Trends and concepts in fern classification (Invited Review)

Fern classification has been highly unstable in the past, largely because the morphological characters selected for emphasis in higher-level classifications have been variable. Christenhusz and Chase review the history of fern classification, and consider that some classifications based on single characteristics have been useful for identifying species but are highly artificial in terms of their views on relationships, whilst classifications using a suite of characters have been considered more ‘natural’, but with the advance of molecular phylogenetics many of these ‘natural’ groups are also been shown to be highly artificial. They highlight which groups/genera have taxonomic issues that remain to be clarified, and propose the use of broader family concepts. Progress in developing a consensus fern classification is reflected in a new proposed classification.