This is a family of five genera,
viz., Aulacopilum Wils., Erpodium (Brid.) Brid., Solmsiella
C. Müll., Venturiella C. Müll, and Wildia (C. Müll.) Broth. A
sixth genus, Microtheciella Dix., was segregated into the family
Microtheciellaceae (Miller & Harrington 1977). Solmsiella is often
combined with Erpodium (Crum 1972, Touw 1992, Vital 1980). Stone (1997)
recently combined Aulacopilum, Wildia, and Venturiella with
Erpodium, thus reducing the family to a single variable genus. Only Erpodium
and Solmsiella are present in Central America.
The Erpodiaceae are usually placed
in the diplolepidous-series of mosses near the Orthotrichaceae. Most
Erpodiaceae are eperistomate, but E. coronatum (Hook. & Wils.) Mitt.
has a peristome reduced to an irregular membrane, and Venturiella sinensis
(Vent. ex Rabh.) C. Müll. and Wildia solmsiellacea C. Müll. &
Broth. have 16 narrowly triangular peristome teeth. Noguchi (1952) and Edwards
(1979) have shown that the peristome in Venturiella is endostomial in
position with somewhat stronger deposition on the outer (dorsal) surface than
the inner (ventral) surface, and that its peristomial formula is 4:2:2. Edwards
(1979) thought this peristome could be haplolepidous. On the basis of such a
peristomial interpretation as well as rbcL-gene sequences Goffinet et
al. (1998) placed the Erpodiaceae in the Haplolepideae, La Farge et al.
(2000) put the taxon in the Pottiales, and Buck & Goffinet (2000)
transferred the family to the Dicranales.
As noted by Edwards (1979),
however, some members of the Orthotrichales have the same peristomial formula
and type of secondary thickening as the Venturiella peristome.
Furthermore, the E. coronatum peristome-type, a reduced irregular
membrane, is also seen in the Bryaceae (e.g., Mielichhoferia or Brachymenium).
This type of peristome reduction, although probably analogous in the
Erpodiaceae and Bryaceae, has no counterpart in the Haplolepideae and so argues
for placing the Erpodiaceae in the Diplolepideae. In our view the confidence
levels of the rbcL gene analyses (e.g., LaFarge et al.
2000 assign Bootstrap values to their analyses of 51, 52, <50) are so low
that these data are inconclusive. Finally, the presence of rhizoidal initials
clustered at the abaxial leaf insertion (see Noguchi 1952, fig. 2 and Stone
1982, fig. 1 a–b) is a characteristic feature of Diplolepideous mosses (see
Hedenäs 1987).