Haplohymenium sieboldii (Dozy & Molk.) Dozy & Molk., Musc. Frond. Ined. Archip. Ind.

Haplohymenium sieboldii (Dozy & Molk.) Dozy & Molk., Musc. Frond. Ined. Archip. Ind. View in CoL , 4: 127, 40. 1846. — Leptohymenium sieboldii Dozy & Molk., Ann. Sci.

Nat., Bot., sér. 3, 2: 310. 1844. — Anomodon sieboldii (Dozy & Molk.) Granzow, Contr. Univ. Michigan Herb. , 21: 243. 1997. Figs. 1, 2, 5.

Plants small, in loose patches, yellowish- to brownish-green, dull. Stems creeping, 1–3 cm long, irregularly branched, central strand absent. Leaves appressed when dry, erect-spreading when wet; stem leaves 0.5–0.9× 0.2– 0.4 mm, from ovate base gradually narrowed into lanceolate acumen, acute, occasionally with 1–3 transparent cells at apex, at lower portion of shoot often fragile; margins flat, entire or weakly crenulate; costa reaching 1/2–3/4 the leaf length, smooth or with few papillae on dorsal side; leaf lamina unistratose; upper and median lamina cells 9–13 µm in diameter, round or round-polygonal, not transparent, thin-walled, with several low branched papillae above lumen, basal marginal cells transversely ovate, to 16 µm wide, basal cells near costa elliptic, transparent, smooth, to 30 µm long; branch leaves ovate-lanceolate, gradually narrowed to acute or subobtuse apex, slightly smaller than stem leaves. Dioicous. Sporophytes unknown from Russian collections. [Perichaetial leaves with oblong base and narrowly lanceolate acumen, to 1.3 mm long. Setae to 5 mm long, yellowish-brown. Capsules erect, symmetric, ovate. Operculum conic. Calyptra with sparse, long hairs, papillose at apex.]

Diagnostic characters of H. sieboldii include ovate-lanceolate, gradually tapered, acute leaves, costa longer the half leaf length, and multipapillose lamina cells. This species is closely related to H. tristie . The latter species differs from it in leaf shape, from ovate base abruptly narrowed into lanceolate or oblong acumen, often rounded at apex and having small apiculus, more fragile, and with shorter costa. Watanabe (1972) points the presence of stem central strand in H. triste and its absence in H. sieboldii ; however, in specimens of H. triste from Asian Russia the stem central strands is very weak or almost absent.

Specimens examined: RUSSIA: Zabaikalsky Territory: Sokhondinsky State Reserve: vicinity Agutsa River , 49°38’N, 111°27’E, ~ 1170 m alt., 17.VII.2010, Czernyadjeva 24-10 ( LE) GoogleMaps , ibidem, Czernyadjeva 28-10 ( LE) . Primorsky Territory: Murav- ’ev Amursky Peninsula , 12 VIII 1929, Transhel s.n. ( LE) ; Sikhote- Alin, Krinichnaya Mountain , 42°59’41’’N, 132°28’58’’E, IX.1990, Czernyadjeva s.n. ( LE) GoogleMaps . Shkotovsky District, vicinity Novokhotunichi Settlement , 3.VIII.1927, Transhel s.n. ( LE) ; vicinity of Vladivostok Sity , 18.VII.1971, Bondartseva s.n. ( LE) .

JAPAN: exsiccata ser. 2 (1948), # 88, Kyushu , X.1946, Hattori s.n. ( LE) ; Buttuji , Hiroshima Pref., XII 1963, Ando s.n. ( LE) ; Honshu , 6 X 1980, Mizutani 6333 ( LE) .

Despite of apparent difference, the species can be identified for sure only when the collection is large enough. The variation in Haplohymenium triste is great and in many collections some less developed shoots might have non-fragile and more gradually tapered leaves. Therefore we presented in Fig.2–3 View Fig View Fig numerous leaves to illustrate the character stressed by Granzow-de-la-Cedra (1997): costa is extended farer than leaf constriction in H. sieboldii and only rarely so in H. triste . In addition, smaller papillae is a useful character of H. sieboldii . Despite of variation of H. triste , its papillae are invariably large, especially on dorsal leaf surface.

As we already mentioned, DNA indicated very close relationships of Haplohymenium triste and H. sieboldii , contrary to H. pseudotriste , which is genetically well differentiated ( Ignatov et al., 2019). The latter species is more southern and so far unknown in Russia, although its findings are possible. This species is the smallest in the genus, with hardly fragile leaves, looking as a strongly minitaturized Anomodon minor .

Papillae of H. pseudotriste and H. triste are rather similar, however their shape depends not only on species, but also of method of preparation: material from the same collection ( Fig. 5A, B View Fig ) looks very different un- der SEM after different specimen preparation. Cross section illustrates that the connection between neighboring cells in Haplohymenium is much less than cell height, thus a considerable portion of the cell is above the level of their joint and can be considered as a large stumpshaped mammilla, further bearing papillae on its upper surface ( Figs. 5 View Fig , 6 View Fig ).

However even more complicated is the case of H. longinerve , which may bring light to papillae development and limits of variation in Anomodontaceae as a whole.

* * *

Haplohymenium longinerve is often separated in identification keys at first couplet by the unipapillose laminal cells. The light microscopy usually allows the high papillae to be seen in distal leaf half. They are especially easily observable in upper third of leaf at back of convex part of acumina ( Fig. 7A View Fig ), “in profile”. Closer to leaf margin ( Fig. 7C View Fig ) and at leaf middle ( Fig. 7D View Fig ) most cells are already multipapillose. The main problem appears with specimens where most or even all cells are multipapillose ( Fig. 7G View Fig and 9A,B View Fig ), which in the case of plants growing in relatively harsh environments, e.g. rock outcrops, but open and sunny, or exposed to winds. Also in plants with well developed high papillae (cf. Fig. 8A View Fig ) some smaller leaves have only multipapillose cells: pictures in Fig. 7G, H View Fig are taken from the same shoot as shown in Figs.7E, F View Fig and 8E, F View Fig .

No unipapillose cells were seen, in e.g. collections from Kurils Islands by Bakalin, which caused erroneous identification as H. triste ( Bakalin et al., 2009) . Its multipapillose cells are shown in Fig. 9A, B View Fig . Molecular phylogenetic analysis of Ignatov et al. (2019), however, demonstrated that it is identical by all DNA markers with other specimens of H. longinerve , including specimen shown in Figs. 7 View Fig , 8 View Fig , 9C–F View Fig , and 10 View Fig .

The large papillae of H. longinerve are similar to those in Anomodon abbreviatus and A. solovjovii ; such papillosity pattern seemingly needs a special term, so we suggest to call it pseudounipapillose. The high conic papillae dominate in places of maximal development, howev- er small round papillae are also present, albeit not well seen in light microscope.

Close to the leaf base papillae on one cell surface are not so unequal, and therefore cells look multipapillose. In some places within transitional zone the variation in papillae structure characterize the type of their development ( Fig. 8C View Fig ). The stump-like mammillae are well seen here and their surface raises in place, forming various type of projections, where one of those may take over others, or not. Such views suggest that the cell wall in this species is plastic and delicate. The structure of papillae is obviously heterogeneous: autofluorescence at higher magnification shows difference in upper part of these high papillae from its median part ( Fig. 8E View Fig ).

Our attempt to take SEM picture of high papillae of H. longinerve (simple mode without special preparation, just coated with gold) brought unexpected results. Stiff plants from the Kuril Islands, where all cells look pluripapillose, resulted in “normal” view ( Fig. 9A, B View Fig ), rather comparable with views of wet plants under CLSM and also comparable with similarly taken pictures for other species ( Figs. 3 View Fig , 4 View Fig , 5 View Fig , 11 View Fig ), whereas in more delicate specimens shown in Figs. 7 View Fig & 8 View Fig the papillae appear terribly damaged by vacuum. Since the preparation was the same and plants were lying side by side during the same session of observation, this difference requires discussion. It seems obvious that cell walls in plants with high conic papillae are extremely delicate and vacuum makes them to collapse, either transforming cell lumina to pit, or leaving at its surface crumplings which we present here with a great hesitation. Few shoots examined this way have no one high papilla in places where they are commonly seen in wet plants in ordinary slide under light microscope and CLSM.

SEM observation of plants fixed in glutaraldehyde and osmium tetroxide, dehydrated in ethanol–acetone series and dried under critical point by the standard protocol also did not succeed to show papillae as large as they are in wet state. Wet papillae are up to 12 µm high ( Fig. 8A View Fig ), while the largest ones in SEM images do not exceed 5 µm ( Fig. 10D View Fig ). The considerable shrinking of large papillae in H. longinerve can be seen even in water slide after its complete drying: photos in Fig. 7A and 7B View Fig provide such an example.

Assuming papillae pattern as unreliable for identification, we consider the base character to be the long, percurrent to excurrent costa, at least in some leaves. The leaf tip, a continuation of costa is often composed of large smooth cells, occasionally with crown papillae at the tip of apical cell ( Fig. 10 View Fig ). This pattern occurs also in H. flagelliforme . The latter species differs in invariably multipapillose cells, and long teeth near leaf apex,which often terminate with crown papillae. Molecular phylogenetic analysis ( Ignatov et al., 2019) placed H. flagelliforme in the basal position in the Haplohymenium clade.