This is a distributable version of the paper with the same pagination. The original published paper is available on request from the authors or from The Bryologist journal.

Pseudosymblepharis and species () new to northwestern North America, and refutation of molecular Chionoloma

Richard H. Zander1 and Patricia M. Eckel

Missouri Botanical Garden, 4344 Shaw Blvd., St. Louis, MO 63110 U.S.A.

ABSTRACT. Three species of Pottiaceae (Bryophyta) of limited distribution worldwide are reported for hyperoceanic northwestern North America. The Asian Pseudosymblepharis angustata is a genus and species new to North America from Alaska. Oxystegus daldinianus, previously known from Europe and the Southern Appalachians in the U.S.A., is reported as new to Canada from British Columbia. A new species is described in Oxystegus from British Columbia. These genera continue to be recognized here instead of the broad molecular systematics concept of Chionoloma in part because of lack of resolution of molecular cladograms reported by a recent study on molecular races in the Pottiaceae.

KEYWORDS. Alaska, British Columbia, dissilient genera, phylogenetics, macroevolutionary systematics, molecular races.

The classification of Pottiaceae subfamily Chiono- reason to abandon them except that they contradict lomoideae R.H.Zander and Trichostomoideae strict phylogenetic monophyly, a stricture that is (Bruch & Schimp.) Limpr. has recently been much arbitrary and enforced simply because use of shared inturmoil(e.g.,Alonsoetal.2018a,b;Kockinger¨ et traits alone in cladistics eliminates evaluation of al. 2010; Werner et al. 2005). In particular, Sollman serial evolution, i.e., taxon to taxon. Primary is the (2000, 2005) reduced 53 species in Chionoloma fact that cladistics is actually hierarchical cluster Dixon,Oxystegus(Limpr.) Hilp.,Pseudosymblepha- analysis using trait changes in a non-ultrametric ris Broth.,TrichostomumHedw., and other genera tree. Sets of traits of two or more species are split to Pseudosymblepharis bombayensis (Mull.¨ Hal.) into pairs and these into pairs again, and so on. This P.Sollman. This in our opinion was a massive loss is not a model of evolution, which should try to of important evolutionary distinctions. The prob- explain at least species to species descent with lem was corrected to a considerable extent by modification. Molecular systematics has additional Alonso et al. (2018a), who re-established several problems (Zander 2013: 58), mainly a lack of distinctive species, but recognized most of these in resolution of evolutionary differences, summarized Chionoloma. in the Discussion. New discoveries are here reported Alonso et al. (2016) merged the genera Oxy- following classical generic concepts. stegus and Pseudosymblepharis with Chionoloma based on putative phylogenetic monophyly in an MATERIALS AND METHODS analysis of nuclear ITS and plastid markers atpB- Herbarium specimens of (Mull.¨ Hal.) rbcL, trnG and trnL-F molecular sequences. They Limpr. (Pottiaceae) associated with the studies of also evaluated traits used by Zander (1993a) as the second author (Eckel 2007, 2010) that were on generic distinctions that matched classical classifi- loan from UBC were restudied in light of the recent cations, but found them less than convincing. We discovery of Asian species of Pottiaceae (Zander have re-examined these distinctions and find no 2017a, 2019a) in the North American hyperoceanic 1 Corresponding author’s e-mail: northwest. Reference slides from previous studies of [email protected] Tortella (Eckel 1990, 1998, 2007a, 2010; Hedenas¨ & DOI: 10.1639/0007-2745-122.4.568 Eckel 2011) and related genera (Eckel 2000a,b,

The Bryologist 122(4), pp. 568–577 Published online: October 21 2019 Copyright Ó2019 by The American Bryological and Lichenological Society, Inc. Zander & Eckel: Pseudosymblepharis and Oxystegus 569

2007b; Zander 1993b, 2017b) plus extensive material Etymology. The epithet is a Latinization of in MO were used to re-evaluate species distinctions. Greek maragna, whip, and phyllon, leaf. Among this material were misidentified specimens Discussion. Oxystegus maragniphyllus is closely of Oxystegus and Pseudosymblepharis. related to O. tenuirostris (Hook. & Taylor) A.J.E.Sm. by the notched and fragile leaf limb, and the broadened leaf base, and stem with central strand. It differs from O. tenuirostris by its stiff, very narrow Oxystegus maragniphyllus R.H.Zander & Eckel, sp. leaf limb, the base commonly fully sheathing the nov. Fig. 1 stem (best seen in stripped leaves with overlapping Caulis funiculo centrali misero praeditus. Folia basal margins), and very narrow leaf apex usually fragilia, limbo inciso, saepe abrupto, basi dilatata with a strongly excurrent costa of 15–30 smooth sed humeris distinctis carentia. Costa ut mucro rhomboidal cells in the very long-linear perichaetial cylindrica, acuta, basi 70–110 lm lata, filo leaves, and 7–15 in cauline leaves. It is known from adaxiali stereidarum e cellulis paucioribus quam only one specimen, but may turn up in other abaxiali constante. Cellulae laminales distalliter hyperoceanic areas, particularly in Asia. Oxystegus 10–13 lm latae, secus margines foliares indiscrete maragniphyllus is similar to O. hibernicus (Mitt.) ascendentes. Hilp. but the former differs in the narrower leaf limb, longer and broader leaf base, and very fragile TYPE: CANADA. BRITISH COLUMBIA: West Vancouver leaves. Crum & Anderson (1958) discuss a very Island, Ocluelet, limestone outcrop in forest, narrow-leaved form of O. tenuirostris (as Trichosto- May 7, 1969, W. B. Schofield 89947 (UBC: mum cylindricum Hedw.) in the Southern Appala- holotype). chians, but their text and illustration concerns a Description. in a dense turf, yellow- population with non-fragile leaf limbs and narrow green, 2–2.5 cm tall. Stem with weak hyalodermis, leaf bases, not broadly ovate and sheathing as is the weakly thickened sclerodermis, central cells all of case with O. maragniphyllus. Although this species is similar size, and central strand weak or rarely recognized from only one specimen, some traits are strong in same , this often red just below qualitative, not quantitative; the range of variation archegonia. Axillary hairs of ca. 8–10 cells, the in all related species is such that there is no overlap distal cells hyaline and basal 4–5 thick-walled or of traits; and distinctiveness is greater between the brownish. Leaves linear-lanceolate from an ovate or new species and related species than between each of elliptical base, fragile and often broken off at mid- those related species. limb, when dry inwardly hooked, incurved-twisted Pseudosymblepharis new to North America and tubulose, when wet spreading at ca. 458 from from Alaska (Fig. 2). The discovery of the Asian sheathing base, apex narrow, gradually merging Pseudosymblepharis angustata (Mitt.) Hilp. in Alaska into a long-pointed, sharp mucro, mid-limb often is not unexpected given that Alaska harbors isolated notched or cracked at margins; costa with adaxial populations of the Asian Pottiaceae species of stereid band smaller than the abaxial; leaf base similar morphology: sikkimense Aziz much broadened, sheathing the stem, shoulders & Vohra (Zander 2017a) and hornschuchi- weakorabsent,notdecurrentatinsertion.Leaf ana (Hook.) Lindb. ex Limpr. (Zander 1978, 2007), cells at mid-leaf quadrate to short-rectangular, 10– both differing in being pleurocarpous. Pseudosym- 13 lmwide,papillaebifid,solid,crowded;basal blepharis angustata is described and illustrated by cells not sharply distinguished from the distal but Saito (1975) and by Li et al. (2001), and an grading through thick-walled, smooth rectangular illustration of the Alaskan plant is provided here. cells, hyaline and thin-walled, ca. 13 lmwide,3– It has leaves that are less strongly squarrose or 5:1, rising up the margins in a narrow line. shouldered than the Latin American P. schimperiana Perichaetial leaves gradually longer and narrower (Paris) H.A.Crum. The European Oxystegus hiber- than the cauline, with longer cusp and more nicus is similar to P. angustata in the large, very elongated sheathing base. Sporophytes and peri- long-lanceolate narrowly acuminate leaves with goniate plants not seen. widened leaf base and plane margins, distal marginal 570 The Bryologist 122(4): 2019

Figure 1. Oxystegus maragniphyllus. A. Habit, moist. B. Habit, dry. C–E. Leaves. F. Stem section. G. Leaf section at mid-leaf. H. Section at leaf base. I. Distal lamina with crack. J. Leaf base. K. Mucro of cauline leaf. L. Mucro of perichaetial leaf. Scale bars: a ¼ 2 mm for A–B, 60 lm for F; b ¼ 60 lm for G–L; c ¼ 50 lm for C–D. Zander & Eckel: Pseudosymblepharis and Oxystegus 571

Figure 2. Pseudosymblepharis angustata. A. Habit, moist. B. Habit, dry. C–E. Leaves. F. Section of stem. G. Leaf section near apex. H. Section at mid-leaf. I. Section at leaf base. J. Basal laminal cells. K. Cells at leaf shoulder. L. Leaf mucro. Scale bars: a ¼ 1 mm for C–E; b ¼ 2 mm for A–B, 90 lm for F; c ¼ 50 lm for G–L. 572 The Bryologist 122(4): 2019 cells with oval lumens with walls weakly projecting identified by M. Alonso, M. J. Cano and J. A. as a minute crenulation, and distal marginal cells Jimenez´ (unpubl.). entire (not even distantly denticulate), but O. CANADA. BRITISH COLUMBIA. Vancouver Island, hibernicus differs in that the costal adaxial stereid Indian Arm, Wigwam Creek, cliff crevices, 15 May band is smaller than the abaxial, the basal cells do 1963, W.B. Schofield 20536 (UBC); mountain just S of not rise into the limb, the basal cells are restricted to Harrison Lake, Skeena River, wet cliff, 13 Jun. 1963, below the shoulders of the sheathing base, and there W.B. Schofield 21266 (UBC); Haida Gwaii, NW is a region of thick-walled rectangular cells between Moresby Island, Paul Inlet Rd., ca. 5 km E of Inlet, the thin-walled extreme basal cells and the distal dry cliff, 22 Aug. 1961, W.B. Schofield 15218 (UBC). laminal cells. Distal laminal cells of O. hibernicus are KEY TO OXYSTEGUS AND PSEUDOSYMBLEPHARIS IN NORTH commonly 10–13 lm wide and short-rectangular, AMERICA NORTH OF MEXICO but those of P. angustata are distinctly smaller, 1. Leaves very long acuminate, abruptly broadened and strongly (6–)8–10 lm wide, quadrate or uncommonly short- sheathing the stem below, apex narrowly acute...... 2 1. Leaves lanceolate to ligulate, not or little sheathing the stem rectangular. below, apex usually blunt, sometimes narrowly so ...... 4 The following citation is for Pseudosymblepharis 2. Costa percurrent or shortly excurrent as a flattened or grooved angustata: U.S.A. ALASKA: Circle Quad., vicinity of apiculus...... Oxystegus hibernicus Eagle Summit, Steese Highway, mile 105–108, 2. Costa excurrent as a cylindric, sharp mucro...... 3 658290N, 1458250W, 1070 m, 18 Aug. 1972, B. M. 3. Leaves not fragile, leaf base with fairly distinct shoulders, basal cells well differentiated from the distal, rising along the margins UBC Murray 72-72 ( ). An ample specimen of Oxy- in 2–3 tapering rows above the sheathing leaf base; costa at base stegus hibernicus from British Columbia was includ- 100–150 lm wide, including ca. 8 guide cells near leaf insertion, ed in the return of a loan from MUB, correctly adaxial stereid band with as many cells as the abaxial; distal annotated by M. Alonso, M. J. Cano and J. A. laminal cells (6–)8–10 lm wide.... Pseudosymblepharis angustata Jimenez,´ but placed in the genus Chionoloma.Itis 3. Leaves fragile and often broken near midleaf, leaf base lacking distinct shoulders, basal cells gradually differentiated from the here placed in the key below, and will be published distal, rising indistinctly above leaf base marginally; costa at base by these authors at a later date. 70–110 lm wide, including 4–6 guide cells near insertion, adaxial Oxystegus daldinianus new to Canada from stereid band with fewer cells than the abaxial; distal laminal cells l British Columbia. The European mountain species 10–13 m wide ...... Oxystegus maragniphyllus 4. Leaves sharply reflexed above a short, broadened base, broadly Oxystegus daldinianus (De Not.) Kockinger,¨ lanceolate to long-ligulate, apex broadly acute to abruptly O.Werner & Ros was reported for eastern North narrowing and broadly acute or rounded acute; base either little America from Southern Appalachians by Zander wider or more usually narrower than limb, less than 1/4 leaf (2017b). This species can at times be difficult to length, basal cells not extending up the margins; leaf margins distinguish from O. tenuirostris but, following usually with a border of smooth or at least cells with fewer and lower papillae, toothed or crenulate, often sharply so; basal cells Kockinger¨ et al. (2010), the abaxial distal costa is little wider than distal cells, concolorous; cells of stem central smooth and of elongate cells in the former, but is cylinder more than 25, central strand absent...... 5 bumpy in parts by short-rectangular bulging cells in 4. Leaves little reflexed above a long half-sheathing base, short- the latter. In addition, the stem commonly has a lanceolate to narrowly long-lanceolate, apex narrowing to an central strand in O. tenuirostris, the leaf base is high abruptly obtuse point; base usually wider than limb, 1/4 to 1/3 of leaf length, basal cells in most specimens extending up the and the leaf limb is fragile. A stem central strand is margins in a narrow line; leaf margins lacking a border of smooth absent in O. daldinianus, the leaf base is short- cells, lacking teeth; basal cells usually much wider than distal sheathing, and the limb is not fragile, often with cells, hyaline; cells of stem central cylinder less than 20, central denticulate distal margins. The discovery in British strand in most specimens present ...... 6 Columbia of three populations in hyperoceanic 5. Leaves distantly dentate in distal 1/2–2/3 by clearly differentiated sharp teeth, apex broadly acute, marginal cells in 2–3 rows non- areas reveals a major geographic disjunction, and papillose and with thicker cell walls in distal 1/2–2/3 ...... one might predict it may be found in montane Asia ...... Oxystegus recurvifolius under other names. 5. Leaves crenulate by projecting cells, abruptly acute to narrowly As this paper was being put together, a loan was acute, margins bluntly and distantly low-toothed in distal 1/4 by projecting end cells of rows, marginal cells thicker walled and MO MUB returned to from (University of Murcia) with fewer and lower papillae in 1–2 rows in distal 1/4–1/3 with two specimens of O. daldinianus (collections ...... Oxystegus daldinianus not listed here) from British Columbia correctly 6. Dioicous; leaves lanceolate to long-lanceolate, often notched and Zander & Eckel: Pseudosymblepharis and Oxystegus 573

fragile, distal laminal cells 10–13 lm in diameter...... between expected relationships and actual phyloge- ...... Oxystegus tenuirostris netic distance. What is that critical distance? 6. Autoicous; leaves short-lanceolate, not notched and fragile; distal It all depends on the extent of molecular laminal cells 8–10 lm in diameter...... Oxystegus spiralis paraphyly involving molecular races. A molecular DISCUSSION race can consist simply of a DNA sequence different Explanations for the inadequacy of molecular from that of another exemplar of the same species. systematics to generate classifications that well- In published molecular studies (e.g., Shaw 2000, reflect evolution have been detailed in previous 2001; Shaw & Schneider1995), genetic diversity is publications (e.g., Zander 2013: 51ff, 110; 2014; common, but we reference here DNA differences at 2016: 322; 2018: 74ff). Nodes in cladograms do not a level critical for classification. Simple multi- represent taxa, they represent sets of the traits of all furcation of branches all leading to exemplars of taxa distal to the node, excepting reversals. Actual the same species may or may not include molecular relationships that molecular cladogram nodes may races because only shared derived traits are recog- represent are genera, species, or molecular races. nized in cladistics. A species, however, of many The ‘‘phylogenetic species concept’’ as the smallest exemplars coming from a multifurcation that set of OTUs that share an ancestor and a certain set includes at least one node beyond the initial of traits is confounded, as detailed below, by the multifurcation implies that there are at least two multiplex branching and paraphyly of molecular different sequences attributable to that one mor- races. phospecies. That these sequences represent races in The most parsimonious or optimal result of nature is evidenced by separate generation, at times, phylogenetic analysis is that which best imposes data of different species from different sequences of the on a dichotomous tree. Given that the possibly same species. viable sorts of morphological mutations of one Enough studies by molecularly oriented system- ancestral species are limited (phyletic constraint), atists have been done to reveal the numbers of then duplication of certain traits in descendant molecular races within species of bryophytes. A species should be common. Such duplication is study has been conducted by Zander (2019b) of the reflected in extension of true evolutionary multi- number of molecular races per species in the furcations into dichotomous phylogenetic sister Pottiaceae. The data for this study were extracted groups. Only consideration of a descendant’s from several recent publications (Alonso et al. 2016; specialization as expressed in autapomorphies can Cano et al. 2009; Grundmann et al. 2006; Kockinger¨ reveal radiation from a generalist ancestral species. et al. 2010; Kuceraˇ & Ignatov 2015; Kuceraˇ et al. Note that if an extant species has all the traits 2013, 2018; Werner et al. 2005, 2009, 2014). expected of an inferred progenitor, then calling the Molecular races were inferred when two exemplars inferred progenitor by a modern name is warranted. of the same species share on a cladogram a trait not Traits used in molecular systematics, because shared by a third exemplar of the same species, i.e., not as open to selection as those of morphology, are an internal node with two exemplars terminal and more liable to preservation as molecular races of one more basal. Molecular paraphyly was identified populations with somewhat different DNA sequenc- as a different species inserted on a node between two es that have neutral value in evolution. Of course exemplars of one species, each taken as a separate genetic diversity is common in (e.g., Shaw molecular race. All entries that showed no evidence 2000, 2001; Shaw & Schneider 1995) but we are of molecular races were ignored because they either focused here on DNA differences at a level critical were of one exemplar or the branches to several for taxonomy and classification. exemplars of one species are multifurcations with- When molecular results are not congruent to out internal branches. Because cladograms only analyses based on morphology, either the morphol- recognize shared traits, only a multifurcation to ogy needs further study or the molecular analysis is multiple exemplars of a single-species with one or faulty. In fact, for species whose morphology is more internal branches was taken as evidence of much reduced and critical traits repressed, molec- more than one molecular race. ular systematics can reject proposed classifications The Zander (2019b) study showed that the when sufficient distance is present on a cladogram published papers included 71 entries of multiracial 574 The Bryologist 122(4): 2019 species including 46 different species. The total Distance of this extent is commonly the reason for number of molecular races among the 46 multiracial phylogenetic erection of cryptic species, genera and species was 208. The average number of molecular families. An example of a dubiously recognized races for the 46 species (i.e., the number of internal cryptic family is Limpr. (La Farge nodes interpretable as signaling differently com- et al. 2002; Zander 2008) at six rather unreliable posed sequences) was 4.5 (i.e. 208/46 ¼ 4.5), for molecular nodes distant from . Tim- both paraphyletic and non-paraphyletic species. miellaceae Inoue & Tsubota may be an acceptable The total number of paraphyletic instances (i.e. segregate family of two genera as it is 12 nodes races that are also paraphyletic) among the 71 distant from Pottiaceae (Inoue & Tsubota 2014), but entries was 29. The number of paraphyletic species the supportive morphological traits, mainly the was 22, or 0.4 of the 46 different known multiracial unique sinistrorsely twisted peristome (the unrelat- species. The number of different apophyletic species ed Gymnostomiella Fleisch. also, however, has (those distal to a paraphyly) in all sequences was 79. opercular cells twisted sinistrorsely implying that The average number of apophyletic species per the lost peristome was similarly twisted), are few and paraphyly was 3.6. A total of 133 nodes were weak. That these taxa may represent an ‘‘early maximally distant (on the cladogram) between diversifying haplolepideous lineage’’ is well paraphyletic exemplars. The average number of supported. That this early lineage or isolated pair of nodes between most-distant paraphyletic exemplars lineages is not the Pottiaceae is not refuted by of the same species was 4.5 (133/29 ¼ 4.5). several clades of quite different morphology branch- Resolution of a molecular cladogram among ing between the two genera and the Pottiaceae. these Pottiaceae species was inferred as poor because The lack of correct resolution of evolutionary (1) large numbers of species have molecular races, relationship only shows up when many exemplars of (2) those with sufficient data to determine the each species are analyzed. Differences in expressed numbers of molecular races average 4.5 molecular traits imply DNA differences but the reverse cannot races, (3) a large percentage (40%) of all demon- be used in taxonomy without considering the strably multiracial species are also paraphyletic, (4) potential distribution of molecular races. Naming the many apophyletic species (an average of 3.6) per molecular races obscures the real relationships of paraphyly indicate that an inferred progenitor- serial descent of the average of 3.6 apophyletic descendant relationship commonly involves many species that apparently are monophyletic descen- species or multiple radiation from one ancestor, and dants from one paraphyletic progenitor species. In (5) the large number of nodes between paraphyletic addition, searching for expressed traits that might exemplars of the same species (4.5) implies that match molecular clades leads to multiple test bias phylogenetic differences between species on a (Zander 2013). molecular cladogram must take into account One way of partially mitigating this uncertainty possible paraphyly by extinct or unsampled material in judging macroevolutionary monophyly is re- of any one species. This means that every species course to expressed traits of classical taxonomy. One represented by only one exemplar on a molecular might expect an average of four new morphological cladogram must be expected to have a circle of traits per speciation event, based on averages of 3.6 possible paraphyly, i.e., of monophyletic uncertain- traits found in studies of (Zander 2016) ty, of 4.5 nodes in all directions. and 4 traits for Streptotrichaceae (Zander 2018) An average of nine concatenated nodes of poor species. Two species differing by only four traits resolution of monophyly between any two single- should not be expected to occur more than two or sample species must be allowed to avoid overlap of three nodes apart on a molecular cladogram, while possible molecular paraphyly. This distance is not species of eight or 10 traits difference should not be unusually great; it is in fact a calculated mean. Thus, expected to have a very close evolutionary distance a species that should be cladistically sister to a whether they are phylogenetically sister groups or second may, in some cases, appear nine nodes not. distant on a cladogram if both species are repre- The first author has written (Zander 2013) that sented by only one surviving molecular race. This ‘‘The names Tortella, , Pseudosymble- distance is not unusual, instead it is an average. pharis and Oxystegus may actually represent only Zander & Eckel: Pseudosymblepharis and Oxystegus 575 one genus, characterized by a usually V-shaped basal genera follow, with explanation. The Basal Missing cell arrangement, a V-shaped transverse section of Link (ultimate inferred ancestor) is predicted or the dorsal stereid band (in most species), an often retrodicted to have a stem central strand present, thick-walled central cylinder, and sturdy, non- ligulate leaves with plane margins and basal cells collapsing hyaloderm cells, but further research is differentiated across the leaf, twisted peristome, and necessary.’’ Further research has evidently not been dioicous sexual condition, costa thick and excurrent done to the extent necessary for the clear modeling in a stout, sharp mucro, and basal cells moderately of Chionoloma as a dissilient genus (a central differentiated as a vee across leaf base. Traits that progenitor with a radiating array of descendant distinguish the taxa in the four lineages (suggested species). Chionoloma sensu Alonso et al. (2016, in part by Zander 1993) from the Missing Link are 2018a) is here refuted because of poor resolution in each assigned significance at the 1 bit level, then the molecular cladograms on which it is based, and summedandtranslatedtoBayesianposterior because strict phylogenetic monophyly cannot probability (BPP) for support of that evolutionary model the radiative generation of descendant relationship (see Zander 2013, 2016, 2018). The species, the most relevant of which are below symbol . means ‘‘generates as a descendant,’’ and presented as dissilient genera. A dissilient genus bold-faced taxa model inferred progenitors. Taxo- has descendant species more closely related (in nomic authorities and circumscriptions are those of number of morphological trait changes) to the joint Zander (1993). progenitor than to each other. Basal Missing Link (ultimate inferred ances- The genus Tortella is molecularly paraphyletic at tor) . the base of the Trichostomoideae clade of Werner et al. (2005) and that of Kockinger¨ et al. (2010), with 1. (Trichostomum complex progenitor, a taxon T. fruchartii and T. inflexa at the base split from the similar to Tr. brachydontium) remainder of the genus. This implies that Tortella is 2. (Trichostomum complex progenitor, a taxon a good candidate for a Basal Missing Link, with similar to Tr. crispulum) reduction of the characteristic vee-shaped differen- 3. (Tortella complex with multiple progenitors tiation of basal cells as evolutionarily thematic. Also, including To. humilis and To. tortuosa . Trichostomum brachydontium Bruch is offered here Pseudosymblepharis . (Chionoloma, Pachyneur- to represent a short-peristomed and dioicous form opsis)) of Tortella humilis (Hedw.) Jenn. as a potential 4. (Oxystegus Missing Link . (O. hibernicus)(O. progenitor of at least Oxystegus and Pseudosymble- spiralis)(O. daldinianus . O. recurvifolius)(O. pharis. This concept is in line with taxa at the base of tenuirostris . O. maragniphyllus,)) morphological (Zander 1993: 47) and molecular (Alonso et al. 2016; Hedenas¨ 2015; Werner et al. (1) Basal Missing Link . Trichostomum brachy- 2005) trees. Zander (1993: 80) noted that the genera dontium, where T. brachydontium is least advanced Pseudosymblepharis and Tortella share many traits, through modification and thus is the possible which would connect them through a Basal Missing unreduced progenitor of a set of species including Link (see below). Zander (2017b) alternatively Trichostomum planifolium, T. urceolare, T. sina- suggested Hydrogonium croceum (Brid.) Kuceraˇ as loense, T. termitarum, T. williamsii, T. unguiculatum a possible model for this missing link. and T. austrocrispum. Peristome short; leaves ovate Very robust specimens of Trichostomum brachy- to long-lanceolate, apex usually rounded; costa thick dontium have the adaxial stereid band larger than and excurrent in a stout, sharp mucro (5 bits, 0.97 the abaxial, much as is the case in Pseudosymble- BPP). pharis species. Zander (1993: 88, 90) suggested (2) Basal Missing Link . Trichostomum crisp- reduction series of species in the various subgenera ulum, where T. crispulum is least reduced of a set of Trichostomum s.l. These should be investigated as including Trichostomum castaneum, T. caespitosum, potential dissilient genera. T. brittonianum, T. subangustifolium, T. atrocaule, Preliminary evolutionary overview and formu- and T. perligulatum. Peristome short; leaves elliptical lae. Modified macroevolutionary formulae for to long-lanceolate, apex rounded-acute, often some- dissilient (adaptive and nearly neutral radiative) what cucullate; distal leaf margins broadly incurved; 576 The Bryologist 122(4): 2019 costa evenly tapering and generally ending before base; leaves strongly reflexed from above leaf base (4 the apex or percurrent (6 bits, 0.99 BPP). bits, 0.94 BPP). (3) Basal Missing Link . Tortella, where Oxystegus sp. Missing Link . O. spiralis through Tortella is a complex genus whose macroevolution spiraled peristome ornamentation; autoicy (2 bits, is not easily dealt with. A progenitor similar to T. 0.80 BPP). tortuosa may have generated species of Pseudosym- The species to species analytic paradigm in blepharis, one or two of which in turn produced evolutionary analysis is effective, and is parsimoni- both Chionoloma and Pachyneuropsis. ous by taking into account all data, given the Basal Missing Link . Tortella through leaves molecular cautions. There are apparently related with V-shaped basal cell region; retention of groups (see Zander 1993) that are left out of this elongate peristome in non-reduced species; generally evolutionary model pending further study, includ- grading in lineage to smaller distal laminal cells (3 ing Pleurochaete Lindb., Streptocalyptra Mull.¨ Hal., bits, 0.89 BPP). Trichostomum subg. Laminanchium R.H.Zander, Tortella . Pseudosymblepharis through peri- and Tuerckheimia Broth. stome short and white; narrowly elongate leaves, base sheathing and with shoulders; broad costa with more guide cells; differentiated basal cells rising ACKNOWLEDGMENTS Loan of specimens by UBC is gratefully noted. The Missouri higher along margins (6 bits, 0.99 BPP). Botanical Garden continues to provide important support for study Pseudosymblepharis . Chionoloma through of Pottiaceae. R. M. Ros communicated a slightly corrected stem tomentum red; leaf margins narrowly in- cladogram for the Werner et al. (2005) study. We thank M. Alonso curved, border denticulate; an even broader costa; for permission to reference the recent relevant findings of her basal cells rising higher; small, incrassate distal group, and for many helpful comments on the manuscript. Two laminal cells bulging adaxially (6 bits, 0.99 BPP). reviewers and the editors of The Bryologist provided much valued criticism and perspective. Also: Pseudosymblepharis . Pachyneuropsis (see Alonso et al. 2018a) through broader costa; distal LITERATURE CITED lamina very narrow or vanishing distally, bi- or tri- Alonso, M., M. J. Cano & J. A. Jimenez.´ 2018a. Taxonomy and stratose (3 bits, 0.89 BPP). Distribution of Pachyneuropsis (Pottiaceae, Bryophyta). Plant (4) Basal Missing Link . Oxystegus sp., where Systematics and Evolution 304: 1069–1076. Oxystegus sp. is a Missing Link itself through short Alonso, M., M. J. Cano & J. A. Jimenez.´ 2018b. Stephanodictyon borneense (Pottiaceae, Bryophyta): new combination, typifica- peristome; leaves lanceolate to linear-lanceolate; tion and update of distribution. Plant Biosystematics 152: 1160– larger laminal cells, basal cells moderately differen- 1164. tiated, often running up the margins somewhat as in Alonso, M., J. A. Jimenez,´ S. Nylinder, L. Hedenas¨ & M. J. Cano. Tortella: basal cells bordered distally by a region of 2016. Disentangling generic limits in Chionoloma, Oxystegus, thick-walled rectangular cells (5 bits, 0.97 BPP). Pachyneuropsis and Pseudosymblepharis (Bryophyta: Pottiaceae): An inquiry into their phylogenetic relationships. Taxon 65: 3–18. Oxystegus sp. Missing Link . O. tenuirostris Alonso, M., J. A. Jimenez´ & M. J. Cano. 2018. Taxonomic identity through leaves lanceolate, fragile, leaf base enlarged of Chionoloma bombayense (Pottiaceae, Bryophyta). Nova and half-sheathing (3 bits, 0.89 BPP). Hedwigia 106: 5–15. O. tenuirostris . O. maragniphyllus through Cano, M. J., J. F. Jimenez,´ M. T. Gallego, J. A. Jimenez´ & J. Guerra. 2009. Phylogenetic relationships in the genus distal lamina narrowly flagellate; leaves hooked (Pottiaceae, Bryophyta) inferred from nrITS sequence data. when dry, base strongly dilated and fully sheathing Plant Systematics and Evolution 281: 209–216. the stem, mucro long and sharp (4 bits, 0.94 BPP). Crum, H. & L. E. Anderson. 1958. Taxonomic studies on North Oxystegus sp. Missing Link . O. daldinianus American mosses. VI–XIII. Journal of the Elisha Mitchell through large size; stem lacking central strand; Scientific Society 74: 31–40. Eckel, P. M. 1990. Tortella tortelloides (Musci: Pottiaceae) new to denticulate leaf margins (3 bits, 0.89 BPP). North America. The Bryologist 94: 84–87. Oxystegus daldinianus . O. recurvifolius Eckel, P. M. 1998. A re-evaluation of Tortella (Pottiaceae, Musci) in through shorter and broader leaves, laminal margins conterminous U.S.A. and Canada. Bulletin of the Buffalo Society with sharp teeth and a border of 2–3 rows of smooth of Natural Sciences 36: 117–191. Eckel, P. M. 2000a. Trichostomum arcticum Kaal. (Pottiaceae), new cells (4 bits, 0.94 BPP). to the Province of Ontario, Canada. Evansia 17: 87–89. Oxystegus sp. Missing Link . O. hibernicus Eckel, P. M. 2000b. Trichostomum tenuirostre var. gemmiparum new through large size, smaller laminal cells; short leaf to the United States. Evansia 17: 61–62. Zander & Eckel: Pseudosymblepharis and Oxystegus 577

Eckel, P. M. 2007a. Tortella. In: Flora of North America Editorial Shaw, J. 2001. Biogeographic patterns and cryptic speciation in Committee, eds. Flora of North America North of Mexico. Oxford bryophytes. Journal of Biogeography 28: 253–261. University Press, New York and Oxford. Vol. 27: 498–511. Sollman, P. 2000. A taxonomic revision of Pseudosymblepharis Eckel, P. M. 2007b. Pleurochaete. In: Flora of North America Broth. (Musci: Pottiaceae) in Asia. Tropical Bryology 18: 129– Editorial Committee, eds. Flora of North America North of 145. Mexico. Oxford University Press, New York and Oxford. Vol. Sollman, P. 2005. A taxonomic revision of the genus Chionoloma 27: 496–498. Dixon (Musci, Pottiaceae) in Africa. Tropical Bryology 26: 35– Eckel, P. M. 2010. Tortella bambergeri in North America and an 38. evaluation of its taxonomy. Bulletin of the Buffalo Society of Werner, O., H. Kockinger,¨ J. A. Jimenez´ & R. M. Ros. 2009. Natural Sciences 39: 1–10. Molecular and morphological studies on the Didymodon La Farge, C., A. J. Shaw & D. H. Vitt. 2002. The circumscription of tophaceus complex. Plant Biosystems 143(Suppl.): S136–145. the Dicranaceae () based on the chloroplast regions Werner, O., H. Kockinger,¨ M. Magdy & R. M. Ros. 2014. On the trnL-trnF and rps4. Systematic Botany 27: 435–452. systematic position of Tortella arctica and Trichostomum Grundmann, M., H. Schneider, S. J. Russell & J. C. Vogel. 2006. arcticum (Bryophyta, Pottiaceae). Nova Hedwigia 98: 273–293. Phylogenetic relationships of the moss genus Pleurochaete Werner, O., R. M. Ros & M. Grundmann. 2005. Molecular Lindb. (: Pottiaceae) based on chloroplast and nuclear phylogeny of Trichostomoideae (Pottiaceae, Bryophyta) based genomic markers. Organisms, Diversity & Evolution 6: 33–45. on nrITS sequence data. Taxon 54: 361–368 ‘‘Corrected’’ version Hedenas,¨ L. 2015. Tortella rigens (Bryophyta, Pottiaceae): relation- of cladogram communicated by R. M. Ros. ships, regional variation, and conservation aspects. Plant Zander, R. H. 1978. Molendoa hornschuchiana new to the Americas Systematics and Evolution 301: 1361–1375. from the Pacific Coast of Alaska. The Bryologist 82: 487–489. Hedenas,¨ L. & P. M. Eckel. 2011. 21. Tortella arctica (Arnold) Zander, R. H. 1993a. A phylogrammatic evolutionary analysis of the Crundw. & Nyholm. In: Ellis, L. T. et al. (eds.), New national moss genus Didymodon in North America north of Mexico. and regional bryophyte records, 29. Journal of Bryology 33: 321. Bulletin of the Buffalo Society of Natural Sciences 36: 81–115. Inoue, Y. & H. Tsubota. 2014. On the systematic position of the Zander, R. H. 1993b. Genera of the Pottiaceae: Mosses of Harsh genus (, Bryopsida) and its allied genera, Environments. Bulletin of the Buffalo Society of Natural with the description of a new family Timmiellaceae. Phytotaxa Sciences 32:1–378. 181: 151–162. Zander, R. H. 2007. Molendoa. In: Flora of North America Editorial Kockinger,¨ H., O. Werner & R. M. Ros. 2010. A new taxonomic Committee, eds. Flora of North America North of Mexico. Vol. approach to the genus Oxystegus (Pottiaceae, Bryophyta) in 27. Oxford University Press, New York, Oxford. Europe based on molecular data. Nova Hedwigia Beiheft 138: Zander, R. H. 2008. Statistical evaluation of the clade ‘‘Rhabdo- 31–49. weisiaceae.’’ The Bryologist 111: 292–301. Kuceraˇ J., J. Kosnarˇ & O. Werner. 2013. Partial generic revision of Zander, R. H. 2013. Framework for Post-Phylogenetic Systematics. Barbula (Musci: Pottiaceae): Re-establishment of Hydrogonium Zetetic Publications, CreateSpace Independent Publishing, and Streblotrichum, and the new genus Gymnobarbula. Taxon Amazon, St. Louis. 62: 21–39. Zander, R. H. 2014. Classical determination of monophyly, Kucera,ˇ J. & M. S. Ignatov. 2015. Revision of phylogenetic exemplified with Didymodon s. lat. (Bryophyta). Part 2 of 3, relationships of Didymodon sect. Rufiduli (Pottiaceae, Musci). concepts. Phytoneuron 2014-79: 1–23. 24: 79–97. Zander, R. H. 2016. Macrosystematics of Didymodon sensu lato Kucera,ˇ J., T. L. Blockeel, P. Erzberger, B. Papp, Z. Soldan, K. (Pottiaceae, Bryophyta) using an analytic key and information Vellak, O. Werner & R. M. Ros. 2018. The theory. Ukrainian Botanical Journal 73: 319–333. complex (Pottiaceae, Bryophyta) revisited: new data support the Zander,R.H.2017a.Anoectangium sikkimense (Pottiaceae, subspecific rank of currently recognized species. Cryptogamie, Bryophyta) new to the New World from Alaska, and its Bryologie 39: 241–257. macroevolutionary relationships. The Bryologist 120: 435–440. Kucera,ˇ J., J. Kosnarˇ & O. Werner. 2013. Partial generic revision of Zander, R. H. 2017b. Oxystegus daldinianus (Pottiaceae, Bryophyta) Barbula (Musci: Pottiaceae): Re-establishment of Hydrogonium new to the New World, evaluated with two new tools for and Streblotrichum, and the new genus Gymnobarbula. Taxon floristics. The Bryologist 120: 50–57. 62: 21–39. Zander, R. H. 2018. Macroevolutionary Systematics of Strepto- Li, X-J., Si He & Z. Iwatsuki. 2001. Pottiaceae. Pages 114–249. In: trichaceae of the Bryophyta and Application to Ecosystem X.-J. Li & M. R. Crosby (eds.), Moss Flora of China. English Thermodynamic Stability. Edition 2. Zetetic Publications, St. Version. Volume 2. –. Science Louis. Press, Beijing, and Missouri Botanical Garden, St. Louis. Zander, R. H. 2019a. Macroevolutionary evaluation methods Saito, K. 1975. A monograph of Japanese Pottiaceae (Musci). extended, consolidated, and exemplified with Anoectangium Journal of the Hattori Botanical Laboratory 39: 373–537. (Pottiaceae, Bryophyta) in North America and the Himalayas. Shaw, A. J. & R. E. Schneider. 1995. Generic biogeography of the Annals of the Missouri Botanical Garden 104: 324–338. rare ‘‘copper moss,’’ Mielichhoferia elongata (Bryaceae). Amer- Zander, R. H. 2019b. Macroevolutionary versus molecular analysis: ican Journal of Botany 82: 8–14. Systematics of the Didymodon segregates Aithobryum, Exobryum Shaw, J. 2000. Molecular phylogeography and cryptic speciation in and Fuscobryum (Pottiaceae). Hattoria 10: 1–38. the mosses, Mielichhoferia elongata and M. mielichhoferiana (Bryaceae). Molecular Ecology 9: 595–608. manuscript received January 31, 2019; accepted August 22, 2019.