Fissidens subgen. Aloma, the scariosus- and bryoides- type of peristome in the light of the phylogenetic tree by Suzuki et al

The scariosus type of peristome and the bryoides peristome were considered by Pursell and Bruggeman-Nannenga to be diagnostic for subgen. Aloma and subgen. Fissidens respectively. Several later authors did not recognize these peristome types nor did they recognize subgen. Aloma. In a recent molecular study, however, subgen. Aloma emerges as a clade of equal rank to subgenus Fissidens. Based on this and on numerous observations of peristomes the scariosus- and bryoides-type peristome are retrieved from oblivion, compared to each other, re-described and illustrated.

and Fleischer (1904) used peristome characters for the classification of Fissidens. They distinguished between taxa with nodose and taxa with spirally ornamented filaments. Allen (1980) published a SEM-study of peristome-types in Fissidens and concluded that in some cases these are correlated with gametophytic groups. He described the scariosusand bryoides-types of peristome using characters of both the exterior and the interior layer of the peristome.
Bruggeman-Nannenga and Berendsen (1990) amended Allen's scariosus-and bryoides-types of peristome based partly on SEM and partly on light microscope observations. Their types are based on features of the exterior peristome layer, recognizing five main types of peristome each with a strong correlation to a gametophytic group. Ishihara and Iwatsuki (1992) studied different features of peristomes in Fissidens and discerned two types of peristome: one not making hygroscopic movements and having nodose filaments, the other making hygroscopic movements and having spirally thickened filaments. The first is found in subgenus Serridium and Pachyfissidens while the second is found in all other infrageneric taxa. Pursell and Bruggeman-Nannenga (2004) published 'refinements to the infrageneric classification of the Fissidentaceae' most amendments being based on the number of exothecial cells around the capsule circumference, peri-stome-and the costal types. The existence of transitional forms led to a reduction of several of Brotherus' sections into Fissidens subg. Aloma (Kindb.) Pursell & Brugg.-Nann. (in the following referred to as subgen. Aloma P&BN). Accordingly, this subgenus is sporophytically homogeneous but gametophytically rather heterogeneous. This classification was followed by Beever (2014). Suzuki and Iwatsuki (2007) published 'new refinements of the infrageneric classification'. They reduced the bryoidesand scariosus-peristome types into one, the fissidens-type. Consequently, they did not recognize subgen. Aloma P&BN which is based mainly on the scariosus type peristome and the presence of only 32 exothecial cells around the capsule. Suzuki et al. (2018) published a molecular study in which subgen. Aloma P&BN is resolved as a clade 'subgen. Fissidens sect. Polyodiopsis + sect. Areofissidens + sect. Aloma + sect. Semilimbidium'. In other words, in addition to sporophytic evidence there is now also molecular support for subgen. Aloma P&BN. This insight makes it necessary to reconsider the peristome types that have been subsumed into the fissidens-type of peristome.

Material and methods
A study of peristomes is complicated for more than one reason. Many collections lack sporophytes. When present the peristomes are frequently damaged. Some peristomes are thin, transparent and poor in contrast. Moreover, many peristomes are hygroscopic and strongly incurved at the bifurcation when wet, making observations of this area difficult. Peristomes are best studied when the operculum has been freshly detached before they become incurved.
For a study of the peristome-types it is not only necessary to understand the general structure of peristomes, it is also prerequisite to understand the peristome types and to know what one is looking for. Therefore, the scariosus-and bryoides-type are here elaborately described and illustrated. Oil-immersion and a lot of focusing are necessary assets in distinguishing the types.

General structure of peristomes
Peristomes are complicated structures and often hard to analyze with the light-microscope. Basic to the study of peristomes is a good understanding of their structure.
Fissidentaceae have haplolepidous peristomes i.e. the peristomes are composed of a single row of 16 teeth.
Each tooth has a basal undivided part. At the bifurcation it is split into two filaments.
Each tooth is composed of two layers of cells (Plate 1: 1, 1: 4), an exterior (OPL) and an interior layer (IPL). During the formation of the tooth, the walls of the OPL-and IPLcells become partly thickened and partly disappear, the cells become 'roofless'.
Each tooth is built up by a single column of OPL cells (Plate 2: 4) and two columns of half IPL-cells (the other halfcell is part of the adjoining tooth). The borderline between these half cells is a characteristic zigzag-line (Plate 2: 5).
The most important features of a tooth are the horizontal walls, the trabeculae and the 'floor' between, called the lamella. Successive trabeculae are sometimes connected by vertical walls. In the bryoides-type the trabeculae of the undivided part are distinct and protruding (Plate 2: 3). In the scariosus-type of peristome the trabeculae (Plate 1: 5-6) do not or hardly protrude and often have an incrassate margin.
All descriptions in this paper are based on light microscopic observations of the OPL of the undivided part and bifurcation.

Results and discussion
Subgenus Aloma P&BN is composed of several not sharply distinct gametophytic groups, sections and subgenera in Brotherus (1924). Because of the occurrence of transitional species Pursell and Bruggeman-Nannenga (2004) united these taxa into subgen. Aloma P& BN. Sporophytically, this subgenus is well defined by two characters, viz. ± 32 exothecial cells around the capsule circumference and a unique peristome-type, the scariosus-peristome. Suzuki and Iwatsuki (2007) reduced the scariosus-and bryoides-peristome types into one, the fissidens-type peristome. In doing so they undermined the morphological basis for subgen. Aloma P&BN. Consequently, they did not recognize this subgenus.
Numerous observations made by me since the 1990-SEM study (Bruggeman-Nannenga and Berendsen 1990) also confirm the bryoides-and scariosus-type of peristomes to be clearly distinct and characteristic for subgen. Fissidens and Aloma respectively, and that ± 32 exothecial cells around the capsule is characteristic of subgen. Aloma. Only a few exceptions were seen. For instance, the subgen. The above makes it clear the scariosus-and bryoides-type of peristome are valuable characters to distinguish subgen. Fissidens from subgen. Aloma P&BN. It has thus become imperative to retrieve these peristome-types from oblivion.
Teeth of both types are hygroscopic and have 'spirally' ornamented filaments.
scariosus-type of peristome (Plate 1: 6) Undivided part -OPL (Plate 1: 4-6). The trabeculae do not or hardly protrude above the lamellar ornamentation. If in doubt, side-viewing the tooth will be helpful. The lamellae often show rows of coarse 'papillae' (Plate 3: 3, 3: 5). In many species the OPL trabeculae of the undivided part have a thickened margin. The bifurcation -OPL. Towards the bifurcation the lamellar ornamentation becomes less conspicuous and the walls surrounding the OPL cells become distinct (Plate 1: 5), encircling the cell completely. Beyond the splitting of the tooth the OPL cells are torn into two halves each surrounded by a wall on three sides. Distal filaments spirally ornamented. The IPL trabeculae of the undivided part of the scariosus type frequently have fimbriae (Plate 1: 7; Allen 1980: Fig. 9, 10).
Plate 1 shows distinctions between the two types. The bryoides-type differs from the scariosus-type by the high protruding OPL-trabeculae of the undivided part (Plate 1: 3). In the scariosus-type these trabeculae do not, or hardly, protrude (Plate 1: 6). Furthermore, the two types differ in the bifurcation: in the scariosus-type the OPL-cells are completely surrounded by walls (Plate 1: 5).
Plate 2 shows examples of bryoides-type peristomes. With the exception of Plate 1: 4 and 1: 7 the figures are from species with typical subgen. Fissidens gametophytes, viz. limbate on all laminae and small to medium sized, smooth laminal cells. Fissidens gladiolus (Plate 2: 4) is a species that combines an unmistakable bryoides-type peristome and more than 40 files of exothecial cells with a reduced gametophyte. Nanobryum dummeri Dixon, the type species of Nanobryum, is a synonym of F. gladiolus. It is considered to belong in subgen. Fissidens .
Plate 3. Ecostate species (Polypodiopsis), species with large cells (Areofissidens) and F. zollingeri Mont. Gametophytically this species could be placed in subgenus Fissidens. However, the sporophyte having ± 32 exothecial cells around the capsule and a scariosus-type peristome, clearly indicates subg. Aloma P&BN. Moreover, the very large and often inflated cells of the vaginant laminae suggest sect. Areofissidens. This species has not been sequenced.
However, in the phylogenetic tree by Suzuki et al. (2018) some of these species are in clade : sect. Aloma (F. exilis, F. pseudoclosteri and F. takayukii), whereas F. pellucidus is in clade sect. Semilimbidium. Here the scariosus type peristome of both F. exilis and F. pellucidus are illustrated. It will be interesting to see what further molecular studies will teach us about elimbate species. From a nomenclatorial point of view it is important to study the DNA of F. pauperculus, the type species of Aloma.
It is interesting to note that in the phylogenetic tree by Suzuki et al. (2018) subgen. Aloma P&BN is a clade of equal rank to subgenus Fissidens subgen. Fissidens and that several subclades are resolved that represent some of Brotherus' and Suzuki and Iwatsuki's taxa (Table 1). It is clear that when more molecular data from more species becomes available the infrageneric classification of the Fissidentaceae will have to be reconsidered. This will almost certainly lead to new combinations. This, however, is outside the scope of this publication and in my opinion should only be done when significantly more molecular results become available. Tanzania