<body>

Overview

Monday, August 12, 2013
The Flora of North America North of Mexico is being created by over 800 authors in an online version and in 30 printed volumes; this is usually called just Flora of North America (FNA). The FNA contains dichotomous identification keys, which allow biologists and serious amateurs to identify the known plants in North America.

This paper analyzes the structure of the FNA keys, pointing out patterns that help visualize and create diagrams of the keys, especially in cases where the the key structure is complex. The patterns documented and the diagramming methods generally apply to all dichotomous keys.

While analyzing the online FNA keys, consistency checks were created to assure that there are no structural errors in the FNA keys. The number of errors found were generally small, but the number of keys is large, so elimination of these errors is essential for algorithmic processing of the keys. The result was a database containing the FNA keys with the structural errors corrected. This database does not change the intended content of the keys; that is, the wording of couplets and resulting taxon names are changed very little, but there were many changes because of  missing or incorrect connections between the couplets and the taxa.  The online FNA was created before the printed version, and many of the errors in the online version were corrected in the printed version.

Taxa in the FNA keys don't occur uniformly across the 65 locations (see Taxon Locations), so keying could be simplified if there were a key for each location (or set of locations). An advantage of using the FNA key database is that the FNA keys can be enhanced with localization information. How this is done and how localized keys can be created is described in the section on Localized Keys.

The 2nd Edition of the Flora of the Pacific Northwest (FPNW) is to be published in 2018. Keys in FPNW have structures similar to FNA, but also have some have additional structural features; these are mostly described in the section Effect of Artificial Keys.

Data created by my analysis are too large to be contained in this blog, so this blog has links to Google Docs spreadsheets. A Google account or a gmail account is required to access these spreadsheets; if you don't have an account, you can go here to create one.


Contents

      Overview
      Couplets, Targets and Rows
      Characters Sets and Paths
      Multiple-Characters-Set Groups
      Subkeys and Concatenated Keys
      Key Structure
      Effect of Artificial Keys
      Taxon List and Counts by Type
      Node Structure and Types
      Node Groups and Relationships
      Taxon Locations
      Localized Keys and Predetermined Characteristics

The first four sections below present concepts needed to describe the basic Key Structure. The Effect of Artificial Keys section extends that. Most of the keys follow Taxonomic Rank, where classification of an unknown species proceeds sequentially from family to genus, then to species and possibly subspecies; that is, they mostly have a synoptic key structure. However, some of the more complex keys mix in an artificial key structure, which makes identification more convenient and reliable (this is sometimes called a diagnostic key structure). The Effect of Artificial Keys section has several examples of how the taxonomic hierarchy can be incorporated into visualizing the key structure of these keys.

0 Comments:

Post a Comment

<< Home