How to Recognize Mosses from Extant Groups among Paleozoic and Mesozoic Fossils

Abstract

This paper describes a range of Paleozoic and Mesozoic mosses and assesses how far they can be referred to extant taxa at the family, ordinal, or class levels. The present study provides new data on Paleozoic mosses of the order Protosphagnales, re-evaluating affinities of some groups previously thought to be unrelated. The leaf areolation pattern, combined with the leaf costa anatomy, results in the subdivision of Protosphagnales into five separate families: Protosphagnaceae (at least six genera), Polyssaieviaceae (at least three genera), and three monogeneric families: Rhizonigeritaceae, Palaeosphagnaceae, and Servicktiaceae. We urge caution in referring Paleozoic and Early Mesozoic fossil mosses as members of Dicranidae and Bryidae, as they may belong to the extinct moss order Protosphagnales. Additional evidence supports the relation of the Permian genus Arvildia to extant Andreaeopsida. We segregate Late Palaeozoic and Early Mesozoic mosses that are superficially similar to extant members of either Dicranales or Polytrichales, into the artificial informal group of Archaeodicranids, distinguishing them from ecostate Paleozoic and Mesozoic mosses, which are combined here into another artificial informal group, Bryokhutuliinids. The latter includes the genus Bryokhutuliinia, widespread in contemporary Asia, from the Middle Jurassic to the Lower Cretaceous, as well as other superficially similar ecostate plants from different regions worldwide, ranging from the Upper Palaeozoic to the Lower Cretaceous. A list of Paleozoic, Mesozoic, and Eocene moss fossils suitable for age calibration in phylogenetic trees is provided.

1. Introduction

Molecular phylogenetic methods have been used in bryophyte taxonomy for only the past 25 years. Within this short period, novel views on relationships between the main lineages became generally stable, suitable for rewriting textbooks and floras. Phylogenetic reconstructions help us to gain a new understanding of the history of the second largest extant plant group, the mosses. However, even the most comprehensive phylogenetic trees cannot provide a complete picture of past moss evolution without fossil records. Fossils are the only source of information on extinct groups, whose diversity remains poorly known due to relatively few and incompletely preserved fossil remains. This paper attempts to combine data on the most common Paleozoic and early Mesozoic mosses and assess if some of them can be referred to the taxa known in the extant moss flora.

The recent publication of Bechteler et al. [1] provided a comprehensive overview of bryophyte evolution, involving all three main lineages: hornworts, liverworts, and mosses. This analysis offered a robust phylogenetic reconstruction based on 228 nuclear genes for 531 species of almost all known orders of bryophytes. The age-calibrated tree found the ages of moss divergence at the subclass level in the Carboniferous and Permian, at the order level mainly in the Jurassic, and rapid diversification at the family level in the Cretaceous. A similar acceleration of diversification during the Cretaceous has already been demonstrated in ferns [2] and liverworts [3]. Earlier publications on moss phylogeny calibration [1,4,5] were based on more limited datasets and achieved less detailed results. However, the general conclusions are similar: the key events of moss diversification occurred earlier than the Jurassic and Cretaceous. Unfortunately, fossil data that might clarify the origin of mosses and characterize their diversity in the past are extremely scarce and not always interpretable with confidence.

Feldberg et al. [6] undertook a comprehensive revision of all fossil liverworts, compiling a list of records suitable for the calibration points in liverwort phylogeny. Their list includes 42 ages (48 species, some combined for one age). Notably, the age of forty listed taxa is between 15 and 35 Ma, five taxa are between 52 and 112 Ma, and only three are older: one from 158 Ma (Jurassic), one from 228 Ma (Triassic), and one from 383 Ma (Devonian). This scarcity is not due to a lack of Mesozoic and Paleozoic specimens: Tomescu et al. [7] listed 85 records from Mesozoic and 16 from Paleozoic. However, most of these liverworts are thallose, with only a few characters that are helpful for assessing their relationships with extant lineages.

In mosses, the situation is even more complicated for at least three reasons. The first is that moss classification, from its early days [8] and up to the most recent [1,9], circumscribe the main lineages by their sporophytic (specifically peristomial) traits, while the vast majority of moss fossils are known only from their gametophytes. The second reason is homoplasy, which hampers the correct evaluation of morphological characters and their value for classification. This led to a significant rearrangement of moss groups at the family and even at the class and subclass levels using the molecular approach [9]. The third reason is that Paleozoic records are rich and morphologically diverse, but this does not help much in their taxonomic interpretation, as their differences from extant mosses are too numerous. This means that they likely belong to extinct groups, tentatively supporting the conclusion of massive extinctions, parallelling those in gymnosperms [5]. However, unlike fossil gymnosperms, which include extinct classes and numerous extinct orders [10], fossil mosses remain largely unclassified above the generic rank, except for one Paleozoic order Protosphagnales [11] and one Early Cretaceous family Tricostaceae [12]. This creates the impression that other Paleozoic and Early Mesozoic mosses can be referred to extant groups, though one may extrapolate that the proportion of extinct and extant families might be similar to, for example, gymnosperms. An important reason for the underdeveloped classification of moss fossils is obvious: mosses are small, and important details of their structure are rarely preserved satisfactorily for describing a taxon of higher rank.

Nevertheless, recent advances in the study of fossil moss anatomy in permineralized specimens [13] and the use of tomography [14] have opened broad new perspectives on the ancient history of mosses. One further method, bulk maceration, was first applied to the study of Paleozoic mosses by Gomankov and Meyen [15]. Their collections from the 1970–1980s were studied further by Ignatov in 1990, who presented new data on as many as 13 moss genera, including 8 genera described as new to science [16]. Most of these were based on the collections from the Lagerstätte Aristovo. In 2022, we gathered considerable additional fossil material from the same locality. Our study has now revealed previously unknown structural features in already described taxa, shedding light on the possible relationships of several genera.

This paper presents novel results on Aristovo specimens and discusses and illustrates selected morphological traits that we propose as important for the classification of fossil mosses. Accordingly, a classification of fossil mosses into several formal and informal groups is suggested. We are aware that the density of data from different ages and different parts of the world remains too sparse for a proper classification, but we believe that our attempt could help in selecting taxa for molecular phylogenetic tree calibration by defining their relationships with extant moss taxa.

2. Materials and Methods

2.1. Locality

The material for the present study was collected in 2022 by Ignatov, Voronkova and Bashkuev from the Vologda Region, Velikoustyugskii District, Russia. The collection site is located on the right bank of the Lesser Northern Dvina River, approximately 100 m upstream of the Aristovo pier, ca. 60.764 N, 46.386 E.

Gomankov and Meyen referred to these deposits as belonging to the Vyatkian Horizon, Upper Tatarian Substage, Upper Permian [15]. Later, Gomankov assigned them to the upper part of the Wuchiapingian Stage, Lopingian Series of the Permian [17], following the new Geological Time Scale.

Moss remains were found in association with Tatarina pteridosperms and numerous leaves of Cordaites. Additionally, the presence of numerous and morphologically diverse oogonia of Charophyta supports previous assumptions [15,17] that the material originates from oxbow lake deposits.

The collected material was deposited in the MHA Herbarium, paleobryological section. Specimens in glycerol jelly from the new Aristovo collections were labelled with the prefix ‘Aristovo-’, which is omitted here in the figure captions where only specimen numbers are provided. Specimens names that started with CUT are slides with 1 or 2 micron sections (thickness marked on slides) and samples used for TEM or SEM.

In addition to new Aristovo collections, the present paper includes new images of Permian collections of Rhizinigerites previously published in [15,17] from Viled (deposited in the Geological Institute of Russian Academy of Science (GIN RAS)), collections of Polyssaievia published in [11] from various localities (listed in the captions of its illustrations), deposited in GIN RAS, and collections of Kulindobryum published in [18] from the Middle Jurassic of Transbaikalia (deposited in the Borissyak Paleontological Institute of Russian Academy of Science (PIN RAS)).

2.2. Materials

The plant-bearing deposits consist of grey and dark grey argillites. Pieces of argillites were transported in plastic boxes in wet conditions and treated with fungicides. In the laboratory, the material was submerged for four weeks in a liquid consisting of one part aqueous saturated solution of ethylenediaminetetraacetic acid (EDTA) disodium salt (to extract calcium content) and two parts hydrofluoric acid. After washing, the plant fragments were sorted and mounted on glass slides in glycerol jelly using a standard protocol, except for specimens prepared for destructive studies (see below).

2.3. Microscopy and Photography

Specimens mounted in glycerol jelly were photographed under an Olympus CX43 light microscope with an Infinity 1–2 camera and Olympus BX53 microscope with an Infinity 3–3 camera. Z-stacks of several images were generated using HeliconFocus 4.50 [19].

2.4. Destructive Studies

Twenty specimens belonging to Protosphagnum nervatum, Palaeosphagnum meyenii, Servicktia undulata, S. vorcutannularioides, and Arvildia elenae were selected for anatomical destructive studies. These specimens were chosen from mostly broken or incomplete specimens. The following methods were employed:

2.4.1. Anatomical Sections

The material, initially kept in water with fungicide, was washed in distilled water, dehydrated in a graded alcohol series (20%, 40%, 60%, 80%, and 96%), followed by an alcohol–acetone mixture (1:1), and pure acetone for 15 min in each solution. The material was then soaked in a graded acetone–resin mixture series (3:1, 1:1, 1:3) for 6, 12, and 1 h, respectively, before being embedded in Epon–Araldite resin, as recommended by the manufacturer. The resin was polymerized at 37 °C for 24 h and then at 60 °C for an additional 24 h. For light microscopy, serial sections 1–2 µm thick were made with an LKB ultramicrotome using glass knives. The sections were placed on glass slides without a mounting medium and photographed using an Olympus BX63 light microscope with a long working distance objective lens (100×/0.80).

2.4.2. TEM Observations

Ultrathin sections, 60 nm thick, were prepared using a LEICA ARTOS 3D ultramicrotome with a diamond knife (45° angle). TEM sections were either unstained or contrasted with uranyl acetate, lead citrate, or ammonium molybdate. The sections were examined using a JEM-1011 TEM (Jeol, Japan) at 80 kV with a CCD ORIUS SC1000W camera, controlled by GATAN Digital Micrograph software, version 2.32, in the Laboratory of Electron Microscopy at the Faculty of Biology, Lomonosov Moscow State University.

2.4.3. SEM Studies

SEM observations were conducted using a ThermoScientific Quattro S equipped with a field emission gun. Material from bulk maceration was glued onto carbon tape affixed to aluminium stubs and observed without gold or platinum coating. Observations were conducted at an accelerated voltage of 15 kV in environmental SEM (ESEM) mode, at an air pressure of approximately 500 Pa, using SE mode.

2.5. Taxa Names

The authorities for the genera and species names used in this paper are listed in the subsection Systematic Paleontology and in Table 1, Table 2 and Table 3.

3. Results

The bulk maceration of argillite pieces yielded numerous plant fossil fragments, including mosses. These mosses are primarily represented by leaf fragments, with the stems bearing partially damaged leaves occurring less frequently. Among the mosses from Aristovo, Protosphagnum nervatum and Intia spp. are the most common, each displaying a range of morphotypes (Figure 1) that are still pending classification. The existing circumscriptions [11,15,16,20,21] provide too few names to adequately cover their diversity. Moreover, significant variation can exist within a single leaf [21,22].

Specimens of other moss genera are much rarer. In the earlier studies of Aristovo flora conducted during the 1970s and 1980s, only a few of their leaves or stem fragments were found. For instance, when the genus Palaeosphagnum was first described, it was known from only four relatively large leaf fragments. Due to such rarity, all specimens were mounted on permanent slides without any thought of destructive studies.

The increased availability of specimens now allows for anatomical studies. Particular attention was given to the costa of Protosphagnum because its subapical branching (Figure 2A–E) corresponds to an anomalous meristematic zone at the leaf apex [23]. Another focus of the study was on the branch buds, which had not been described previously in Protosphagnum. Although not always abundant, some stems contain 2–4 buds within a 5 mm segment. The recent bryological literature extensively discusses the shape and arrangement of foliose structures around the branch primordia, particularly in Hypnales, and to a lesser extent in Bartramiales and Hedwigiales. The branch primordia of Protosphagnum are relatively easy to observe (Figure 2F–K).

Our attempts to make cross-sections of Protosphagnum leaves revealed that anatomical studies of these fossil mosses are challenging. The delicate bodies of these plants, when sectioned at 2 µm from a resin block, appear to lack any discernible structure (Figure 3A–C). The cell contents in the TEM images appear almost homogeneous, with cell boundaries rarely visible and uncertain at best. The walls of hyaline cells in Protosphagnum are thinner than half a micron. Figure 3 presents several images of these sections. Under a light microscope, hyaline cells can be seen with fibril-like structures across them (Figure 3K), which are actually the cell walls, with approximately 2 μm spacing at points where they appear as a double line (i.e., where they join the dorsal and ventral surfaces of the hyaline cell). The costa ‘section’ fragment in Figure 3I (arrowed) shows its unistratosity, an unusual structure in mosses.

Given the limited anatomical information obtained from the resin-embedded sections of Protosphagnum leaves, we studied them using SEM. Leaf sections cut with a razor blade and observed without a metal coating in environmental mode provided more informative images, confirming our expectation of costa unistratosity (Figure 4). The original descriptions of the genera Intia and Protosphagnum characterized their costa as multistratose, although they acknowledged partial unistratosity [11]. We examined leaves from 10 specimens with slightly different areolation patterns. In all leaf sections, the costa appeared unistratose, even in areas where it looked somewhat thicker (Figure 4A, arrow). Despite the costa appearing multistratose under a light microscope in frontal view (Figure 2), likely due to thick-walled costa cells with a narrow lumen equal to or narrower than the cell thickness (Figure 4), we never observed any apparent costa multistratosity in Protosphagnum.

While some species in this group might have costa with more than one cell layer, most part of the costa in Protosphagnum is unistratose. For descriptive purposes, we will refer to the costa as unistratose, as shown in Figure 4, while admitting some discontinuous bistratosity. A re-examination of the previous collections of protosphagnalean mosses [23] confirmed that the costa in their leaves has a very similar structure. We also found unistratose costa in Rhizinigerites, which is apparent even in frontal views (see below).

In contrast to Protosphagnum, the costa in other Paleozoic genera with cell dimorphism, such as Palaeosphagnum and Servicktia, were found to be multistratose. Details on the structure of these genera and their taxonomic implications are discussed in the taxa descriptions below. Similarly, new findings on the genus Arvildia from Aristovo will be presented in the discussion of this genus, where its systematic position is strongly supported by the new Aristovo collections.

4. Discussion

This section begins by discussing the Permian mosses in the order Protosphagnales and re-evaluating their conspectus. It then addresses Paleozoic and Mesozoic fossils that may be compared with extant moss taxa, selecting the most appropriate candidates for phylogenetic tree calibrations. Next, it reviews mosses with ambiguous systematic positions and finally provides general comments on the search for the oldest fossils of certain moss lineages.

A complete list of fossils with comments on their possible relation to extant groups has been published [21], including some doubtful and incomplete records. Most of these latter records are not included in the current discussion, although further studies could significantly increase their importance. Instead, here we focus on a selected group of better-known taxa that are maximally useful for the practical purposes outlined in this paper’s title. Where relevant, we have inserted a “Caution” paragraph to prevent misplacing a fossil record into an unrelated group.

4.1. Protosphagnalean Mosses

Mosses of the order Protosphagnales were first described by Neuburg as an extinct group [11,20] and have since been widely accepted as such [15,16,21,24]. However, the distinctions between these mosses and other groups have been interpreted differently by various authors, reflecting our increased understanding of their structure.

Neuburg [11,20] described nine fossil moss genera from the Lower to Upper Permian of Angaraland (now in Siberia and NE European Russia) and classified them into two orders: Bryales and Protosphagnales. The latter included three genera with clearly dimorphic laminal cells: Junjagia, Vorcutannularia, and Protosphagnum. Six other genera with more or less monomorphic laminal cells were retained in Bryales (at that time understood as mosses other than Sphagnales and Andreaeales). Abramov and Savicz-Lyubitskaya even suggested placing Intia, one of the genera described by Neuburg, in the extant family Mniaceae [25]. The leaf shape and size, distinct leaf border, and the leaf variation series in Paleozoic protosphagnalean mosses and the genus Intia are quite similar to those in modern Mniaceae. However, Protosphagnum and Intia exhibit numerous transitional morphotypes (Figure 1) and are apparently closely related [16]. These transitions were also noted in the original discussion of Protosphagnum and were interpreted as evidence of the origin of Protosphagnales from an Intia-like ancestor, as the latter is more common in older deposits than Protosphagnum and other genera with lamina cell dimorphism [11].

To accommodate these intermediate morphs across the numerous studied localities in NE European Russia, the genus Syrjagia Fefilova was described with a Protosphagnum-like leaf structure in its proximal part, while distally it is indistinguishable from Intia [26,27]. However, further studies [16,21,22] have shown that the co-occurrence of protosphagnalean and bryoid areolation within a single leaf is even more common, posing a challenge for formal classification. This phenomenon has been discussed in terms of cell divisions [22], showing that the irregular development of cell T-triads, TT-tetrads, or TTT-pentads [22] results in the protosphagnalean areolation pattern appearing in different parts of leaves without obvious rules (Figure 1C,E).

Protosphagnalean mosses also exhibit unique pathways of leaf cell differentiation that differ significantly from the pattern almost universally present in all extant mosses. In the latter, cell divisions continue near the leaf base longer than in the upper parts [28]. In protosphagnalean mosses [23], active cell divisions may continue longer in other parts of the leaf, including the mid-leaf or apical region (Figure 2B,D). This trait of Protosphagnum likely corresponds to the unusual costa pattern, which has distal and lateral branching associated with unistratosity.

The results from the new Aristovo collections reveal that mosses with cell dimorphism include groups with uni- and multistratose costae, different patterns of laminal cell dimorphism, the presence or absence of a limbidium (with several different patterns if present), different leaf shapes and sizes, stem branching, and varying coverage of branch primordia by foliose structures. This structural polymorphism makes protosphagnalean mosses too diverse to be classified as a single, undivided taxon. Therefore, we propose a classification of the order Protosphagnales into five families.

4.2. Systematic Paleontology

Order Protosphagnales Neuburg, Trudy Geol. Inst. Akad. Nauk SSSR 19: 61. 1960.

Description. Stems branched; branch initials either surrounded by foliose structures apart from the apical cells and surrounded the branch base or foliose structures all occur on branch. Leaves broadly ovate, lingulate or lanceolate, usually limbate throughout, with the border cells having oblique transverse walls, rarer elimbate, costa reaching usually 0.7–0.9 the leaf length, unistratose or multistratose; laminal cells usually dimorphic with protosphagnalean arrangement followed the T-pattern in cell divisions [22], rarer monomorphic.

The order has been described based on the genus: Protosphagnum Neuburg, Trudy Geol. Inst. Akad. Nauk SSSR 19: 75. 1960.

The order includes five families; all of them are described below.

4.2.1. Protosphagnaceae Ignatov fam. nov.

Type genus: Protosphagnum Neuburg, Trudy Geol. Inst. Akad. Nauk SSSR 19: 75. 1960.

Type of the genus: Protosphagnum nervatum Neuburg, Trudy Geol. Inst. Akad. Nauk SSSR 19: 75, pl. 71–74, pl. 75: 1–2, pl. 76–78; f. 43–52. 1960.

Description. Stems sparsely irregularly branched, leaves widely spaced but rosette-like crowded terminally. Rhizoids rare, absent in most specimens. Foliose structures present on stems around branch primordia, cochleariform, broadly rounded or truncate, well developed and regularly retained after full branch development. Leaves spreading to patent, broadly to narrowly ovate or elliptic, usually broadly rounded to apex, rarely acute; auriculate, cordate or gradually narrowed to the base; very shortly decurrent; margins plane, bluntly serrate, bordered throughout, occasionally in young leaves less so in apical part; border cells unistratose, uni- or biseriate, overlapping with cells above and shortly bent outwards; laminal cells varing from hexagonal to arranged in T-triads, TT-tetrads or TTT-pentads; costa terminating shortly below leaf apex; moderately broad, unistratose, with lateral branches. Spherical or pyriform brood bodies covered by rhizoids sometimes present on the stem or dorsal side of costa.

Genera included: Protosphagnum Neuburg, Intia Neuburg, Kosjunia Fefilova, Vorcutannulatia Neuburg, Junjagia Neuburg, Bulbosphagnum Maslova & Ignatov.

Illustrations: Figure 1, Figure 2, Figure 3 and Figure 4; [11]: Plates 1–18, 59–78; [15]: Plate 2, Figures 1, 2, Text-Figure 5; [16]: Plate 7, Figures 84–93, Text-Figure 20; [21]: Figures 1, 2, 4, 6; [22]: Figures 2–3, 5–6; [23]: Figures 2, 4–79; [26]: Plates 1–5, 12–16, 18–22.

Caution: Young leaves from thin stems (likely grown in unfavorable conditions) might be much smaller, have shorter costa, have a weak to almost absent limbidium, and have a totally homogeneous areolation [29]. This variant is rare; it is reminiscent of Bryum plants left in a wet state in a plastic bag for days, so they develop thinner stems with leaves lacking a limbidium and sometimes becoming ecostate.

Comments: The important traits of protosphagnalean mosses were discussed in a previous publications focusing on the development of the protosphagnalean leaf areolation pattern [22], the distribution of the zones of active cell division within the leaves [23], and young leaf variation [29]. All their results pertain to the Protosphagnaceae, as established here.

4.2.2. Rhizinigeritaceae Ignatov fam. nov.

Type genus: Rhizinigerites S.V. Meyen, Trudy Geol. Inst. Akad. Nauk SSSR 401: 28. 1986.

Type of the genus: Rhizinigerites neuburgiae S.V. Meyen, Trudy Geol. Inst. Akad. Nauk SSSR 401: 28, f. 6. 1986.

Description. Stems sparsely branched, with three types of branches: (1) leafy branches similar to maternal stem, but with somewhat smaller leaves; (2) rhizoidophores, leafless axes with bundles of rhizoids; (3) prerhizoidophores, similar to rhizoidophores but lacking rhizoids. Foliose structures on stem around branch primordia absent. Rhizoids commonly present, some apparently arising from basal parts of stem with abundant rhizoids covering the stems, which generate them. Leaves erect-spreading to patent, ovate-lanceolate, gradually tapered to narrow linear-lanceolate upper part, at apex abruptly broadly acute and the leaf tip blunt, towards the base slightly rounded; margins plane, unbordered, bluntly serrulate due to some marginal cells slightly protruding beyond the leaf outline, maximally one cell wide; laminal cells short elongate or rectangular, with rounded ends, homogeneous or arranged in T-triads, and in some areas in leaf middle arranged in distinct oblique rows, where neighboring cell rows alternate in cell orientation: in one row elongate cells (1.5–2:1) are elongated in the oblique row direction, whereas in the next row, the cells have generally the same proportions (1.5–2:1), but are elongated in the direction parallel to the costa, thus forming areolation pattern similar to sphagnoid but without one cell of T-triads. Thus the leaves have numerous, more or less regularly arranged one-cell perforations; costa rather slender, disappearing shortly below leaf apex; moderately broad below but two cells wide above, looking unistratose in many places, with occasional presence of lateral branches formed by longer and darker cells, continued into lamina as unistratose, 1–2 seriate veins; the expression of veins varied, being apparent only in large, better developed specimens, especially near the leaf base. Brood bodies unknown.

Genus included: Rhizinigerites Meyen.

Illustrations: Figure 5; [15]: Plate 2, Figures 3, 4, 6, Text-Figure 6 p.p.; [16]: Plate 5, Figures 60–67, Plate 6, Figures 68–80, Text-Figures 15–18; [21]: Figure 5.

Caution: Leaf cells in the apical, submarginal, and juxtacostal regions are usually more homogeneous, as well as the smaller leaves from thin branches (Figure 5).

Comments: The presence of veins, similar to Polyssaievia (see below), was not previously reported for this genus.

4.2.3. Palaeosphagnaceae Ignatov fam. nov.

Type genus: Palaeosphagnum Ignatov, Palaeontographica, Abt. B, Paläophytol. 217: 177.

Type of the genus: Palaeosphagnum meyenii Ignatov, Palaeontographica, Abt. B, Paläophytol. 217: 177. f. 24; pl. 7: f. 89–92. 1990.

Description. Stem unknown. Leaves ovate to ovate-lanceolate gradually tapered to acute apices, slightly rounded to base; margin entire, bordered by unistratose, 2(–3)-seriate limbidium, limbidium cells separated by oblique cell walls; laminal cells dimorphic: the narrow darker cells forming loops embracing several cells, often 7 in a specific pattern (3 + 1 + 3), but in more cases with additional cells, so the areolation often looks composed of a fairly irregular loops of 4 to 20 cells; costa stout and broad, multistratose, disappearing below leaf apex, without lateral branches. Brood bodies unknown.

Genus included: Palaeosphagnum Ignatov.

Illustrations: Figure 6; [15]: Plate II, Figures 10–13, as Muscites sp. SVM-2; [16]: Plate 7, Figures 89–92; Text-Figure 24.

Caution: The areolation of Palaeosphagnum is unlikely to be confused with other mosses if the leaf fragment includes at least 50–100 laminal cells. The smaller fragments may be confused with the basal fragments of other protosphagnalean mosses, e.g., Vorcutannularia or Servicktia.

Comments: Ignatov [16] described the genus based on only four fragments from the leaf middle parts with the strange areolation and stout costa. Based on the new collections, we added to the description of the genus brief data on the leaf apex, limbidium structure, costa anatomy, and overall variation.

4.2.4. Servicktiaceae Ignatov fam. nov.

Type genus: Servicktia Ignatov, Palaeontographica, Abt. B, Paläophytol. 217: 18. 1990.

Type of the genus: Servicktia acuta Ignatov, Palaeontographica, Abt. B, Paläophytol. 217: 181. f. 27: e, h; pl. 8: 94–95, 98–101. 1990.

Description. Stem unknown. Leaves oblong or ovate-triangular and gradually tapered to acute apex or from broadly oblong base reflexed and tapered to broadly lanceolate upper portion; plane, keeled or undulate; margin entire, bordered by unistratose, 2(–3)-seriate limbidium, limbidium cells separated by oblique cell walls, longer and usually broader than cells inwards; laminal cells small, quadrate or polygonal throughout the leaf and thus monomorphic, or in lower leaf part dimorphic with scattered large pellucid cells or with a regular and clear protosphagnoid areolation by cells forming T-triads or TT-tetrads; costa stout and broad, multistratose, disappearing below leaf apex, without lateral branches, at least sometimes with enlarged cells of the ventral epidermis. Brood bodies unknown.

Genus included: Servicktia Ignatov; ? Ignatievia Ignatov, see comment 2.

Illustrations: Figure 7, Figure 8 and Figure 9; [15]: Plate 2, Figures 5–8, as Muscites sp. SVM-1; [16]: Plate 8, Figures 94–106, Plate 9, Figures 107–109, Text-Figure 27.

Comments: (1) The type of S. acuta that is the type species of the genus Servicktia has small leaves with homogeneous cells. The very specific limbidium formed by the large cells with oblique cell walls, however, links several species, which differ from one another in their leaf shapes. One of them, S. vorcutannularioides, has a rather regular laminal cell dimorphism, albeit not identical to that in Protophagnaceae. Following the discovery of a leaf with a typical protosphagnoid pattern, it is described here as a new species.

(2) Ignatievia Ignatov has been described as a separate genus [16] having mammillosely prorate laminal cells, resembling Bartramiaceae and, therefore, is placed in Bryidae [21]. The presence in Ignatievia of the border similar to that in Servicktiaceae in the cell arrangement and also the observation of protruding cell walls near the joints with neighboring cells in Servicktia undulata (Figure 7) tentatively suggests the placement of Ignatievia in the Servicktiaceae but retained in a separate genus because of the longer laminal cells and coarsely serrate leaf apex. However, the final decision on its position should be delayed until the discovery of additional collections of this very rare moss.

The description of the new species:

Servicktia tatyanae Ignatov sp. nov. Figure 9.

Holotype: Aristovo-100B-1 (MHA, paleobryological collection!). Figure 9.

Locality: Aristovo, Vologda Region, Velikoustyugskii District.

Age: Upper Permian, Lopingian [15,17].

Diagnosis. Servicktia tatyanae differs from other species of the genus in clearly protosphagnoid structure in the lower part of leaf.

Description. Leaves no less than 3.5 mm long from a broadly oblong base 1.3 mm wide, reflexed at 2/3 the leaf length, narrowed to broadly lanceolate upper portion that is gradually tapered, decreasing in width from 0.9 mm to 0.5 mm over 1.3 mm, apparently keeled; margin entire, bordered by unistratose, 2(–3)-seriate limbidium, limbidium cells separated by oblique cell walls, 70–130 × 12–16 µm below, 30–65 × 10–13 µm distally, longer than cells inwards; laminal cells in distal leaf portion quadrate, 8–22 × 8–24 µm, in subtrasversal rows; cells of the basal part dimorphic, with clear protosphagnoid areolation by T-triad cells, cells forming loops 23–27 × 10–12 µm, isodiametric cells inside the loops ca. 15 µm, close to the leaf base and near costa loops are larger, up to 150–200 × 35–60 µm, with one or less often two cells inside the loop, costa strong, broad, no less than 300 µm wide, composed by elongate cells to 160 × 15–20 µm, without lateral branches.

Etymology: the species is named in honor of Tatyana V. Voronkova, a Russian plant physiologist who was deeply involved in various bryophyte studies, including fossils.

Caution: The laminal cells of S. undulata, S. acuta, and cells in the upper part of the leaf of S. tatyanae are isodiametric, and in the case of unclear expression of limbidium, they can be compared with many groups of acrocarpous mosses, e.g., Ditrichaceae. The border cells with oblique cell walls are somewhat similar to those in Fissidens, where border cells are also longer than laminal cells, which are fairly uniform [16]. However, Fissidens leaves have a unique structure having dorsal and vaginant laminae, which is absent in Servicktia.

4.2.5. Polyssaieviaceae Ignatov fam. nov.

Type genus: Polyssaievia Neuburg, Dokl. Akad. Nauk SSSR 107(2): 322. 1956.

Type of the genus: Polyssaievia spinulifolia (Zalessky) Neuburg, Dokl. Akad. Nauk SSSR 107(2): 322, 7–10. 1956. Basionym: Walchia spinifolia Zalessky, Probl. Paleontol. 1: 234, f. 16. 1936.

Description. Stems regularly pinnate branched. Leaves erect to spreading, straight or recurved near leaf middle, from the broadly ovate base gradually narrowed into lanceolate-triangular upper part, rounded to base; margin entire, bordered by unistratose, 1–2-seriate limbidium, limbidium cells indistinctly differentiated; laminal cells in distal leaf part elongate to linear, with narrowly acute upper and lower ends, towards the leaf base shorter, rectangular or rhombic, monomorphic throughout the leaf or forming ‘net venation’ by narrow, usually two cell wide veins formed by longer and darker cells; veins spreading from the base and far in the lamina, ending among laminal cells or fused, but never reaching the leaf margin and never extending into the upper part of the leaf, where all cells are narrow; vein expression varies from the most pronounced development when veins occur throughout the broadened basal part of the leaf and are absent only in the most distal part, to the ‘net venation’ characterized by only few veins near the leaf base, or it is totally absent, at least in most leaves. Distinct oblique rows of laminal cells sometimes present, but within the areas with ‘net venation’ the rows are not as long as in leaves where the ‘net venation’ is poorly developed or undeveloped. Sphagnalean areolation pattern rarely expressed; when present, it is seen in a few areas near the leaf base as alternating cells rows where cells in neighboring rows are elongated perpendicularly.

Genera included: Polyssaiaevia Neuburg, Uskatia Neuburg, Baidaevia Neuburg, and ? Salairia Neuburg.

Illustrations: Figure 10; [11]: Plates 22–52; [21]: Figure 3.

Caution: With their regular pinnate branching, densely arranged acute leaves and narrow laminal cells with acute upper and lower ends, mosses of this family, especially Uskatia, resemble extant pleurocarpous mosses; moreover, ‘net venation’ is usually not expressed in Uskatia leaves.

Comments: (1) The genus Salairia somewhat resembles Protosphagnacae (large rhombic cells, though without apparent triads), but also to Polyssaieviaceae (cells towards the margins are markedly narrower). However, since bulk maceration studies were not conducted for this genus, its familial placement remains uncertain.

(2) Uskatia dentata Fefilova is likely closer to Intia than to Uskatia and hence, is placed in the Protosphagnaceae family.

(3) T-patterns are rare in Polyssaieviaceae, being observed only near the leaf base (e.g., Plate 24, Figure 3 in [11]); however, the areolation at leaf bases in Polyssaievia is similar to that in the juvenile leaves of Protosphagnum (Figure 10J). The finding of thin veins within the leaf lamina, also in Rhizinigerites now provides additional evidence for the placement of Polyssaieviaceae in Protosphagnales.

4.3. Paleozoic and Meseozoic Mosses Putatively of Extant Classes, Subclasses, or Orders

This section first discusses Permian mosses in the Andreaeopsida, whose placement in this class is supported by the new Aristovo collections. We then review other taxa where the plant structure is sufficient for certain placement in the extant subclass Bryidae. Following this, we discuss mosses that likely belong to an extant group, but whose features are insufficient for placement in a specific order.

4.3.1. Andreaeopsida

The distinct position of the genus Andreaea has been recognized since the late 19th century, when mosses were classified into three main groups: Sphagnum, Andreaea, and all others. Almost all subsequent classifications, including the most recent [1,30], support this position of Andreaea. Despite the apparent great age of this lineage, there have been no fossil records of Andreaeobryales and Andreaeales.

In its original description, the genus Arvildia from the Upper Permian was compared with Pottiaceae, especially Didymodon, due to a generally similar habit. The families Grimmiaceae, Orthotrichaceae, and Seligeriaceae were also considered. Later, its affinity to Andreaeales was reported as a possibility [21,31]. We can now confirm its placement in Andreaeopsida with stronger evidence (Figure 11), further supported by the bulk maceration of new collections from Aristovo.

Figure 11A,D illustrates the Andreaea pattern of areolation in the subapical part of the leaf, where the apical cell undergoes unifacial divisions, resulting in a uniseriate subapical cells. The leaf then becomes biseriate, and abruptly, four or more seriate [31]. This pattern of unifacial division in the apical cell is unique in Andreaea (Figure 11C). The arrangement of cells resulting from the typical bifacial cell division of the apical cells in mosses has been previously illustrated [31,32]. A similar pattern occurs in the small leaves of Andreaeobryales (Figure 11D), although in larger leaves, the apical cell is bifacial. Further studies of the genus Arvildia may reveal additional insights into its relationships with Andreaeales and Andreaeobryales, though we recommend using it as a reference for the minimal age of the lineage, Andreaea + Andreaeobryum.

4.3.2. Bryopsida Subclass Bryidae

Mosses with an apparent affinity to other extant groups are listed in Table 1. The correlation between their position in the phylogenetic tree and the age of their records is weak, largely due to the rarity of unambiguous fossil evidence. Fossils are lacking for the classes Takakiopsida, Oedipodiopsida, Tetraphidopsida, and the subclasses Buxbaumiidae, Gigaspemidae, Funariidae, and Timmiidae. Ancient groups, such as Sphagnopsida, Polytrichopsida, and Diphysciidae, have unequivocal fossil records only from the Mesozoic era and onwards. Therefore, the fact that fossil mosses referable to Dicranidae are only known from the Cretaceous—later than the subclass Bryidae, which is known from the Jurassic—can likely be attributed to the fortuitous discovery of fossils with conspicuous characteristics sufficient for their placement in the extant orders Splachnales and Hypnales.

The placement of Kulindobryum taylorioides [18] within Bryidae is particularly notable. This species was described based on the compressions of capsules that exhibit putative peristome linear elements at the mouth, appressed to the capsule wall from the outside. Two capsules allowed us to count the number of teeth, which were 32—a rare characteristic, known only in some species of the extant genus Tayloria, Splachnaceae (Figure 12). Despite its suboptimal preservation, this unique structure, discernible in the rocks, allows for a reasonably confident identification, making it useful for calibration purposes.

The genus Paleodichelyma, from the Upper Jurassic of Transbaikalia, was originally placed in the Hypnales [33]. The taxonomic placement of P. sinitzae became even more certain after the description of a second species in this genus, P. kiritchkovae [34], with numerous capsules in a lateral position and a laminal areolation pattern of the prosenchymal cells, which is typical for Hypnales.

The presence of Bryidae in the Jurassic is also consistent with the findings of the representatives of the most advanced order of Bryidae, the Hypnales, as early as in the the Valanginian age of the Early Cretaceous [12,35]. The permineralized Tricosta [12] and Krassiloviella [35] from British Columbia exhibit all the characteristics necessary to refer them to Hypnales, including lateral perichaetia and branch buds surrounded by juvenile leaves. These two genera were assigned to the extinct family Tricostaceae [12] due to their tricostate leaves, a feature unusual in extant mosses. Other mosses with tricostate leaves were described from the Late Jurassic of the Russian Far East (based on bulk maceration material) [36] and were later found in several locations in Transbaikalia [37] as imprints with well-recognizable cellular structures. Their close relationship with Tricostaceae is highly likely. If better preserved collections of Asian plants confirm their structural similarity, the age of Tricostaceae could be extended to the Jurassic. However, even if this is not confirmed, the Valanginian age of Tricostaceae, which is undoubtedly Hypnales, the crown clade of Bryidae definitively supports the Jurassic age of the Bryidae lineage as a whole.

4.3.3. Bryopsida Subclass Dicranidae

The Cretaceous records of Dicranidae are unequivocal. The Campylostelium is known from both its gametophyte, with a complex costal structure, and its sporophyte, which exhibits a Dicranales s.l. structure of peristome [38]. The costa anatomy in some fossils is sufficient to place them within extant genera Campylopus and Dicranodontium [14].

Most families of Dicranidae are represented by a single species and often by a single collection (Table 1), with the exception of Leucobryaceae, which are represented by four species. One reason for the more numerous records of Leucobryaceae is the highly specific costal structure characteristics of some genera, including Campylopus and Dicranodontium. A second reason is that Leucobryaceae diverged from the main evolutionary line of Dicranidae earlier than Dicranaceae [1], suggesting that Leucobryaceae may have been more diverse during the Mesozoic. A third reason is a higher diversity of Leucobryaceae than of Dicranaceae in the tropics. Notably, Cretaceous Dicranidae from higher latitudes are represented only by two findings from Alaska (Rhabdoweisiaceae: Cynodontium luthii [39]) and the Antarctic (Dicranaceae or Rhabdoweisiaceae, Livingstonites gabrielse [40]).

Unfortunately, poor preservation prevents the detailed observation of the costal structures and thus precludes the unequivocal generic and familial placement of some Dicranidae specimens. Such cases are discussed below in the context of the ‘Archaeodicranid’ group.

Table 1. Paleozoic and Mesozoic mosses referred to extant orders or classes (* Leucobryaceous moss referred to Leucobryaceae [7] according to [21] may belong to Octoblepharaceae; ** Livingstonites gabrielae was not compared with any extant moss family, but cross section of costa is typical for Dicranales). Following Tomescu [13], we retain the Ordovician record of Sphagnum [41] with a ‘reservation of caution’ because of the too big gap from other records. The taxa are arranged according to phylogenetic trees [1]. A few Eocene records are included to document selected families not known so far from earlier deposits.

Class, Order, or Family	Species/Material	Age	Region	Reference
		Permian and Mesozoic
Sphagnales	Chlorosphagnum cateficense Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Andreaeopsida + Andreaeobryopsida	Arvildia elenae Ignatov	Permian, Lopingian	Russia, NE European	[30]
Polytrichaceae	Meantoinea alophosioides Bippus, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[42]
Polytrichaceae	Polytrichastrum incurvum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Polytrichaceae	Eopolytrichum antiquum Konopka, Herend., G.L. Merr. & P.R. Crane	Cretaceous, Campanian	USA, Georgia	[43]
Diphysciaceae	Physcidium tortuosum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Diphysciaceae	Physcidium simsimiae Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Grimmiaceae	Tricarinella crassiphylla Savoretti, Bippus, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[44]
Leucobryaceae	Canaliculidium fissuratum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Leucobryaceae	Dicranodontium minutum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Leucobryaceae	Campylopus lusitanicus Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Leucobryaceae	Leucobryaceae gen. sp.	Cretaceous, Aptian/Albian	Portugal	[14]
Dicranellaceae	Campylopodium allonense Konopka, Herend. & P.R. Crane	Cretaceous, Santonian	USA, Georgia	[38]
Leucobryaceae or Octoblepharaceae *	Leucobryaceous moss	Cretaceous, Valanginian	Canada, British Columbia	[7]
Calymperaceae	Calymperites burmensis Heinrichs, Schäf.-Verw., Hedenäs, Ignatov & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[45]
Rhabdoweisiaceae	Cynodontium luthii Bippus, G.W. Rothwell & Stockey	Cretaceous, (Santonian/Campanian to Maastrichtian?)	USA, Alaska	[40]
Dicranales **	Livingstonites gabrielae E.I. Vera	Lower Cretaceous	Antarctic, South Shetland Islands	[40]
Ditrichaceae	Ditrichaceae gen. sp.	Cretaceous, Aptian/Albian	Portugal	[14]
Splachnaceae	Kulindobryum taylorioides Ignatov	Middle or Late Jurassic	Russia, Transbaikalia	[18]
Hypnodendrales	Vetiplanaxis pyrrhobryoides N.E. Bell	mid-Cretaceous amber	Myanmar	[46]
Hypnodendrales	Vetiplanaxis espinosus Hedenäs, Heinrichs & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[47]
Hypnodendrales	Vetiplanaxis longiacuminatus Hedenäs, Heinrichs & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[47]
Hypnodendrales	Vetiplanaxis oblongus Hedenäs, Heinrichs & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[47]
Hypnales	Palaeodichelyma kiritchkovae A. Frolov, Kazan. & Enushch.	Early Jurassic, Plinsbachian	Russia, Irkutsk	[34]
Hypnales	Palaeodichelyma sinitzae Ignatov	Late Jurassic or Early Cretaceous	Russia, Transbaikalia	[33]
Hypnales	Tricosta priapiana C. Blanco-Moreno, Valois, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[12]
Hypnales	Krassiloviella limbelloides G.W.K. Shelton, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[35]
		Eocene
Sphagnaceae	Sphagnum sect. Cuspidata (Lindb.) Schimp.	Eocene	Ukraine, Rovno amber	[48]
Sphagnaceae	Sphagnum sect. acutifolia Wilson	Eocene	Baltic amber	[49]
Polytrichaceae	Atrichum P. Beauv.	Eocene	Baltic amber	[50]
Grimmiaceae	Grimmia Hedw.	Eocene	Baltic amber	[51]
Leucobryaceae	Brothera Müll. Hal.	Eocene	Baltic amber	[51]
Leucobryaceae	Campylopodiella Cardot	Eocene	Baltic amber	[50]
Leucobryaceae	Campylopus Brid.	Eocene	Baltic amber	[50]
Rhachitheciaceae	Hypnodontopsis Z. Iwats. & Nog.	Eocene	Baltic amber	[50]
Mniaceae	Trachycystis Lindb.	Eocene	Baltic amber	[50]
Mniaceae	Trachycystis Lindb.	Eocene	Ukraine, Rovno amber	[52]
Rhizogoniaceae	Pyrrhobryum Mitt.	Eocene	Baltic amber	[53]
Aulacomniaceae	Arrhenopterum heterostichoides (Janssens, D.G. Horton & Basinger) Ignatov comb.nov. (basionym: Aulacomnium heterostichoides Janssens, D.G. Horton & Basinger, Canadian Journal of Botany 57: 2153. f. 4–5, 7–8, 10–18. 1979.)	Eocene	Canada, British Columbia	[54]
Ephemeropsidaceae	Ephemeropsis K.I. Goebel	Eocene	Germany	[55]
Leskeaceae	Haplocladium (Müll. Hal.) Müll. Hal.	Eocene	Baltic amber	[50]
Herpetineuraceae	Herpetineuron (Müll. Hal.) Cardot	Eocene	Baltic amber	[56]
Sematophyllaceae	Aptichella (Broth.) Herzog	Eocene	Baltic amber	[50]

Table 1. Paleozoic and Mesozoic mosses referred to extant orders or classes (* Leucobryaceous moss referred to Leucobryaceae [7] according to [21] may belong to Octoblepharaceae; ** Livingstonites gabrielae was not compared with any extant moss family, but cross section of costa is typical for Dicranales). Following Tomescu [13], we retain the Ordovician record of Sphagnum [41] with a ‘reservation of caution’ because of the too big gap from other records. The taxa are arranged according to phylogenetic trees [1]. A few Eocene records are included to document selected families not known so far from earlier deposits.

Class, Order, or Family	Species/Material	Age	Region	Reference
		Permian and Mesozoic
Sphagnales	Chlorosphagnum cateficense Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Andreaeopsida + Andreaeobryopsida	Arvildia elenae Ignatov	Permian, Lopingian	Russia, NE European	[30]
Polytrichaceae	Meantoinea alophosioides Bippus, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[42]
Polytrichaceae	Polytrichastrum incurvum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Polytrichaceae	Eopolytrichum antiquum Konopka, Herend., G.L. Merr. & P.R. Crane	Cretaceous, Campanian	USA, Georgia	[43]
Diphysciaceae	Physcidium tortuosum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Diphysciaceae	Physcidium simsimiae Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Grimmiaceae	Tricarinella crassiphylla Savoretti, Bippus, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[44]
Leucobryaceae	Canaliculidium fissuratum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Leucobryaceae	Dicranodontium minutum Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Leucobryaceae	Campylopus lusitanicus Hedenäs, Bomfleur & E.M. Friis	Cretaceous, Aptian/Albian	Portugal	[14]
Leucobryaceae	Leucobryaceae gen. sp.	Cretaceous, Aptian/Albian	Portugal	[14]
Dicranellaceae	Campylopodium allonense Konopka, Herend. & P.R. Crane	Cretaceous, Santonian	USA, Georgia	[38]
Leucobryaceae or Octoblepharaceae *	Leucobryaceous moss	Cretaceous, Valanginian	Canada, British Columbia	[7]
Calymperaceae	Calymperites burmensis Heinrichs, Schäf.-Verw., Hedenäs, Ignatov & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[45]
Rhabdoweisiaceae	Cynodontium luthii Bippus, G.W. Rothwell & Stockey	Cretaceous, (Santonian/Campanian to Maastrichtian?)	USA, Alaska	[40]
Dicranales **	Livingstonites gabrielae E.I. Vera	Lower Cretaceous	Antarctic, South Shetland Islands	[40]
Ditrichaceae	Ditrichaceae gen. sp.	Cretaceous, Aptian/Albian	Portugal	[14]
Splachnaceae	Kulindobryum taylorioides Ignatov	Middle or Late Jurassic	Russia, Transbaikalia	[18]
Hypnodendrales	Vetiplanaxis pyrrhobryoides N.E. Bell	mid-Cretaceous amber	Myanmar	[46]
Hypnodendrales	Vetiplanaxis espinosus Hedenäs, Heinrichs & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[47]
Hypnodendrales	Vetiplanaxis longiacuminatus Hedenäs, Heinrichs & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[47]
Hypnodendrales	Vetiplanaxis oblongus Hedenäs, Heinrichs & A.R. Schmidt	mid-Cretaceous amber	Myanmar	[47]
Hypnales	Palaeodichelyma kiritchkovae A. Frolov, Kazan. & Enushch.	Early Jurassic, Plinsbachian	Russia, Irkutsk	[34]
Hypnales	Palaeodichelyma sinitzae Ignatov	Late Jurassic or Early Cretaceous	Russia, Transbaikalia	[33]
Hypnales	Tricosta priapiana C. Blanco-Moreno, Valois, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[12]
Hypnales	Krassiloviella limbelloides G.W.K. Shelton, Stockey, G.W. Rothwell & Tomescu	Cretaceous, Valanginian	Canada, British Columbia	[35]
		Eocene
Sphagnaceae	Sphagnum sect. Cuspidata (Lindb.) Schimp.	Eocene	Ukraine, Rovno amber	[48]
Sphagnaceae	Sphagnum sect. acutifolia Wilson	Eocene	Baltic amber	[49]
Polytrichaceae	Atrichum P. Beauv.	Eocene	Baltic amber	[50]
Grimmiaceae	Grimmia Hedw.	Eocene	Baltic amber	[51]
Leucobryaceae	Brothera Müll. Hal.	Eocene	Baltic amber	[51]
Leucobryaceae	Campylopodiella Cardot	Eocene	Baltic amber	[50]
Leucobryaceae	Campylopus Brid.	Eocene	Baltic amber	[50]
Rhachitheciaceae	Hypnodontopsis Z. Iwats. & Nog.	Eocene	Baltic amber	[50]
Mniaceae	Trachycystis Lindb.	Eocene	Baltic amber	[50]
Mniaceae	Trachycystis Lindb.	Eocene	Ukraine, Rovno amber	[52]
Rhizogoniaceae	Pyrrhobryum Mitt.	Eocene	Baltic amber	[53]
Aulacomniaceae	Arrhenopterum heterostichoides (Janssens, D.G. Horton & Basinger) Ignatov comb.nov. (basionym: Aulacomnium heterostichoides Janssens, D.G. Horton & Basinger, Canadian Journal of Botany 57: 2153. f. 4–5, 7–8, 10–18. 1979.)	Eocene	Canada, British Columbia	[54]
Ephemeropsidaceae	Ephemeropsis K.I. Goebel	Eocene	Germany	[55]
Leskeaceae	Haplocladium (Müll. Hal.) Müll. Hal.	Eocene	Baltic amber	[50]
Herpetineuraceae	Herpetineuron (Müll. Hal.) Cardot	Eocene	Baltic amber	[56]
Sematophyllaceae	Aptichella (Broth.) Herzog	Eocene	Baltic amber	[50]

4.4. Paleozoic and Mesozoic Mosses Insertae Sedis

‘Archaeodicranid’ group

The majority of Paleozoic and Mesozoic fossils that do not belong to Protosphagnales exhibit an easily recognizable morphotype that cannot be confidently assigned to any extant moss orders or even subclasses. These fossils are relatively large plants with erect, densely foliate stems, and appressed, erect, or erect-spreading leaves that are narrow, typically 5–10 (–20) times longer than wide. The single costa extends to the leaf tip and, although often not clearly visible, it is apparently present even in poorly preserved specimens. Fossils of such plants could be interpreted as belonging to Dicranidae, Polytrichales, or some orders of Bryidae (e.g., Bartramiales), but usually without sufficient clues to determine a specific order.

Mosses of this group are widely reported (Table 2). For simplicity from this point forward, we use the informal name ‘Archaeodicranids’ to refer to fossils that superficially resemble species of Dicranaceae s.l. or Polytrichaceae, but whose available structural features do not preclude their placement in multiple orders. Archaeodicranids may also include plants from the early stages of the Polytrichopsida and Dicranidae lineages, before they acquired their characteristic conductive tissues shown clearly from the Cretaceous. Referring to a specimen as an Archaeodicranid indicates that its taxonomic position is too broad for precise age calibration. This classification, however, helps to separate these fossils from those with ecostate leaves or otherwise different morphologies.

Several Archaeodicranids merit further comment.

Merceria augustica from Antarctica [57] is known from their transverse sections of stem and leaves, along with some fragments of the frontal views of lamina cells. The combination of a multistratose, the moderately homogeneous costa in transverse section, leaves up to 2.5 mm wide, and elongate-rhomboidal to linear laminal cells with narrowly acute upper and lower ends is not known in extant mosses. Smoot and Taylor compared Merceria with many groups, including genera described by Neuburg, but did not find sufficient similarity [57]. The narrowly acute cells are particularly unusual, as this feature is characteristic of advanced Bryidae groups, such as Hypnales. Representatives of Dicranidae rarely have cells of this shape, though some early divergent groups (e.g., Flexitrichum, Distichium) have submarginal cells in broad sheathing leaf bases with oblique ends somewhat similar to the Merceria cells. However, the differentiated costa in Flexitrichum and Distichium makes their affinity with Merceria rather unlikely. The comparison with Polyssaieviaceae may warrant additional attention, as the cell shape in the upper part of the leaf in Polyssaievia (Figure 10B,E,F) and Uskatia [11] is very similar to that of Merceria. The absence of a ‘net venation’ in Polyssaieviaceae in the distal part of the leaves raises the possibility that this similarity is not merely superficial. Although the costa in Polyssaievia appears quite thick, further anatomical studies are needed to confirm or refute a possible relationship with Protosphagnales. Some evidence for similarity between the vascular plant flora of Angaraland and Gondwana in the Permian [58] raises the possibility that protosphagnalean mosses might occur in the Permian in the latter region too, being not confined only to Angaraland and Subangaraland, where they are known now.

Heinrichsiella from the Jurassic of Patagonia [59] presents another challenge in distinguishing between Polytrichales and early divergent Dicranidae, especially Timmellaceae. The original description suggested possible affinities with both groups. Interestingly, the bistratose lamina occurs in the earliest divergent Polytrichales (e.g., Alophozia, Bartramiopsis, Lyellia, and Lower Cretaceous Meantoinea), some early divergent Dicranidae (e.g., Timmiellales), and Diphysciidae (e.g., extant Diphyscium and Cretaceous Physcidium [14]). A bistratose lamina is not common among extant mosses, and its occurrence in Archaeodicranids warrants further investigation. Similar to the case of Heinrichsiella, we refrain from assigning Lower Permian Paleocampylopus [60] to either Dicranidae or Polytrichales, despite its conspicuous rosette-like leaves at the stem tip, which are similar to the perigonia of extant Polytrichum.

Table 2. Paleozoic and Mesozoic fossil mosses referred to Archaeodicranid group.

Species	Age	Region	Reference
Merceria augustica Smoot & T.N. Taylor	Permian	Antarctica	[57]
Palaeocampylopus buragoae Ignatov & Shcherb.	Permian	Russia, Primorsky	[60]
Talchirophyllites indicus Sh. Chandra	Permian	India	[61]
Yguajemanus yucapirus I.C.C. Souza, Ricardi-Branco & Y. León	Permian	Brazil, southern	[62]
Viledia minuta Ignatov	Permian	Russia, NE European	[16]
Dwykea goedehoopensis J.M.Anderson & H.M. Anderson	Permian	South African Republic	[63]
Dwykea araroi Ricardi-Branco, J.S. Costa, I.C.C. Souza, Ronh, Longhim & R.S. Faria	Permian	Brazil	[64]
Saksenaphyllites saksenae Sh. Chandra	Permian	India	[61]
Umariaphyllites acutus Sh. Chandra	Permian	India	[61]
Capimirinusriopretensis I.C.C. Souza, Ricardi-Branco & Y. León	Permian	Brazil	[62]
Moss cuticles (Figure 4E)	Triassic	Antarctic, East	[65]
Stachybryolites zhoui X.W. Wu, Xiu Y. Wu & Y.D. Wang	Lower Jurassic	China, Xinjang	[66]
Muscites sp.	Lower Cretaceous	Argentina	[67]
Muscites sp. 1–3	Lower Cretaceous	Russia, Transbaikalia	[37]
Heinrichsiella patagonica Bippus, Savoretti, Escapa, García Massini & Guido	Jurassic	Argentina	[59]
Muscites cretaceus Debey & Ettingsh.	Cretaceous	Germany	[68]
Ningchengia jurassica Heinrichs, X. Wang, Ignatov & M. Krings	Upper Jurassic	China	[69]

Table 2. Paleozoic and Mesozoic fossil mosses referred to Archaeodicranid group.

Species	Age	Region	Reference
Merceria augustica Smoot & T.N. Taylor	Permian	Antarctica	[57]
Palaeocampylopus buragoae Ignatov & Shcherb.	Permian	Russia, Primorsky	[60]
Talchirophyllites indicus Sh. Chandra	Permian	India	[61]
Yguajemanus yucapirus I.C.C. Souza, Ricardi-Branco & Y. León	Permian	Brazil, southern	[62]
Viledia minuta Ignatov	Permian	Russia, NE European	[16]
Dwykea goedehoopensis J.M.Anderson & H.M. Anderson	Permian	South African Republic	[63]
Dwykea araroi Ricardi-Branco, J.S. Costa, I.C.C. Souza, Ronh, Longhim & R.S. Faria	Permian	Brazil	[64]
Saksenaphyllites saksenae Sh. Chandra	Permian	India	[61]
Umariaphyllites acutus Sh. Chandra	Permian	India	[61]
Capimirinusriopretensis I.C.C. Souza, Ricardi-Branco & Y. León	Permian	Brazil	[62]
Moss cuticles (Figure 4E)	Triassic	Antarctic, East	[65]
Stachybryolites zhoui X.W. Wu, Xiu Y. Wu & Y.D. Wang	Lower Jurassic	China, Xinjang	[66]
Muscites sp.	Lower Cretaceous	Argentina	[67]
Muscites sp. 1–3	Lower Cretaceous	Russia, Transbaikalia	[37]
Heinrichsiella patagonica Bippus, Savoretti, Escapa, García Massini & Guido	Jurassic	Argentina	[59]
Muscites cretaceus Debey & Ettingsh.	Cretaceous	Germany	[68]
Ningchengia jurassica Heinrichs, X. Wang, Ignatov & M. Krings	Upper Jurassic	China	[69]

4.5. ‘Bryokhutuliinid-Group’

The genus Bryokhutuliinia was first described from Mongolia [70] and later found in three localities in Transbaikalia [71,72,73], dating from the Upper Jurassic to the Lower Cretaceous, as well as in the Lower Jurassic of the Irkutsk coal basin [34].

These plants are known from imprints and are characterized by their large size and conspicuous appearance. They typically have stems that are regularly pinnately branched, with foliage that ranges from terete to subcomplanate, and forms rosettes terminally. The leaves are ovate-oblong to ovate-lanceolate, ecostate, and limbate, with lamina cells that are rectangular (length to width ratio mostly 2–3:1) and border cells that are linear and thick-walled. Sporophytes are known for two of the five species in the genus and are found on short lateral branches. The capsules are immersed or slightly emergent, ovate, and have a conic operculum. Mamontov and Ignatov [18] discussed the taxonomic position of Bryokhutuliinia and concluded that its most probable affinity is with Dicranidae rather than Bryidae (e.g., Hypnales). However, their evidence is not yet sufficient for calibration purposes and more data on this moss are needed.

Several fossils with similar features other than Bryokhutuliinia are listed in Table 3. Whether these mosses represent a natural group, or their similarities are due to their shared aquatic ecology (a possibility supported in some cases by the associated entomofauna for Bryokhutuliinia [74]) remains an open question, which warrants further studies. Several ecostate mosses with appressed leaves are included in Table 3 with question marks.

Table 3. Fossil mosses referred to the Bryokhutuliinids group, ecostate mosses, mostly with spreading leaves and rarer with erect leaves (marked with asterisk).

Species/Images	Age	Region	Reference
Moss cuticles (Bomfleur et al., 2014, Figure 2)	Permian	Antarctic, East	[65]
Muscites brickiae P. Moisan, S. Voigt, J.W. Schneid. & Kerp	Triassic	Kyrgyzstan	[75]
Bryokhutuliinia ignatovii A. Frolov, Kazan. & Enushch.	Lower Jurassic	Russia, Irkutsk	[34]
Bryokhutuliinia crassimarginata Ignatov, Karasev, Sinitsa & Maslova	Middle or Upper Jurassic	Russia, Transbaikalia	[71]
Bryokhutuliinia jurassica Ignatov	Upper Jurassic or Lower Cretaceous	Mongolia	[70]
Bryokhutuliinia ingodensis (Srebrod.) Ignatov	Upper Jurassic or Lower Cretaceous	Russia, Transbaikalia	[72,73]
Bryokhutuliinia obtusifolia Ignatov & Shcherb.	Lower Cretaceous	Russia, Transbaikalia	[37]
Baigulia complanata Ignatov, Karasev & Sinitsa	Upper Jurassic	Russia, Transbaikalia	[73]
Muscites fontinalioides Krassilov	Upper Jurassic	Russian, Khabarovsk (Bureya)	[36]
Muscites samchakianus Srebrod.	Upper Jurassic or Lower Cretaceous	Russia, Transbaikalia	[72]
Baiguliella minuta Ignatov, Karasev & Sinitsa	Upper Jurassic	Russia, Transbaikalia	[73]
* Muscites sp. (plate II, f. 4–7)	Triassic	Kyrgyzstan	[75]
* Muscites guesceliniae Townrow	Permian	South African Republic	[76]

Table 3. Fossil mosses referred to the Bryokhutuliinids group, ecostate mosses, mostly with spreading leaves and rarer with erect leaves (marked with asterisk).

Species/Images	Age	Region	Reference
Moss cuticles (Bomfleur et al., 2014, Figure 2)	Permian	Antarctic, East	[65]
Muscites brickiae P. Moisan, S. Voigt, J.W. Schneid. & Kerp	Triassic	Kyrgyzstan	[75]
Bryokhutuliinia ignatovii A. Frolov, Kazan. & Enushch.	Lower Jurassic	Russia, Irkutsk	[34]
Bryokhutuliinia crassimarginata Ignatov, Karasev, Sinitsa & Maslova	Middle or Upper Jurassic	Russia, Transbaikalia	[71]
Bryokhutuliinia jurassica Ignatov	Upper Jurassic or Lower Cretaceous	Mongolia	[70]
Bryokhutuliinia ingodensis (Srebrod.) Ignatov	Upper Jurassic or Lower Cretaceous	Russia, Transbaikalia	[72,73]
Bryokhutuliinia obtusifolia Ignatov & Shcherb.	Lower Cretaceous	Russia, Transbaikalia	[37]
Baigulia complanata Ignatov, Karasev & Sinitsa	Upper Jurassic	Russia, Transbaikalia	[73]
Muscites fontinalioides Krassilov	Upper Jurassic	Russian, Khabarovsk (Bureya)	[36]
Muscites samchakianus Srebrod.	Upper Jurassic or Lower Cretaceous	Russia, Transbaikalia	[72]
Baiguliella minuta Ignatov, Karasev & Sinitsa	Upper Jurassic	Russia, Transbaikalia	[73]
* Muscites sp. (plate II, f. 4–7)	Triassic	Kyrgyzstan	[75]
* Muscites guesceliniae Townrow	Permian	South African Republic	[76]

5. Conclusions

This paper summarizes the history of bryophytes through a collection of fossil data spanning the Paleozoic and Mesozoic eras and various regions of the Earth. Most fossil mosses are known from only a few specimens collected at single localities. There are, however, two notable groups that existed for over 20 million years and were distributed across at least two thousand kilometres. These are the mosses of the order Protosphagnales from the Permian and Bryokhutuliinia from the Jurassic and Lower Cretaceous of Central Asia. Both groups disappeared relatively quickly as the former seems to have been a part of the Permo–Triassic extinction event, while the latter did not survive the Cretaceous terrestrial revolution.

Both groups were apparently aquatic and such specialization likely influenced their extinction during a period of rapid global environmental change. Some Protosphagnales described here exhibited a less developed ‘sphagnalean areolation pattern’ [22] and multistratose costa, suggesting they might have been adapted or adaptable to terrestrial ecosystems. However, linking these ‘largely protosphagnoid’ mosses, such as Servicktia, to extant taxa remains challenging without a complete lineage of descendants.

Consequently, for these extinct aquatic plants, including Protosphagnales and Bryokhutuliinia, we do not recommend searching for representatives among extant taxa. Instead, we suggest that ‘habitually terrestrial’ Archaeodicranids, if their costa structure is preserved, could provide more informative insights. Notably, no Paleozoic mosses with complex costa structures have been identified. Only slightly larger cells in the ventral epidermis of Merceria and scattered substereids in Arvildia demonstrate a case of poor structural differentiation of the costa.

The case of Arvildia is particularly noteworthy due to its rare trait of cell arrangement in the apical part of the leaf, which could be of significant value, even though it is not present in every leaf. Similarly, Kulindobryum is promising due to its capsules having 32 pendent peristome teeth—a unique trait, although preservation issues may cast some doubt (Figure 12).

The main target for future studies should be better preserved Archaeodicranids, where the structure of the costa can be useful for taxonomic classification. Such studies should consider that any group has three distinct ages [77,78]: (1) the ‘stem age’, representing the divergence of the clade from its extant sister group; (2) the ‘morphological age’, when the ancestor acquires traits that make it definable as a member of the extant group; and (3) the ‘crown age’, indicating the age of the most recent common ancestor of all living representatives of the group. Examples from angiosperms [78] show that the stem age can be twice as long as the crown age, and if the morphological age is close to the crown age, then much of the lineage evolution may occur without a distinct ‘morphological signature’. Thus, only a sufficiently dense sampling of fossils and/or unique features may reveal transitions between superficially dissimilar groups. Although the overall scarcity of fossil mosses might not be conductive to this, the protosphagnalean mosses discussed here may provide a promising example.