1. Introduction
Mosses are the most diversified lineage among Bryophytes [
1,
2]. Recent years have witnessed important advances in the elucidation of the phylogenetic relationships among the major lineages of mosses, e.g., order level and above [
3,
4,
5]. However, much remains to be clarified at lower levels, which is a key issue to establish a robust molecular-based classification, especially in the case of larger families.
Orthotrichaceae Arn. is the second most speciose family of mosses, with an estimated 900 species [
6,
7]. Most are epiphytic taxa, both from tropical and temperate environments, although in each of these major climatic regions one of the two subfamilies of the group predominates. Macromitrioideae Broth. are cladocarpous mosses with almost exclusively tropical distribution, whereas Orthotrichoideae Broth. includes acrocarpous mosses that inhabit temperate regions of both hemispheres and high tropical mountains [
8]. Orthotrichoideae is better known, both in terms of specific diversity, e.g., [
9] and phylogenetically, e.g., [
6,
10,
11]. Recently, Draper et al. [
12,
13] have provided new and more complete insights in the phylogenetic framework of the subfamily, besides some of the evolutionary patterns underlying its complexity. According to these works, Orthotrichoideae is composed of fourteen genera. Ten of them are grouped into Orthotricheae, a tribe that in turn integrates two well differentiated lineages recognized as subtribes: (i) Orthotrichinae, that includes
Orthotrichum s.str., the most diversified genus; (ii) and Lewinskyinae, which includes
Lewinskya and
Ulota, the other two major genera of the tribe. The other four recognized genera are currently integrated into Zygodonteae [
12], with
Zygodon s.str. as the most species-rich genus of this tribe (
Table 1).
The phylogenetic reconstruction obtained by Draper et al. [
12] did not conclusively resolve the analysed representation of Zygodonteae, as it separated its components in two well-supported clades but for which no robust conclusions could be drawn about their sister relationships. The results showed that one of the genera,
Codonoblepharon, could constitute a separate lineage from Zygodonteae. Nevertheless, the placing of that lineage in a polytomy, together with a clade containing the rest of the genera of Zygodonteae and a clade containing the genera of Orthotrichoideae, advised postponing the possible consequences for the systematic of the subfamily until obtaining robust evidence.
Codonoblepharon, a genus initially conceived as grouping very heterogeneous taxa [
14], was later circumscribed by Goffinet and Vitt [
8] to segregate a section from
Zygodon (Sect.
Bryoides Malta), mainly characterized by the absence of papillae in its leaf cells [
15]. Thus conceived,
Codonoblepharon contains about seven mainly tropical and southern hemisphere species [
14,
16], although there are discrepancies about the ascription of
C. forsteri (Dicks.) Goffinet, the only one restricted to the northern hemisphere. While studies based on morphological or biogeographical evidence suggest the exclusion of this species from
Codonoblepharon [
8,
14], others based on phylogenetic reconstructions with partial taxa representation [
6,
12] indicate that it should be recognized as part of this genus.
In the present work we evaluate the hypothesis, based on the results obtained by Draper et al. [
12], that
Codonoblepharon represents a lineage independent of Zygodonteae. Our main objective is to achieve a robust phylogenetic reconstruction that would unequivocally reflect the relationships of
Codonoblepharon within the subfamily Orthotrichoideae. Besides, we pursue to obtain more information about the phylogenetic structure of
Codonoblepharon and the taxonomic position of
C. forsteri.
3. Discussion
This work presents an updated molecular phylogeny with representation of all the accepted genera of the subfamily Orthotrichoideae based on Draper et al. [
12], except for the recently described
Rehubryum F.Lara, Garilleti and Draper [
13] from Lewinskyinae. Special attention was paid to the tribe Zygodonteae and the genus
Codonoblepharon. The variability and large number of analyses that have been conducted, consistent with each other, minimize the possibility of topological inconsistencies. The only analysed marker that has shown ambiguous alignments due to its high variability is nuclear
ITS2. Nevertheless, the exclusion of these ambiguities did not affect phylogenetic results.
Orthotrichoideae is a taxonomically complex subfamily, which has led to numerous supraspecific reassignments. The species of
Zygodon s.l., including the group with smooth leaf cells now segregated in the genus
Codonoblepharon, have long been considered a large and important natural group. Unlike
Orthotrichum s.l., the other traditional large genus of acrocarpous Orthotrichaceae,
Zygodon s.l., includes mostly dioicous mosses, with sporophytes bearing a long seta and usually producing vegetative propagules on variably long and branched filamentous supports arising from the stem, never directly from the leaves, although propagules may be grouped in the leaf axils. Other distinguishing characteristics are the growth of the colonies forming lax turfs, rarely mats or cushions, the relatively small leaves with non or slightly recurved margins and little or no cell differentiation along the lamina, and the cucullate non-plicate calyptra, typically devoid of hairs. They typically grow on old or decaying bark of large trees or stumps. As in other groups of Orthotrichaceae, several species are saxicolous, either facultative or, more rarely, obligate [
15,
16] and [
17] (pp. 15–135).
Malta [
15], in the only world monograph on the genus, recognized a total of 77
Zygodon species grouped into four sections. Most of them (ca. 86%) belonged to the globally distributed section
Euzygodon Müll.Hal. Section
Stenomitrium Mitt. included a single species, with Andean and Patagonian distribution and deviant morphology due to its robust and creeping stems, leaves in pentastichous arrangement and dimorphic basal leaf cells. Section
Obtusifolii Malta also contained a single species characterized by lingulate leaves with a rounded apex and papillose calyptra and a wide disjunct distribution including populations in Southeast Asia, Australasia, South America and Mexico. Finally, section
Bryoides Malta included nine species, mainly distributed in the southern hemisphere and, furthermore, occurring in the north in some tropical localities and in Europe, characterized by smooth leaf cells. Malta [
15] argued that this latter section represented a natural group that might merit subgeneric rank, as smooth cells are unusual among Orthotrichaceae. Malta’s taxonomic delimitation was basically followed by Calabrese [
16], even though Goffinet and Vitt [
8] had reinstated a few years earlier the genus
Codonoblepharon for most of the species of section
Bryoides and
Bryomaltaea Goffinet to segregate
Zygodon obtusifolius Hook (
Table 3).
Zygodon s.l. has been treated at various taxonomic ranks, including family (Zygodontaceae Schimp.) and subfamily (Zygodontoideae Broth.). These mosses have numerous characters shared with other Orthotrichaceae, but some deviate and more closely resemble representatives of other families, such as Ditrichaceae Limpr., Grimmiaceae Arn. or Pottiaceae Hampe [
8,
18,
19], that are not phylogenetically close. In fact, Schimper [
20] proposed the segregation of
Zygodon as a family that he considered intermediate between Orthotrichaceae and Weissiaceae Schimp. (=Pottiaceae p.p.). Most of its components are currently included in Pottiaceae, and the proposal also included
Amphidium Schimp., now placed in Amphidiaceae M.Stech. Among the characters that can be considered deviant are those that have been highlighted as being particularly significant [
21]: the cucullate non-plicate calyptra and the orthotropic stems forming short turfs, not cushions or creeping mats as in other Orthotrichaceae. The current consideration of
Zygodon s.l. as a tribe within Orthotrichoideae is relatively recent [
8]. However, regardless of the taxonomic rank given to the group, it has traditionally been considered a fairly homogeneous entity, well differentiated from the rest of Orthotrichaceae. This idea that Zygodonteae is a natural and quite uniform group was only partially questioned in early molecular studies [
6], which allowed the segregation of
Bryomaltaea obtusifolia (Hook.) Goffinet as part of a phylogenetically distant lineage. However, this did not diminish the general consideration of Zygodonteae. Recently, however, Draper et al. [
12] concluded that the genus
Zygodon is a polyphyletic artificial group and their results supported the distinction of
Pentastichella with the inclusion of
Pleurorthotrichum Broth. and the establishment of the new genus
Australoria (
Table 1 and
Table 3). Thus, Malta’s sections of
Zygodon [
15] are now treated as separate genera,
Zygodon s.s. being currently restricted to the representatives of the section
Euzygodon (
Table 3).
The results of the present work support the distinction of
Codonoblepharon as a separate lineage, as earlier suggested by Goffinet et al. [
6] and Draper et al. [
12]. Unequivocal evidence for this was obtained by inclusion in the analysis of a wide representation of the former Zygodonteae, including species with smooth leaf cells and a greater number of representatives with papillose leaf cells. The phylogenetic reconstruction obtained here places the monophyletic group constituted by
Codonoblepharon as a sister group of the clade that includes all the other Orthotrichoideae. This suggests the need for recognition of this lineage as a separate tribe, which we propose to name Codonoblepharonteae; it is the third within the Orthotrichoideae, along with Zygodonteae and Orthotricheae. Thus,
Codonoblepharon changes from being considered just a section of
Zygodon to a major independent lineage among Orthotrichoideae. This taxonomic proposal is morphologically supported by smooth leaf cells, which is an exclusive character of Codonoblepharonteae within the subfamily. This classification is paralleled by Macromitrioideae, which contains a single genus characterized by entirely smooth leaf cells,
Schlotheimia Brid., being also considered a separate tribe, Schlotheimieae Goffinet [
8]. Future phylogenetic studies on Macromitrioideae may reveal the true relationships of this group with smooth cells within the subfamily.
The phylogenetic reconstruction (
Figure 1) also shows that the lengths of the branches and the topology within Codonoblepharonteae are similar to what can be observed in other main lineages, such as Zygodonteae, which includes up to three genera. In contrast, in the new tribe all the terminals belong to a single genus,
Codonoblepharon. This leads us to consider that the taxonomic diversity of the group is yet to be investigated and that the segregation of
Codonoblepharon into several separate genera could be possible. Further in-depth studies are needed to unravel this possibly overlooked diversity, giving special attention to
C. pungens, which, according to Malta [
15], constitutes the systematic weak point of the group and to
C. minutum (Müll.Hal. and Hampe) Matcham and O’Shea which in our reconstructions appears as the sister species of all the congeners included in the analysis. Further evidence of the possibly overlooked diversity among Codonoblepharonteae is the strong phylogenetic structure obtained for
C. menziesii which could have considerable taxonomical and biogeographical significance. The New Zealand samples of
C. menziesii are separated from the Australian and Californian ones, which could imply that they correspond to different taxa and might support the idea of Shevock [
22] that the occurrence of
C. menziesii in western North America is due to a recent introduction from Australia. Since Malta [
15] already recognized a strong intraspecific morphological variability within
C. menziesii, the current concept could hide a complex of species and calls for a deep integrative study.
Codonoblepharon forsteri is limited to Europe and northwestern Africa, making it the only representative of the genus with a Holarctic distribution [
23]. Regarding its phylogenetic reconstruction, the samples are nested within the lineage of Codonoblepharonteae. Their segregation in a subclade together with
C. pungens is significant as they are the only two autoicous species of the genus included in the analysis. In Orthotrichoideae, most genera are either dioicous or autoicous [
12,
13], which could support their segregation into a separate genus, although, once again, it is preferable to await the results of a more complete morphological and molecular study to resolve this in a robust and accurate manner.