GrassBase - The Online World Grass Flora

Notes For Users

The database attempts to emulate the traditional description's portrayal of structural diversity and diagnostic detail. Secondary objectives are to provide a discipline for testing comparative and diagnostic properties, and a framework for accumulating observations of potential interest. If used as an aid to identificatiion some ingenuity is required to circumvent subjective, variable or imperfectly known characters whose coding is often erratic at present.

It is confined to morphological characters. Anatomical and other cryptic characters are less numerous, sparsely recorded and arguably better handled at generic level.

Infraspecific taxa and hybrids (except for a few commonly treated on a par with species) are omitted.

The general principles governing the choice of characters and states are:

1. Terminology should conform to traditional usage. A few unfamiliar expressions ("companion sterile spikelets") have been used to secure generality, but are not intended as new terms.

2. Characters should be broken down to simple elements. The principal exceptions are a few syndrome characters useful for tribal separation, and mixed collections of unusual states.

3. All structural features should be retrievable.

4. All diagnostic characters should be retrievable. Except fine distinctions between critical species for which text characters are provided.

5. There should be a place for all characters, or at least suitable pegs on which to hang qualifying comment; some loosely defined facies characters fall into this category. A few of the commonest qualifiers have been given character status.

6. The wording should be such that the characters can stand alone, but otherwise key–writing is ignored.

7. Strict standardization of subjective character–states is regarded as impractical, but consistent distinctions within a genus should be attainable.

The character set is based upon phrases encountered during a wide experience of reading and writing descriptions, but inevitably it is a compromise between conflicting demands, not the least of which is the desirability of keeping it to a manageable size.

In the process of sorting out generic boundaries species have occasionally been accommodated in genera for which no combination has been made. No particular significance should be ascribed to these tentative assignments.

The descriptions are mostly taken from literature with minimal herbarium checking, many of them from a time when the character set was considerably shorter. They all need revision and we welcome your co–operation to improve their quality and enhance their usefulness.

The dataset is released, without charge, on the understanding that corrections and additions are communicated to us from time to time for incorporation in the master ITEMS file. In descriptive mode a character for notifying changes is provided for this purpose; it should be used as a flag, to be cleared when update implemented. We hope that, with your help, we can develop this into an international database, giving all agrostologists access to morphological information on a world–wide scale.