The Moss Families of the British Isles


L. Watson and M. J. Dallwitz


This data set is generated from a DELTA database (Dallwitz 1980; Dallwitz, Paine, and Zurcher 1993), and comprises draft descriptions of the families, illustrated from Berkeley (1863) and Dixon and Jameson (1924), with a file of source references and hints on using the program INTKEY for interactive identification and information retrieval. The generic nomenclature and family interpretations mostly conform with Smith (2004), with a few minor adjustments; cf. Vitt (1984), Buck and Goffinet (2000), and Ochyra and O’Shea (2004).

The descriptive data on families were originally compiled from the family, generic, and species descriptions of Smith (1978), Dixon and Jameson (1924) and Watson (1978), cross referenced with their keys; and they have subsequently been extensively modified via Smith’s 2004 edition. The package incorporates no new observations on mosses, and no original taxonomic research has been brought to bear in preparing it, but considerable effort has been invested in trying to render the family descriptions consistent with those of their constituent genera and species. This proved an unsatisfying experience, given the extensively non-comparative nature of the available descriptions at all levels (a problem not confined to moss taxonomy, cf. Watson 1971), and the certainty that comparative DNA studies will soon result in continuing, major changes to family circumscriptions; but the effort was necessary, in view of numerous discrepancies between the family descriptions currently being purveyed and those of the taxa they are supposed to encompass. Smith’s 2004 treatment of the British representatives seems to fairly represent current world views on the classification of mosses, but the latter evidently leaves much to be desired at family level and above; as indeed he freely acknowledges, remarking with some exasperation, for example, that “DNA studies are unlikely to help in identification (p. 810; see aso his complaint about a modern re-classification of Pottiaceae, in which the author discusses subfamilies and tribes in terms of “clades and ancestral nodes”, without providing adequate group descriptions). Hedderson et al.’s (2004) conclusions from comparative chloroplast DNA studies that while the Dicranidae (= Dixon’s Aplolepidae, the haplolepideous mosses) seem to be monophyletic, modern families and orders within that huge assemblage are “poorly supported”, need therefore come as no surprise.

While it has been apparent for about forty years that comparative DNA studies would ultimately provide conclusive evidence for understanding phylogeny, classifications derived via DNA sequencing will never be definitive unless the taxonomic coverage is adequate, and the DNA sampled can be confidently assumed represent complete genomes. Many DNA-based classifications fail to meet even the first criterion, and none seem to satisfy the second; indeed, plant phylogenies deduced entirely from chloroplast DNA do not directly represent nuclear genomes at all. It should be unnecessary to point out that purported phylogenies, whether based on DNA or not, are impossible to evaluate as such in the absence of external evidence from character correlations; and that even if a classification truly reflects evolutionary relationships, it is useless if it does not bestow the ability to generalize about the contents of groups in terms of their structure, physiology, ecology, cytology, etc. cytology, etc. For too long, people have been peddling revised classifications in the absence, fully operational, new or properly revised group descriptions. Editors of taxonomic journals should long ago have been automatically rejecting classificatory papers that fail to meet the obvious requirements of both good science and of practical applications of taxonomic systems, regardless of the kinds of data analyzed.

Most of the smaller moss families currently accepted seem homogeneous in terms of the data assembled here, and are probably sound taxonomic entities (though not necessarily justifiable at family level); but while all the large families include suites of closely related genera, too many seem unsatisfactorily circumscribed, and several (e.g., Amblystegiaceae v. Brachytheciaceae, Pottiaceae versus Dicranaceae, Ditrichaceae and Rhabdoweisiaceae) are incompletely separable. In clear recognition of poor family circumscriptions, users of moss floras have long been offered keys direct to the genera rather than progressing via the families, and the present compilation cannot be expected to prove very satisfactory for pursuing identifications to family level. On the other hand, it does afford the opportunity to make Berkeley’s pleasant illustrations and H.G. Jameson’s fine interpretations of Dixon’s meticulous drawings available to a wider audience, with nomenclaturally updated legends; and to present a preliminary character list, which with expert input could readily be improved and increased in scope, with a view to extending the interactive option to the levels of genus and species.