Reinstatement of the Patagonian moss Ulota glabella Mitt. (Bryophyta, Orthotrichaceae)

In 1842, J.D. Hooker collected a number of mosses on Hermite Island (Cape Horn region). From one of those gatherings, Hooker 141, four species of Ulota have been described: U. luteola, U. fuegiana, U. glabella, and U. eremitensis. The fi rst two species are widely accepted, whereas the identity of the latter two has been recently discussed, and the names are now synonymized under U. fuegiana, the more widely distributed species in the Tierra del Fuego archipelago. Our studies, based on recent collections of Orthotrichaceae from Patagonia, show that specimens diff erent from those of U. fuegiana and agreeing with the protologues of both U. glabella and U. eremitensis are common in Patagonia. Comparisons with type material of all four names demonstrate that the type for U. glabella is in such bad condition that it cannot be used, and an epitype should be selected. In this paper, we comment on the whereabouts of the collection Hooker 141 and the species described from it, discuss the distinct identity of U. glabella and its relationship with U. eremitensis as well as its diff erentiation from other species, present a diagnostic description of U. glabella and, fi nally, select an epitype to fi x the application of this name.

From his collection 141 from Hermite Island, Hooker described in collaboration with William Wilson (Hooker 1847) Orthotrichum luteolum Hook.f. & Wilson (≡ Ulota luteola), with more or less crisped leaves and a hairy calyptra. They also indicated a "var. ß", with a naked calyptra, an uncommon feature for the genus Orthotrichum Hedw., and certainly rare in Ulota D.Mohr. Nevertheless, this variety was not validly published since it was not given a name. Hooker renumbered those specimens as no. 141b for O. luteolum and 141a for the var. ß, and these were distributed to Kew (now at BM), E, NY, and PC. Mitten (1860) studied the duplicate now held at NY and recognized four morphotypes, which led him to take the following actions: 1) rename Orthotrichum luteolum as Ulota fulvella Mitt. nom. inval.; 2) validate Hooker & Wilson's var. ß creating U. glabella; and 3) describe two new species, U. eremitensis based on a morphotype with a naked calyptra and shorter capsule than U. glabella, and U. fuegiana based on another morphotype with a hairy calyptra. The specimens seen by Mitten (NY) were incomplete and he could only study intact peristomes in U. eremitensis (poorly described, but drawn on a label with the holotype of this species, NY[00737687]!), and he was unable to describe the spores of any of the species. Thus, his descriptions were mainly based on gametophytic characters (leaf margin crenulation, number of rows of differentiated basal cells, calyptra hairiness, but also the relative size of capsule and seta), lacking the very important discriminatory characters of the peristome. Mitten (1860) supplemented the meagre description of Ulota luteola (≡ U. fulvella) provided by Hooker and Wilson, establishing that this is the only species of the group of taxa from Hooker's collections 141a and 141b with a thin band of differentiated basal marginal cells, only one to three cells wide, whereas the other species have broad bands of more than five cells. Another important character is the entire, noncrenulate leaf margins, whereas the exostome is vaguely described as formed by eight pairs of teeth split at the apices, a character that Mitten deemed to be of importance. Concerning the other species that he described, U. fuegiana is similar to U. luteola, being discriminated from the latter by its less densely hairy calyptra, the leaves with margins irregularly eroded towards the apex, and the many series of pellucid cells at the basal margins. Ulota glabella was characterized by its naked calyptrae, leaves not crisped, and the finely crenulate ("margine minute crenulato") leaf margins. Mitten did not find peristomes in the specimens he studied (NY[00737690]!, NY[00737691]!, Fig. 1A-D) and therefore nothing is said about this taxonomically important structure. Ulota eremitensis was very similar to U. glabella, since it also has naked calyptrae and finely crenulate margins, and the differences were related to the growth-form of the colonies -subpulvinate in U. glabella and in little tufts in U. eremitensis -and the seta : capsule size relationship. Mitten described the peristome of U. eremitensis with an exostome of eight pairs of teeth divided at the apices, and an endostome of eight very thin segments ("angustis capillaribus"), almost as long as the teeth. Mitten drew this peristome in the holotype of this species (NY[00737687]!). The lack of detailed peristome descriptions sometimes makes the interpretation of Mitten's taxa challenging. See Table 1 for the main differences between these four species according to Mitten (1860).
The revisionary work by Malta (1927) is of particular interest, as he assessed the full scope of the genus in South America through the study of almost all the available material at the time, including original specimens at BM and PC that Mitten could not study. Malta considered U. luteola and U. fuegiana as different, well-defined species. Regarding U. eremitensis and U. glabella, Malta did not find any differences between them except for the size and color of the capsules, so he considered them as conspecific, and synonymized U. eremitensis under U. glabella.
Malta highlighted three diagnostic characters in his description of U. glabella, but he used several others in the differentiation of the species. Two of them were previously noted by Mitten, namely the naked calyptrae and the finely serrate leaf margins ("Rändern fein gesägten"), which can be interpreted as the finely crenate margins of Mitten's work. The third character is a new feature not previously described, the sheathing perichaetial leaves with a broad, high base that abruptly tapers to a blunt tip. Other important characters are the papillose costa, usually an overlooked feature, the exostome of eight teeth splitting into 16, papillose and with lines in the upper part, and the endostome of eight or 16 segments somewhat knotted and papillose. It is not possible to safely say on which specimen this peristome description is based, as all the duplicates revised by Malta have lost this structure. One possibility is that it was based on specimens of U. eremitensis, as this is the only one that seems to have had perfect peristomes at that time. Nevertheless, it is also possible that Malta did not use original specimens for his description, but rather another intriguing specimen collected by Dusén (S) from Westhoff Island cited in his study. This particular specimen will be discussed later.
The recent revision by Wang & Jia (2016) reduced the number of accepted species. They concurred with Malta's opinion about the conspecificity of Ulota glabella and U. eremitensis, but considered that "U. glabella is virtually identical with U. fuegiana in terms of leaf crisping, peristome structure, stomata position and calyptra hairiness" and consequently they listed U. glabella and U. eremitensis under U. fuegiana. These authors did not find serrate leaf margins and considered that U. fuegiana can have naked calyptrae. In support of their opinion on the variability of the hairiness of the calyptra in U. glabella, they referred to the very sparsely hairy calyptra of the dubious specimen from Westhoff Island cited by Malta (1927).
In summary, currently, there are two interpretations of the two species with naked calyptrae described by Mitten. The first one by Malta (1927), who considers that a single species is involved, different from others in the genus, and the second one by Wang & Jia (2016), who considers that the forms with naked calyptrae are simply part of the variability in Ulota fuegiana.

Material and methods
This work is based on the study of the original collections by Hooker (BM, duplicates at E, NY and PC), as well as a significant number of recent collections (MAUAM, NY, VAL), including some from the original locality of U. glabella, Hermite Island in the Cape Horn region. In total, 124 modern specimens were studied ( Original specimens were studied under a stereo microscope to avoid damage. Modern collections were also prepared on permanent slides and fully studied under a compound microscope.

The South American species of Ulota D.Mohr. with naked calyptrae
Based on the revision by Wang & Jia (2016), there would only be one species of Ulota in South America with constantly glabrous calyptrae, U. macrocalycina. In addition to this feature, this species has other relevant differentiating characters that strongly support its discrimination: 1) the prostrate growth with upright secondary branches, forming dense tufts; 2) the very short leaves (1.3-1.5 mm long), erect, straight or slightly undulate with blunt to obtuse apices; 3) the perichaetial leaves differentiated, longer than the vegetative ones; 4) the stomata located at urn base; 5) the endostome of eight completely smooth and hyaline segments; and 6) the verrucose spores, 30-35 µm wide. The growth form and the position and size of the leaves are uncommon characters sufficient to safely distinguish this characteristic species with a stereoscopic microscope or even a hand-lens. Nevertheless, at least one other species with naked calyptrae consistently appears in Patagonia and, especially, in Tierra del Fuego. This plant is clearly different from Ulota macrocalycina and U. fuegiana, even considering the variability of U. fuegiana to include naked calyptras, as supported by Wang & Jia (2016). The more outstanding characters (Figs 3-4) for its discrimination are: 1) leaves undulate to somewhat curved with crenulate-serrate margins and papillose costa; 2) perichaetial leaves differentiated; 3) naked calyptra with a dark beak; 4) stomata present throughout the capsule; 5) brown-orange to pale orange exostome, formed by eight pairs of teeth with a tendency to split, but usually remaining joined at the tips; and 6) brown-orange to pale orange endostome of eight segments, occasionally with remains of incomplete intermediates, ornamented on both sides with irregular small and compact tufts of very high papillae.

Reinterpretation of Ulota glabella Mitt.
The morphological similarities between the second South American species of Ulota with naked calyptrae and the protologue of U. glabella (Mitten 1860) or the more precise description and drawings by Malta (1927) suggest that they are the same species. On the other hand, the analyses of abundant modern specimens show that U. fuegiana and U. glabella are actually two distinct species that are well distinguished by a set of characters (Table 2, Fig. 4), of which the most important are the hairiness of the calyptrae, the color of the peristomes, the endostome ornamentation, the highly developed exostomial trabeculae in U. fuegiana, and the papillosity of the leaf margin and costa in U. glabella. Of these, the characters with the greatest discriminatory value are the peristomial ones.
The holotype of U. eremitensis (NY[00737687]!) preserves remains of the peristome, including basal portions of endostome segments. Although the state of the material discourages the use of microscopic preparations to analyze these structures, the stereoscopic microscope shows that the segments are colored in the same way as those of U. glabella. The main difference between U. glabella and U. eremitensis as defined by Mitten (1860) was the size relationship between capsule and seta (1 : 2 in the former and 1 : 3 in the latter). In our study of recent specimens, we found a continuous gradation between both extremes, supporting Malta's view of a single variable species. Nevertheless, some differences in the peristome coloration and ornamentation are found along this size gradient and it cannot be discarded that U. eremitensis may be a different species. More intensive studies with an integrative taxonomic approach are needed to answer this issue. Until these studies can be completed, we consider U. glabella and U. eremitensis to belong to the same taxon.

Description
Plants growing in small cushions. Leaves non crisped, undulate to somewhat curved, (2.0-)2.5-3.15 × 0.4-0.6 mm; leaf margin crenulate-serrate, more strongly towards apex, with protruding papillae. Costa dorsal cells papillose. Lamina cells with 1(-2) single or ramified papillae. Differentiated basal marginal cells in more than five rows. Perichaetial leaves differentiated, with sheathing base, abruptly narrowed towards lamina and with blunt to obtuse apex. Vaginula naked, 0.4-0.7 mm long. Calyptra naked with dark beak, 1.5-1.9 mm long. Seta (2.75-)4.0-5.7 mm long. Capsule fusiform when dry and full of spores, cylindrical when dry and empty, somewhat wider below its middle, 1.35-2.2 mm long including neck; ribs moderately marked, concolorous with rest of exothecium. Stomata scattered from urn base and neck to near capsule mouth. Exothecial bands thin, scarcely differentiated, formed by 2-3 cell rows with walls not strongly thickened. Operculum plane-convex, with broad reddish basal rim. Peristome double. Exostome of 8 pairs of teeth with tendency to split, but usually remaining joined at tips, brown-orange, sometimes paler, 230-325(-356) µm long; external side (outer peristome layer = OPL) ornamented in basal section by reticule with tall papillae, in upper ⅓-¼ with longitudinal crests and aligned papillae, sometimes uniformly papillose; inner side (primary peristome layer = PPL) with scattered tall papillae, becoming denser at tips and sometimes longitudinally aligned at base. Endostome of 8 segments, occasionally with remains of incomplete intermediates, brown-orange, sometimes paler, filiform and fragile, sometimes wider and more persistent; uni-biseriate at base, uniseriate above, almost as long as teeth; external side (PPL) smooth in basal ⅔, in the upper ⅓ smooth but ornamented with clusters of high papillae, with appearance of tufts, which can protrude laterally, much less frequently with lines; inner surface (inner peristome layer = IPL) with base ornamented by low reticulum with coralloid aspect or by papillae, in upper parts smooth and with papillae in clusters as in PPL, rarely more or less homogeneously papillose with these clusters not or hardly developed; trabeculae differentiated. Spores finely papillose, 20-30 µm in diameter.

Discussion
The assimilation of U. glabella with U. fuegiana by Wang & Jia (2016) was based on the interpretation that the calyptrae of the latter vary from very hairy to naked and on the lack of serrate margins in the specimens they studied. The two other cited characters, the stomata position and the crisping of the leaves, are more or less variable within this species and, in the case of leaf position, sometimes difficult to describe, which results in the same curvature of leaves being described differently by different authors. The specimens we identified as U. glabella have leaves with crenulate-serrate or crenulatepapillose margins (Fig. 3E), exactly like those described by Mitten (1860) and drawn by Malta (1927).
Concerning the development of hairs on the calyptra, although this feature can show variability both in the genus Ulota (Caparrós et al. 2011;Garilleti et al. 2020) and other Orthotricheae , it rarely displays a complete gradation from hairy to naked calyptrae (Caparrós et al. 2014). Among the numerous specimens of U. fuegiana that we have studied, none have been found with naked calyptrae; all of them have short, thick and papillose hairs variably abundant. The identity of the specimen from Westhoff Island with some hairs cited by Malta (1927) and used by Wang & Jia (2016) in support of their position is doubtful. In addition to the presence of hairy calyptrae, the spores of this specimen (Malta 1927) are noticeably large (40-50 μm), far greater in size from those of U. fuegiana, which measure, according to our observations, 25-30(-34) μm. Such large unicellular spores are found in South America only in U. magellanica (Mont.) A.Jaeger and U. pusilla. The discriminant characters for U. magellanica are discussed in Garilleti et al. (2015). Ulota pusilla is morphologically close to U. glabella, as it has crenate-serrate leaf margins, an orange-tinged peristome and a similar shape of the capsule. Its endostome has 8 + n segments, with the intermediates sometimes well-developed. It is not impossible that the sample from Westhoff Island, used by Malta to describe the spores and peristomes of U. glabella, may in fact correspond to U. pusilla. Malta himself pointed out the great similarity of this specimen with the latter species. Confirmation of the identity of the Werthoff Island specimen has not been possible at the present time, since the Stockholm Herbarium is currently under renovation and access to its collections is closed.

An epitype for Ulota glabella Mitt.
As shown in this paper, U. glabella is a distinct species with clear morphological characters discriminating it from U. fuegiana. Nevertheless, the holotype of U. glabella was already in imperfect condition ( Fig. 1A-D) when this species was published and today all the extant isotypes are similarly damaged (e.g., Fig. 1E). Thus, the discriminatory characters for this species, particularly those differentiating it from U. fuegiana, are lost, suggesting that recent confusion with this species has likely occurred because of the poor condition of the type material, which prevents any critical comparison between the species.
Since all original material is "demonstrably ambiguous and cannot be critically identified for purposes of the precise application of the name to a taxon" (ICN Art. 9.9, Turland et al. 2018), it is necessary to select an epitype with well-preserved differentiating features, mainly those of the peristome, to serve as an interpretative type that fixes the application of this name. We have chosen as epitype a specimen collected by R. Garilleti from Hoste Island, Tierra del Fuego archipelago (Chile), belonging to the same biogeographic region as Cape Horn. Details of this specimen as well as the holotype and isotypes supported by it are in the description of the species. It is completed with the details in Figs 3 and 4B, D.