How opposite may differ from opposite: a lesson from the peristome development in the moss Discelium

Abstract

Abstract Previous morphological studies and molecular phylogenetic reconstructions resolved Discelium in the diplolepideous-opposite peristome group of mosses among the early-diverging mosses with arthrodontous peristomes. However, sporophyte development in Discelium differs from that of the other ‘diplolepideous-opposite’ families, Funariaceae and Encalyptaceae, in that the transverse sections of the peristome in the early stages of sporophyte differentiation exhibit diplolepideous-opposite, diplolepideous-alternate and haplolepideous patterns. Although the proportion of diplolepideous-opposite vs. haplolepideous patterns increases as the capsules mature, a haplolepideous peristomial formula persists in about one-third of the peristome sectors, reducing in frequency only in the lower parts of the teeth. This is the first evidence of the presence of the haplolepideous pattern in the ‘diplolepideous-opposite’ lineage; although appearing in the course of development it does not end in a really haplolepideous peristome, as its endostome and exostome elements remain opposite due to adhesion throughout their length. In contrast to Discelium, the peristome in Encalyptaceae maintains a typical diplolepideous-opposite pattern of cell divisions from the earliest stages of development, as determined by the unusually thick cells of the inner peristomial layer. The presence of the haplolepideous pattern in Discelium fills an enigmatic gap between the earliest-diverging arthrodontous lineage Diphysciaceae and the terminal lineages Dicranidae, in which the haplolepideous pattern prevails, and Bryideae, in which it appears only as a transitional stage towards the more complex structure. The diplolepideous-opposite peristome may not represent a synapomorphy for the ‘diplolepideous-opposite’ group of mosses as a whole (including Discelium), thus supporting treating Discelium in its own order.