Fabronia pusilla

Fabronia pusilla View in CoL , F. rostrata , F. matsumurae

The distinction between F. pusilla and F. ciliaris is also sometimes questioned. At least some problematic specimens were mentioned by, e.g., Crum & Anderson (1981). Specimens of F. pusilla from westernmost localities are easily recognized due to large, multicellular teeth at leaf margins. However, at the eastern edge of its distribution range, where F. pusilla co-occurs with F. ciliaris , plants with smaller marginal teeth consisting not more than three cells are often collected. Contrary to F. major and F. altaica , F. pusilla usually has ovate leaves abruptly narrowed into acumen, short laminal cells and short, ovate capsules, which is similar to F. ciliaris . In our dataset, only one specimen of F. pusilla in question with small marginal teeth was included ( F. pusilla 2, Italy). It appeared in F. pusilla -clade in ITS-based tree ( Fig. 1 View Fig ), belonging the the main haplotype in the haplotype network ( Fig. 2 View Fig ). On the other hand, specimens with occasional 2–3-celled marginal teeth were observed far beyond the distributional range of F. pusilla , e.g., in South Urals and Khabarovsk Territory. They were resolved within F. ciliaris ( F. ciliaris 20 and 15 in Figs. 1–3 View Fig View Fig View Fig ).

We were able to confirm only one specimen of F. pusilla from Siberia: it is a specimen from Ol’khon Island of Baikal Lake. It is old and could not be sequenced; however, it undoubtfully belongs to this species due to leaves with very large, multicellular marginal teeth ( Fig. 10 View Fig : 4). Closest localities of F. pusilla are known in Middle Asia ( Turkmenistan, Tajikistan and Kyrgyzstan); we confirm these identifications in LE. Fabronia pusilla was also reported from SW China, Xizang and Yunnan ( Gao & Fu, 2002b); however, illustrations in Gao & Fu (2002a: page 96, Fig. 43) do not show the main character of F. pusilla , namely multicellular teeth at leaf margins.

Fabronia rostrata is a rare species, only recently found in Russia. The only sequenced specimen showed very subtile difference from F. ciliaris in ITS and none in more variable IGS1. However, a unique combination of such characters as eperistomate capsule and operculum with long, oblique beak easily separates this species from F. ciliaris (it is worth to note, however, that one specimen of F. ciliaris from the Caucasus (Kabardino-Balkaria 1) also had numerous capsules with long, oblique beak, but with perfectly developed peristome). Further studies are needed to understand such discrepancy better.

Another specimen from the Russian Far East was once identified as F. matsumurae due to the absence of normally developed peristome. It appeared to be identical to F. ciliaris in both ITS and IGS1 sequences. Some capsules in this collection have rudimentary or practically absent peristome, while in other ones pale teeth ca. 50 µm long, with papillose-striolate ornamentation were observed. This feature is unknown in F. ciliaris , but for F. matsumurae only totally eperistomate capsules were reported. In other characters, i.e., leaf shape, marginal teeth, cell areolation, capsule and operculum shape and spore size these plants fit F. ciliaris . We consider peristomes in this collection as abnormally developed and refer this specimen to the latter species.

NrIGS1 vs. nrITS

The IGS1 is not a very widely used marker for phylogenetic studies as it is often too variable so precluding proper aligning, variable in length ( Mateos & Markow, 2005), and has big reversions ( Wicke et al., 2011).

However, in some rapidly evolving groups IGS provides a strong and clear phylogenetic signal, which is well congruent with that from ITS (Logacheva et al., 2011). In moss species of the genus Schistidium , it possessed a conservative and species-specific substitutions ( Milyutina et al., 2015). The present study indicates a great value of IGS1 for supporting ITS-based conclusions of the species delimitation.