Anomodontaceae and

Family Anomodontaceae and View in CoL the position of Anomodon attenuatus and A. giraldii

The polyphyly of Anomodon found in the present molecular phylogenetic analysis may appear extreme, especially with Anomodon attenuatus and A. giraldii deeply nested in the Neckeraceae . Although Anomodon was originally segregated from Neckera ( Hooker & Taylor, 1818) , at that time the pleurocarp genera were very different from their modern circumscriptions. The familial classification of pleurocarps took shape in the mid XIX century and from the beginning papillose leaf cells were evaluated as a character of great weight: Schimper first segregated Leskeaceae ( Schimper, 1856) , and shortly after that Thuidiaceae ( Schimper, 1860a). Anomodon was placed in Leskeaceae in the former publication of Schimper, but not transferred to Thuidiaceae . This was later done by Fleischer (1923) and Brotherus (1925), and most authors of 20th century ( Podpera, 1954; Smith, 1978; Crum & Anderson, 1981) followed this placement until the publication of Buck & Vitt (1986). The latter authors crucially revised the whole pleurocarpous moss system, and among other changes placed Anomodon not only in its own family Anomodontaceae (originally segregated by Kindberg (1897) but disregarded for almost a century), but also in the order Leucodontales , whereas Thuidiaceae and Leskeaceae were referred to the Hypnales . Most subsequent floras and checklists accepted the family Anomodontaceae . Buck & Vitt (1986) included in Anomodontaceae the genera Anomodon , Haplohymenium , Herpetineuron , Lindbergia , Myurella , and Thelia , united mostly by their papillose laminal cells. Their placement in other families has already been mentioned, and the position of Herpetineuron will be discussed below.

Granzow-de la Cerda (1997) published a thorough worldwide revision of Anomodon (incl. Haplohymenium ) and Herpetineuron , applying cladistic methodology based on morphology. He submerged Haplohymenium into a section Haplohymenium of Anomodon , along with sect. Anomodon and the newly described monospecific sect. Thraustus . More importantly, the genus Anomodon was subdivided in two subgenera; the first, subgen. Anomodon , included the three sections just mentioned, while the second, subgen. Pseudanomodon Limpr. , included A. attenuatus , A. giraldii , A. longifolius and A. rostratus . Thus, Granzow-de la Cerda (1997) accepted the latter subgenus with almost the same circumscription as Limpricht (1895), adding only the East Asian A. giraldii .

Limpricht (1895) was the first author who found Anomodon heterogeneous in its current circumscription, and he subdivided it into two subgenera (although without a definite indication of rank). Subgen. Pseudanomodon (‘ Pseud-anomodon ’) included A. attenuatus , A. longifolius and A. rostratus , while A. viticulosus , A. rugelii , and A. (Haplohymenium) triste formed subgen. ‘ Euanomodon ’. The difference between these subgenera included: the presence vs. absence of a neck with an air space in the lower part of the capsules; mostly acuminate vs. broadly rounded leaf apices; and elongate and smooth vs. mostly isodiametric and papillose cells of the perichaetial leaves. Although A. attenuatus was characterized as a somewhat atypical member of the subgenus (the leaf is broadly acute), Granzow-de la Cerda (1997) selected this species as the type of Pseudanomodon . He also found other important distinctions in the branching pattern between Anomodon subg. Pseudanomodon and Anomodon s. str. (obtuseleaved species related to A. viticulosus ): only the species of Pseudanomodon have a truly “Isobryalean” branching architecture, i.e. have a clearly defined stoloniform “primary stem” giving rise to the secondary stems with much larger leaves and pinnate branching, contrary to the main stem of, e.g., A. viticulosus , albeit creeping and with smaller leaves, but having simple branches which never subdivide extensively. According to Granzow-de la Cerda (1997), this difference is clearly correlated with another character, the central strand, which appears to be absent in subgen. Pseudanomodon and also Haplohymenium ; however, the latter genus is characterized by much smaller plants with thin stems, and accordingly may lack a central strand for that reason. In addition, Granzow-de la Cerda (1997) found the position of the perichaetia in Anomodon attenuatus and A. giraldii to be confined to older parts of the plants and never present beyond the last branching point. This character never occurs in other core Anomodon species and Haplohymenium , with the exception of Haplohymenium sieboldii . However, the free branching of the latter species makes evaluation of this character rather difficult, and thus we refrain from confirming or rejecting it; moreover, this rare species never occurs with sporophytes in Russia ( Czernyadjeva & Ignatova, 2019).

These distinctions help to explain the strange position of subgen. Pseudanomodon in the Neckeraceae , namely in a supported clade with Homalia , as already mentioned by Tsubota et al. (2002) and Olsson et al. (2009a). The Neckeraceae as a whole, and Homalia in particular, have a sharp differentiation between creeping stoloniform “primary stems” and pinnately branched secondary stems. The presence of papillae in the leaf cells makes Anomodon quite an odd member of the Neckeraceae ; however. the ‘ Pinnatella -lineage” of the Neckeraceae ( Olsson et al., 2009a) includes Homaliodendron papillosum , a moss with distinctly unipapillose laminal cells. We failed to include material of it in our own study, but the nad5 sequence from GenBank points towards a considerable similarity with Anomodon attenuatus and A. giraldii (cf. Fig. 5 View Fig ).

Summing up, there is no alternative to the placement of two latter species in a separate genus within the Neckeraceae , and raising the subgenus Pseudanomodon to generic level is obviously a preferable solution.

Pseudanomodon (Limpr.) Ignatov & Fedosov , stat. nov. – Basionym: Anomodon [unranked] Pseudanomodon (‘ Pseud-Anomodon ’), Laubm. Deutschl. 2: 774. 1895. Lectotype (selected by Granzow-de la Cerda, 1997): Anomodon attenuatus (Hedw.) Huebener.

Species included:

Pseudanomodon attenuatus (Hedw.) Ignatov & Fedosov , comb. nov. – Basionym: Leskea attenuata Hedw., Sp. Musc. Frond. 230. 1801.

Pseudanomodon giraldii (Müll. Hal.) Ignatov & Fedosov , comb. nov. – Basionym: Anomodon giraldii Müll. Hal., Nuovo Giorn. Bot. Ital. View in CoL , n.s. 3: 117. 1896.

Diagnosis: Plants rather large, yellowish green. Primary stem creeping, without central strand, stoloniform, proximal branch leaves broad (wider than long); secondary stems with much larger leaves, arcuate, attenuate, pinnately branched, foliage subcomplanate; leaves ligulate, broadly acute, entire or with a few teeth near apex; costa ending shortly below apex; laminal cells moderately to sparsely papillose, papillae not pedicellate. Gametangia only on old axes. Perichaetial leaves with smooth or slightly papillose cells of elongate shape. Capsules symmetric, cylindrical, with neck. Annulus not differentiated; operculum obliquely short-rostrate. Exostome teeth striolate proximally, finely papillose distally. Endostome with segments about as long as teeth, cilia reduced or occasionally developed, short. Spores small. Calyptra smooth.

The original description of the subgenus Pseudanomodon included also A. rostratus and A. longifolius . The former species was found nested in a clade of Claopodium ( Gardiner et al., 2005) , and accordingly was placed in that genus ( Ignatov et al., 2006). Anomodon longifolius will be discussed in more detail below.

The two species referred here to Pseudanomodon form a clade in ITS and rpl16, whereas in the concatenated dataset and in atpB-rbcL analyses Homalia trichomanoides appears to be closer to P. attenuatus than to P. giraldii . We do not intend to give a detailed analysis of this puzzling situation; a detailed discussion was promised some time ago by Olsson et al. (2009a), and presumably will appear as a comprehensive special study. It seems worthy mentioning that in the same Pinnatella -clade Enroth et al. (2010) revealed a species more similar to Homaliadelphus than to Pinnatella and any related genera.

Among the additional characters differentiating both species of Pseudanomodon from the Anomodontaceae ( Anomodon and Haplohymenium ) is the much better developed endostome.

At the same time, we must note a considerable difference in the proximal branch leaves, which are compound (sensu Ignatov & Spirina, 2012) in P. giraldii , and much more similar to some Neckeraceae ( Spirina & Ignatov, 2015) , while P. attenuatus has proximal branch leaves similar to Anomodon ( Figs. 12–13 View Fig View Fig ). The papillae in Pseudanomodon are different from the massive papillae of Anomodon ( Fig. 8 View Fig ), which was noted by Granzow-de la Cerda (1997), who also pointed out that the papillae in P. giraldii and P. attenuatus are quite different. As shown in Fig. 8 View Fig , P. attenuatus has rather thin cell walls and small papillae which are occasionally forked, while the papillae in P. giraldii are more scattered, always simple, and the dorsal and ventral cell surfaces are thickened. In Anomodon and Haplohymenium the cell outlines are convex, thus the outlines of the dorsal and ventral surfaces are crenate, while they are plane in both of the Pseudanomodon species.

A new circumscription of the genus Anomodon and family Anomodontaceae

The controversial topologies inferred from the analyses of different DNA regions obviously cannot be followed literally and require a compromise solution.

The core group of Anomodon , i.e. A. viticulosus , A. minor , A. thraustus , A. abbreviatus and A. solovjovii form a maximally supported clade in most analyses, and only a few comments regarding the two latter species are needed. Their laminal cells have papillae of a very unusual shape ( Fig. 9 View Fig ) among other Anomodon . Iwatsuki (1963) segregated them in a separate section, whereas Granzow-de la Cerda (1997) lowered their status to subsectional. The present molecular phylogenetic results showed only a slight segregation, thus supporting the latter view. Although appearing very peculiar, the cells in A. abbreviatus and A. solovjovii are in fact pluripapillose, i.e. the same as in other core Anomodon species, although one papilla “branch” overtops the others and in most parts of the leaf each cell has one papilla much larger than the rest ( Fig. 9 View Fig ). A somewhat similar case is observed in Haplohymenium longinerve ( Czernyadjeva & Ignatova, 2019) . Closer to the leaf margin ( Fig. 9C, D View Fig ) and just above the smooth basal cell area ( Fig. 9F View Fig ) the multipapillosity of Anomodon solovjovii is especially apparent.

Granzow-de la Cerda (1997) included Haplohymenium in Anomodon subgen. Anomodon , as a separate section. It seems that this decision was quite adequate for his study, in demonstrating that Haplohymenium is much closer to Anomodon than to Pseudanomodon and Claopodium rostratum . However, as Granzow-de la Cerda did not put his study in the broader context of pleurocarp systematics at familial level, any recognition that these taxa did not belong to Anomodon at all was impossible. The data available now are rather in favor of the recognition of Haplohymenium as a separate genus. It has a distinct morphology and its clade is situated on a long branch and has maximal support in most analyses ( Figs. 1-6 View Fig View Fig View Fig View Fig View Fig View Fig ). The only odd topology was found in rpl16, where Anomodon minor fell within the Haplohymenium longinerve clade, which we cannot explain based on the present limited sampling. However, the independent status of the genus Haplohymenium requires a fuller discussion, with decisions regarding the statuses of Anomodon longifolius and A. rugelii .

Anomodon longifolius View in CoL was referred to Pseudanomodon View in CoL (as a subgenus) by both Limpricht (1895) and Granzow-de la Cerda (1997), though sharing only a few characters with A. attenuatus View in CoL and A. giraldii View in CoL : the presence of flagelliform branches, the absence of perichaetia beyond the point of last branching (although in A. longifolius View in CoL only thin branches occur on the stem distally from the perichaetia), and elongate cells in the perichaetial leaves. The papillae of Anomodon longifolius View in CoL are different from the core species of Anomodon View in CoL , Pseudanomodon View in CoL , and probably from any other pleurocarp ( Figs. 10 View 20 µm ). Higher magnification of moist leaf cells shows that the papillae are not centered over the cell lumen, as is usually assumed, but paired. SEM images taken from young leaves ( Fig. 10A, B View 20 µm ) and from dry herbarium material ( Fig. 10C, D View 20 µm ), show the papillae of two neighboring cells originating from the wall between these two cells (sometimes a longitudinal, sometimes a transverse one). In the dry state ( Fig. 10C, D View 20 µm ), the papillae are inclined towards the center of the lumen and their tips cover the concavities that appear as a result of cell drying.

The proximal branch leaves of Anomodon longifolius View in CoL are narrow and occasionally compound, sometimes similar to those in Anomodon rugelii View in CoL and Heterocladium dimorphum View in CoL and related species, while they are broad in Anomodon View in CoL , Haplohymenium and Pseudanomodon View in CoL ( Figs. 12 View Fig ).

The peristome of Anomodon longifolius differs strongly from the peristomes of core Anomodon : its exostome teeth are relatively narrow, transversely striolate below, longitudinally striolate to prominently papillose above, and finely papillose on the inner surface, whereas in core Anomodon the teeth are massive, smooth below on the outer surface and indistinctly papillose above ( Figs. 14 View Fig ).

The position of Anomodon longifolius in ITS trees in the Neckeraceae – Lembophyllaceae – Echinodiaceae – Ortostichellaceae clade, and in some analyses sister to Heterocladium dimorphum (albeit with very low support), and its basalmost position in organellar trees (or sister to A. rugelii , both keeping a basa-l position in this case) indicate the necessity of excluding A. longifolius from the genus Anomodon . Therefore it is separated here in its own genus. Its familial placement requires further study, but for now, in a view of the topologies of the organellar trees, we retain it in the Anomodontaceae .