How quantifying wood charcoal

Identifications may be scored in two ways: by ubiquity (qualitative data) and by fragments counts (quantitative data).

Ubiquity score

For a qualitative analysis, the analyst counts the number of samples in an archaeological level in which each identified taxon appears. A minimum of 10 samples per level is necessary for such an analysis, but a smaller number can provide useful information (although the interpretations in this case should be discussed carefully). Only the taxa present in at least two samples of a given level are significant in term of interpretation. Use ubiquity score mainly when you intend to deal with vegetation history.

Table illustrating ubiquity scoring. Each taxon determined is recorded by a "1" if present or by a blank if absent.The last column shows the sum of presence.

Fragment counts score

For a quantitative analysis, a fragment count is the fastest way to proceed. The number of fragments identified for each taxon is reported in a table, and then percentages are calculated, excluding unidentified individuals (meaning fragments showing not enough recognisable criteria to be identified). A number of 50 fragments identified per sample is a minimum, but 100 fragments analysed for each sample is recommended (50 fragments from the 4 mm fraction and 50 fragments from the 2 mm fraction). A good range of samples to proceed is 5 to 10 samples. Considering that qualitative and quantitative analysis improve to be combined (see Data Analysis section), analysing a minimum of 10 samples per level is recommended (as far as possible, of course).

Table illustrating fragment counts scoring. The number of fragments for each taxon determined is recorded and then summed.The last column shows the relative frequency of each taxon for the whole period studied.

Why exclude unidentified fragments of quantitative analysis?

Unidentified fragments are excluded of quantitative analysis, since they don't provide any kind of information (ecological or economical). Unidentified fragments have to be taken into account when it is about assessing the quality of the material study, otherwise they don't have much meaning in terms of interpretation.