Measurements of cnidae from sea anemones (Cnidaria: Actiniaria): statistical parameters and taxonomic relevance
R.B. Williams

A statistical study of actiniarian cnida measurements has been carried out, using Urticina eques (Gosse), Nematostella vectensis Stephenson, Haliplanella lineata (Verrill), Sagartia elegans (Dalyell) and Sagartiogeton laceratus (Dalyell) as examples. Standard methods for obtaining samples of the same kinds of tissue from different anemones, and for measuring cnidae in tissue samples for subsequent data analysis are described. The method of examination of tissue samples ensured that cnida measurements were random and that they comprised a continuous frequency distribution. The distributions of cnida lengths were shown here for the first time to be normal (Gaussian), with no within-sample variation (U. eques, N. vectensis, H. lineata, S. elegans). Hence, parametric inferential statistical tests could be performed validly on the measurements obtained. Reliable estimates of means and ranges of sizes were provided by measurements of ≥40 of each cnida type in a tissue sample. Statistically significant differences occurred between mean lengths of cnidae in replicate samples from the same specimen (U. eques, S. lacerutus), and between samples from different specimens of the same species (S. lacerutus). There was no predictable relationship between the size of an anemone and the mean lengths of its cnidae, at least for the one asexually produced species examined (S. laceratus). The most convenient and reliable measure of an anemone's size was its pedal disk diameter (PDD). Based upon these methods and results, an outline protocol for obtaining actiniarian cnida measurements and for their statistical treatment was developed. It should include descriptions of at least: 1 ) specimens examined (species; numbers; localities whence collected; sizes; whether live or, if not, how preserved); 2) tissue samples taken (identity; position on specimen whence taken; how obtained); 3) examination of samples (preparation; microscopical method; magnification; accuracy of measurements; randomness of measurements); 4) statistical treatment of cnida measurements (descriptive statistics - at least the mean size, number, standard deviation, largest and smallest; type of frequency distribution and, if necessary, transformation of data; inferential statistics - parametric or non-parametric significance tests). It was concluded that statistically significant differences between mean lengths of particular cnida types from different samples of the same kind of tissue cannot be used alone to distinguish actiniarian species. This does not preclude other uses of statistical analysis of cnida sizes for anatomical or physiological studies.

Keywords: Actiniaria, cnidae, measurements, protocol, statistics, taxonomy.
Contents of this volume Sci. Mar. 60(2-3) : 339-351 Back PDF
Please, send comments to
ICM (CSIC) - Passeig Marítim de la Barceloneta, 37-49. E-08003 Barcelona (Spain)