NC STATE UNIVERSITY NCSU Libraries

About pare Coding

p (present) = IAWA feature number can be present or unknown(?) or varies(v)
A search using a feature coded as "p" (present) will return a list of woods in the database that have that IAWA feature listed in their description, or coded as variable (v) or unknown(?). For example, if you search for septate fibers (65p), this will return descriptions that have 65 (septate fibers obviously present) or 65v (septate fibers sometimes present or rare and can be overlooked) or 65? (not known if the wood has septate fibers, common for fossil woods).

a (absent) = IAWA feature number can be absent or unknown(?) or varies(v)
A search using a feature coded as "a" (absent) will return a list of woods in the database that do not have that IAWA feature listed in their description, or it's coded as variable (v) or unknown(?). For example, if you search for "fusiform parenchyma absent" (90a), this will return descriptions that do not include 90 or have 90v (fusiform parenchyma sometimes present, sometimes absent, or rare and can be overlooked) or 90? (not known if the wood has fusiform parenchyma, information in the literature doesn't give information on parenchyma strands).

r (required present) = IAWA feature number must be present
The IAWA feature number must be present in the result set. NOTE: Woods in the database that have the feature coded as ? or v in the database will NOT be in the list of possible matches.

e (required absent) = IAWA feature number must be absent
The IAWA feature number must be absent in the list of possible matches. NOTE: Woods in the database that have the feature coded as ? or v in the database will NOT be in the list of possible matches.

Be careful with using "r" or "e" in your searches, particularly for fossil woods. Features that can be hard to determine in a fossil might not be mentioned in a published description and those features are followed by a "?" in the database. For example, many fossil woods have not had the vessel-ray pitting described and in the database v-rp features are coded as "?". So if you used 30r [vessel-ray parenchyma pits, at least some of them, have to be similar to the intervessel pits], then you'll lose all the descriptions with 30? or 30v.

Caution: Particularly for the quantitative features (vessel diameter, vessels / mm2, ray width), when trying to identify an unknown, if you code one of the features in a set of related features (multistate characters) present, do NOT code the adjacent feature absent. For example, there are species in which some samples clearly have exclusively uniseriate rays (feature 96) while others clearly have rays 1-2 seriate (feature 97). The database description for such species will have "96 97". So if you code 96 p (uniseriate rays present) and 97a (rays 1-3 seriate absent), this species would be eliminated.

For this type for variability it seems best to have the description include both features present (96 97), rather than code 96v 97v as if you coded 96r or 96e the species would be eliminated. [4 May 2009 - I'm checking the descriptions of the 96v 97v types now to determine how best to have the description set to avoid the r or e elimination problem.]

Caution: Wood is variable and many of the records in the database were based on only 1 or a few wood samples. Consequently, the wood you are working with, even though it might belong to a species in the database, may differ somewhat from the description in the database. Moreover, wood anatomists vary in how they interpret the wood features. Thus, in describing an unknown wood, allowing mismatches can be useful. If the 1st search of the database yields no results, then we recommend that you try the search again using mismatches.

Using 0 mismatches (the default) results in a list of woods with all of the descriptors you used in your search criteria.

Allowing 1 mismatch results in a list of woods with all of the descriptors you used and woods with all but 1 of the descriptors. Allowing 2 mismatches results in a list of woods with all of the descriptors you used and woods with all but 1 or 2 of the feature items..... etc.

When describing an unknown wood, often there are feature numbers that you are sure of their presence or absence and consider important and you would want the search results to list species that matched those particular feature numbers.

For example, if you were sure that a modern wood had simple perforation plates (IAWA feature 13) and did not have homocellular rays (IAWA feature 104), the coded description of the wood could look like this: 1p 5p 13r 104e 136p 142p