and ongoing interdisciplinary research initiatives. Inclusion of dendrochronological data in the SEAD infrastructure will allow for multidisciplinary studies combining major scientific and societal questions. SEAD is already adapted to this purpose, a pilot study has been undertaken and digitisation workflows and time estimates have been confirmed for the more than 70,000 samples archived at the Lund University dendro laboratory, where the project is hosted, and its partners. Broad coverage of research networks, stakeholder interaction and strategic support from the cultural heritage community are guaranteed by the project partners and an established international and multidisciplinary reference group.


## S7-PR-001 {.unnumbered}
 
**Isotope dendrochronology in New Zealand**

Gretel Boswijk^1^, Neil Loader^2^, Giles Young^2,\ 3^, Alan Hogg^4^

*^1^ School of Environment Te Kura Taiao, University of Auckland Waipapa Taumata Rau, New Zeeland. ^2^ Department of Geography, Swansea University Prifysgol Abertawe, Wales. ^3^ Natural Resources Finland (LUKE), Helsinki, Norway. ^4^ Te Aka Mātuatua School of Science, University of Waikato Te Whare Wānanga o Waikato, New Zeeland.*

<g.boswijk@auckland.ac.nz>

Ring-width dendrochronology has enabled the development of precise chronology in Aotearoa New Zealand. Whilst much of this work has focussed on kauri (*Agathis australis*) from the upper North Island, other species such as matai (*Prumnopitys taxifolia*), miro (*P. ferruginea*), and totara (*Podocarpus totara*) are also present in the environment as living trees, as conserved artefacts (taonga), or in the archaeological record. Unfortunately, inter-species dating using existing kauri chronologies has so far proved unsuccessful, so the development of well-replicated tree-ring chronologies for these species, which are also long-lived and have national coverage would represent an important step in understanding the socio-cultural and environmental histories of these islands. Despite the optimism of early researchers, recent analysis of ring-width variability in matai and miro has shown these species to be very challenging for ring-width dendrochronology. Analysis of the chemical (stable isotope) composition of tree-rings may however provide an alternative approach to support chronology building and precision dating for these species. This paper presents initial proof-of-concept for the further investigation and development of isotope dendrochronology in New Zealand.


## S7-PR-002 {.unnumbered}

**Solar bursts recorded in tree rings**

Fusa Miyake^1^

*^1^ Institute for Space-Earth Environmental Research, Nagoya University, Nagoya, Japan.*

<fmiyake@isee.nagoya-u.ac.jp>

Tree rings record past extreme solar bursts as radiocarbon spikes. Thus far, several radiocarbon spikes have been reported, such as the 774 CE, 993 CE, and \~660 BCE events. Such signatures of extreme solar bursts are not only important for solar physics research, but also useful for accurate dating. In this presentation, I will introduce the detected radiocarbon spikes and a further exploration efforts of past radiocarbon spikes.


## S7-PR-003 {.unnumbered}

**Genetic analysis of „The Bed of Roses"**

Hilke Schroeder^1^, Lasse Schindler^1^

*^1^ Thuenen Institute of Forest Genetics, Grosshansdorf, Germany.*

<hilke.schroeder@thuenen.de>

A restorer of old furniture asked the Thuenen Institute of Forest Genetics to help him with a genetic analysis of a bed declared as Victorian from the 19. Century but to his opinion clearly older. A first dendrochronological analysis resulted in the 18. Century and best accordance to American oaks. With genetic analysis no age determination is possible but species and origin can be identified. We got samples from the bedposts of the old bed and performed genetic analysis. We found that the bed was made of an European oak species and also the analysis of the origin resulted in Central Europe with a possibility of Britain.

With more expert reports of other scientists (as radiocarbon analysis, another dendrochronological analysis, detailed research of the used paints and some more) it became more and more clear that the restorer maybe bought the „Bed of Roses" -- the bed in which Henry VII and Elizabeth of York put an end to the war of roses.

Using this example, we will explain the use of genetic methods to identify species and origin of wood samples for a broad range of applicabilities. And, of course, we will tell something more about the „Bed of Roses".


## S7-P-001 {.unnumbered}

**The Rosewoods, where fine arts and nature conservation meet: the importance of identifying wood to the species level and the methods to do that**

Kévin Liévens^1^, Victor Deklerck^2^, Mélissa Rousseau^1^, Nils Bourland^1^, Hans Beeckman^1^

*^1^ Service of Wood Biology, Royal Museum for Central Africa (RMCA), Tervuren, Belgium. ^2^ Royal Botanic Gardens Kew, Richmond, Surrey, United Kingdom.*

<kevin.lievens@africamuseum.be>

Rosewood is a commercial term encompassing hardwood species with a pink to brown heartwood, from different botanical taxa from tropical America, Africa and Asia. Rosewoods have often been used for style furniture, interior joinery and sculptures. Harvests historically originated from Asia, Central America and South America. Today, rosewood species continue to be intensively sought after for furniture, especially in Asia, resulting in high pressure on the remaining populations in continental Africa and Madagascar. Especially wood from the Dalbergia and Pterocarpus genus is of particular high concern for forest and species conservation. There is evidence that taxa from the latter genus are now being harvested mostly unlawfully and at a rate that could reach economic and biological extinctions in some range States. Consequently, many Dalbergia and some Pterocarpus species are listed in CITES Appendixes, making measures of conservation and enforcement of trade regulations dramatically needed. Notably because of CITES requirements, those measures ask for identification of the timber to the species level. This is not always feasible with classical wood anatomical assessments. A approach has been proposed based on DART-FOMS analysis of the metabolites in the heartwood after a wood anatomical screening. We discuss this approach with four case-studies of wood identification of historical furniture: an armchair design by Josef Hoffmann in 1901, a virginal built by Henri Van Casteel in 1770, a Chinese table Huanghuali and a batch of rosewood guitars recently confiscated by the Belgian customs.


## S7-P-002 {.unnumbered}

**Reconstruction of forest development in the medieval ore mountains using wood density of mining timber**

Svenja Ahlgrimm^1^, Tobias Scharnweber^1^

*^1^ Institute of Botany and Landscape Ecology, University of Greifswald, Greifswald, Germany.*

<Svenja.Ahlgrimm@uni-greifswald.de>

The Ore Mountains are one of the most important medieval mining areas in Europe. During excavations of 12th and 13th century mining complexes, thousands of construction timbers (Abies alba M., Picea abies L.) were collected and partly dendrochronologically dated. Permanent settlement in the mountain complex only started during this period. It can therefore be assumed that the forests used for mining timber are representative for a natural mountain forest not affected by human interventions before.

To improve our knowledge about the development of the medieval Saxon primeval forest we selected a subset of these historical wood samples. We use tree-ring width (TRW) supported by maximum latewood density (MXD) for dendroprovenancing and to investigate disturbance events as well as climate-growth relationships.

During sample preparation we noticed that timber quality was strongly affected by post-sedimentary deposits in the mines, resulting in high iron and manganese contents in the wood. First measurements revealed that these elements can bias the density parameters, especially medium (decadal) time frequency signals. To reduce this bias, we have developed a protocol for treating the samples in advance, removing the metal concretions prior to density measurements.

Despite these methodological difficulties, we assume that the use of MXD parameter increases the reliability of the data set. Since MXD is strongly dependent on summer temperatures, it is likely to provide a more robust signal than TRW and thus supports cross-dating of undated series. Considering the climatic differences along the altitudinal gradient, the use of MXD can also improve the accuracy of dendroprovenancing.