WOOD CHARCOAL : IDENTIFICATION
Using a reflected light microscope
Identification of wood charcoal is made by using a reflected light microscope with bright and dark field (BF/DF) and a magnification range from 100x to 500x, the maximum range of such a microscope being generally 50x to 1000x. Use of a microscope with such magnification is essential since certain key features for wood charcoal determination can only be seen at a high (100x or 200x) or very high (500x or 1000x) one.
However, using a low-powered microscope is useful for a pre-classification of the types of wood present in a sample. Based on some of the criteria observed in transverse section, wood charcoal can be pre-classified depending on whether they are homoxylous (Gymnosperms) or heteroxylous (Angiosperms), then (for heteroxylous wood) regarding the porosity and the distribution of the vessels. This pre-classification can be made either on the field (if a low-powered microscope with a 40x magnification is available) or in the laboratory, simultaneously with the sorting process.
Identification of wood charcoal is supported by use of reference material: wood anatomy atlases and a reference collection of modern charred wood. Wood anatomy atlases are catalogues of photographs of thin sections of modern fresh wood, generally from a particular geographical area, here covering Europe, Northern Africa and Middle East. Wood anatomy atlases also include identification keys.
Wood anatomy atlases
Here is a list of wood anatomy atlases, followed by some websites where some identification material is also available. Wood atlases present in the laboratory of the University of Sheffield are in bold-enhanced type.
Greguss, P. (1959): Holzanatomie der europaïschen Laubhölser und Strächer.
Fahn, A., Werker, E., Baas, P. (1986): Wood Anatomy and Identification of Trees and Shrubs of Israel and Adjacent Regions.
Parsapajouh, D., Schweingruber, F.H. (1988): Atlas des bois du nord de l'Iran.
Schweingruber, F.H. (1990): Anatomy of European Woods.
Gale, R., Cutler, D. (2000): Plants in Archaeology.
Hather, J. (2000): The identification of northern European woods.
Neumann, K., Schoch, W., Détienne, P., Schweingruber, F.H., Richter, H. (2000): Woods of the Sahara and the Sahel.
Benkova, V., Schweingruber, F.H. (2004): Russian Wood Anatomy.
http://pcwww.liv.ac.uk/~easouti/Index.html: A website compliling the reference collection of Cecilia A. Western.
http://www.wsl.ch/land/products/dendro/: A website where an identification key for wood of Central Europe is available.
http://www.holzanatomie.at/: An electronic atlas made by Andreas Heiss can be downloaded here. Comprehensive reference about wood of Northern America and Europe, but crucial lack of photographs of wood (not ideal for students).
There are two types of identification keys: the first is organised according to the main criteria of identifications (type of wood, porosity, width of rays…) as in Anatomy of European Wood (Schweingruber 1990); the second is organised according to an exhaustive list of features observed under microscope as in Anatomy of Trees and Shrubs of Israel and Adjacent Regions (Fahn et al. 1986). A reference collection of modern charred wood is however more important than wood anatomy atlases. A reference collection consists in a set of modern fresh wood which has been carbonised. This provides a direct comparison for the archaeological material. Such a reference material is built up by gathering pieces of wood of different diameters for each ligneous species (perhaps from a particular geographic region). Such a reference collection is now available in the laboratory of the University of Sheffield. List of species gathered for this collection can be found here.
Microscopic work on wood charcoal
Wood identification process
Wood charcoal are broken by hand or by using a razor blade to expose the three different sections of wood: transverse (perpendicular to the vertical axis of wood), longitudinal tangential (perpendicular to the rays) and longitudinal radial (parallel to the rays). Identification of wood charcoal is based on recognition of different cells and their arrangement. There are two types of wood: homoxylous and heteroxylous. An homoxylous wood is a wood without fibres, only tracheids, and parenchyma; it appears homogeneous in transverse section, and corresponds to wood of Gymnosperms (conifers). An heteroxylous wood is a wood with fibres, vessels and parenchyma; it appears heterogeneous in transverse section, and corresponds to wood of Angiosperms (leaved trees). Nomenclature used for descriptions follows the one established by the International Association of Wood Anatomists (IAWA).
Sections should be examined in the following order:
- examining the transverse section (TS),
- examining the longitudinal tangential section (TLS),
- examining the longitudinal radial section (RLS).
Our own experience shows that following this order is the best way to proceed. Transverse section allows generally narrowing strongly the range of possible species, so this is the most important section to begin with. Then, tangential longitudinal section helps to narrow the identification further more and, sometimes, examining radial longitudinal section is just a way to confirm an identification reached through the previous two sections.
Anatomical features used during the identification process are viewable here. A small identification key designed for use with the reference collection of the laboratory in Sheffield is also provided. However, the level of identification achieved in anthracology rarelly goes to the species, but most likely to the genus of the plant, and sometimes only to the family or the sub-family. This is because most of the species of a same genus or sub-family are anatomicaly very close and that the criteria allowing to distinguish them are not always present (or simply nonexistent). Below is the nomenclature used for the different levels of identification achieved.
Level of identification achieved
Here is an illustration of the nomenclature used in anthracology, depending on the level achieved during the identification process. This example is based on Crataegus azarolus (azarole hawthorn), a member of the Rosaceae family (sub-family of the Pomoideae).
- Family: the taxon was identified at the family level.
- Sub-family: the taxon was identified at the sub-family level.
- Genus 1/Genus 2: hesitation between two genera.
- cf. Genus: the taxon was most likely identified at the genus level, however other genera are also possible.
ex.: cf. Crataegus
- Genus sp.: the taxon was identified at the genus level.
ex.: Crataegus sp.
- Genus species 1/species 2: the taxons was identified at the genus level, hesitation between two species.
ex.: Crataegus azarolus/monogyna
- Genus cf. G. species: the taxon was identified at the genus level, a species is highly supposed.
ex.: Crataegus cf. C. azarolus
- Genus species: the taxon was identified at the species level.
ex.: Crataegus azarolus
Dendro-anthracology provides complementary information upon wood charcoal identification. It involves the observation of different features in the wood (Marguerie and Hunot 2007). The most interesting feature of this is growth-ring curvature (weak, moderate or strong). Its evaluation can provide information about the size of the piece of wood that was burnt:
- fragments showing a weak curvature represent most likely big pieces of wood (trunk, big branch),
- fragments showing a moderate curvature represent most likely medium pieces of wood (medium branch),
- fragments showing a strong curvature represent most likely small pieces of wood (small branch, twig).
Complementary information about the size of a piece of burnt wood is provided by the presence (or absence) of the following features: bark, pith, reaction wood and tyloses. Presence of bark implies presence of the last ring of growth in the fragment. In association with a strong or weak curvature, this reinforces the category of size at which the fragment was assigned. Presence of pith (or heartwood) only is not very useful. It is with the additional presence of bark. In that case the calibre of the twig or stem can be measured (but the reduction of size of the wood during carbonization must be taken into account). Reaction wood occurs in some branches and leaning trunks. If associated with strong curvatures, it confirms that the piece of wood was originally from a branch. If associated with a weak curvature, it could mean that the tree was growing either under the wind or on mountain flanks. Tyloses are cells of parenchyma growing out in the vessels and filling them. Their presence only occur in vessels near the heartwood. In association with a weakly-curved fragment, this confirms that the piece burnt was rather big, so that it was a big branch or a trunk that were used.
Finally, information about the state of wood before carbonization is provided by the presence (or absence) of the following features: fungal hyphae, insect degradation, radial cracks and vitrification features. Fungal hyphae can penetrate in dead or dying wood from fungus living on the surface. Their presence indicates that wood burnt was dead wood. The presence of insect degradation channels could provide the same information, as long as these channels are left by dead wood borrowing insects. Radial cracks can occur in damp wood, but their occurence depends also of the temperature and of the density of the wood. Vitrification features can be cause by various factors, such as a rapid combustion at high temperatures, a high degree of humidity..., but all the cause of their occurence in the wood are not known yet and there are studies in progress for a better definition of this phenomenon.
Recording the information acquired
During the identification process, the information concerning each wood charcoal observed is recorded on a specially designed score-sheet. This includes the identification (or the fact that the piece can't be identified) and the different dendro-anthracological features.
The different categories that must be filled in are the sample code (to be able to find quickly a previously identified taxon), the fragment number (which fragment is identified first, which one second, and so on...).
It is essential to record each charcoal observed in the right order, so building a saturation curve (below) could be possible. For each wood charcoal observed, the following information is then recorded: the taxon identified, the growth-ring curvature (w: weak; m: moderate; s: strong; none: indetermined), presence of bark (B), presence of pith (P), presence of reaction wood (RW), presence of tyloses (T), presence of fungal hyphae (FH), presence of insect degradation (ID), degree (I, II, III) of vitrification (V) and presence of radial cracks (RC). Degrees of vitrification are as follow: low brilliance (I); strong brilliance (II); dense, non recognisable masse (III).
A saturation curve is built by scoring after how many fragments a new taxon appears during the analysis. Practically, it consists in reporting the fragment number for which each taxon appears the first time in the sample studied (using this record sheet). A saturation curve allows to determine the optimal number of fragments to identify to have a good array of taxa present in one sample. The optimal number is found by reading the point where the curve levels up (see figure below).
- Fahn A., Werker, E., Baas P. (1986). Wood Anatomy and Identification of Trees and Shrubs of Israel and Adjacent Regions. Jerusalem: Israel Academy of Sciences and Humanities.
- Maguerie, D., Hunot, J.-Y. (2007). Charcoal analysis and dendrology: data from archaeological sites in north-western France. Journal of Archaeological Science 34: 1417-1433.
- Schoch, W., Heller, I., Schweingruber, F., Kienast, F. (2004). Wood anatomy of central European Species. Online version: http://www.wsl.ch/land/products/dendro/welcome.html
- Schweingruber, F. (1990). Anatomy of European Woods. Stuttgart: Haupt.