Orthodontiaceae Goffinet

The morphological differences between the N.Z. genera are substantial and preclude the drafting of a meaningful family description. This strongly suggests that undue emphasis has been placed by Goffinet and others on molecular information to the exclusion of morphological information.

The status of the Orthodontiaceae and the relationships of its core genus Orthodontium are confused. Orthodontium has traditionally (Brotherus 1924) been placed in the Bryaceae within a small subfamily (Orthodontioideae), characterised by an absent or very low endostomal membrane and filiform endostomal segments. Sainsbury (1955) accepted this placement, albeit without reference to the subfamily, as did Scott & Stone (1976). The elevation of the subfamily Orthodontioideae to family level as the Orthodontiaceae was proposed by Goffinet (in Buck & Goffinet 2000). Goffinet placed two genera within his new family: Orthodontium and the very small Asian Orthodontopsis. Shaw (2012) treated the family for Australia using Goffinet’s 2000 circumscription and recognised only the single genus Orthodontium.

Bell et al. (2007) published an influential molecular study of pleurocarp relationships using mitochondrial and chloroplast DNA. Among their conclusions was that Orthodontium is closely allied with Hymenodon, Leptotheca (which both occur in N.Z.), and the more recently described Asian Orthodontopsis. By way of context, Hymenodon was placed in the Rhizogoniaceae by both Brotherus (1924) and Sainsbury (1955), and Leptotheca was placed in the Aulacomniaceae by those same two authors. Their family placements for Hymenodon and Leptotheca were followed by Scott & Stone (1976) and by Churchill & Buck (1982).

Goffinet et al. (2009) reinterpreted the systematic significance of the Bell et al. (2007) study. Reflecting the fluidity of modern moss family (and higher rank) circumscription, Goffinet et al. (2009) proposed the transfer of both Hymenodon and Leptotheca (as well as Orthodontopsis) to the Orthodontiaceae. They placed this newly circumscribed family in the Rhizogoniales in the general relationship of the Rhizogoniaceae and Aulacomniaceae.

The family definition of Goffinet et al. (2009) is followed here less from conviction than from convenience. Their definition emphasises molecular data to a far greater degree than morphological characters. For example, Orthodontium has linear laminal cells, while those of Leptotheca and Hymenodon are isodiametric, and the three genera have different costal anatomy. (Orthodontium has a costa with a central stereid band; Leptotheca has median guide cells and both abaxial and adaxial stereids; and Hymenodon has a few, often two, guide cells and abaxial stereids.) Orthodontium is variably autoicous, with antheridia occurring in the axils of outer perichaetial leaves (paroicous), or on short perigonial buds located either below the perichaetium or at the base of fruiting shoots (gonioautoicous) or occasionally on well-developed male shoots. Leptotheca is dioicous with terminal perichaetia, while Hymenodon is rhizautoicous with perigonia bud-like and nidulant among basal rhizoids or vegetative leaf axils. Both Orthodontium and Leptotheca have striate capsules with double peristomes (albeit with reduced endostomal membranes), while Hymenodon has smooth capsules with a single exostomal peristome.

The morphological differences between the N.Z. genera are substantial and preclude the drafting of a meaningful family description. This strongly suggests that undue emphasis has been placed by Goffinet and others on molecular information to the exclusion of morphological information.