﻿Phylogeny of the genus Loxospora s.l. (Sarrameanales, Lecanoromycetes, Ascomycota), with Chicitaea gen. nov. and five new combinations in Chicitaea and Loxospora

﻿Abstract Loxospora is a genus of crustose lichens containing 13 accepted species that can be separated into two groups, based on differences in secondary chemistry that correlate with differences in characters of the sexual reproductive structures (asci and ascospores). Molecular phylogenetic analyses recovered these groups as monophyletic and support their recognition as distinct genera that differ in phenotypic characters. Species containing 2’-O-methylperlatolic acid are transferred to the new genus, Chicitaea Guzow-Krzem., Kukwa & Lendemer and four new combinations are proposed: C.assateaguensis (Lendemer) Guzow-Krzem., Kukwa & Lendemer, C.confusa (Lendemer) Guzow-Krzem., Kukwa & Lendemer, C.cristinae (Guzow-Krzem., Łubek, Kubiak & Kukwa) Guzow-Krzem., Kukwa & Lendemer and C.lecanoriformis (Lumbsch, A.W. Archer & Elix) Guzow-Krzem., Kukwa & Lendemer. The remaining species produce thamnolic acid and represent Loxospora s.str. Haplotype analyses recovered sequences of L.elatina in two distinct groups, one corresponding to L.elatina s.str. and one to Pertusariachloropolia, the latter being resurrected from synonymy of L.elatina and, thus, requiring the combination, L.chloropolia (Erichsen) Ptach-Styn, Guzow-Krzem., Tønsberg & Kukwa. Sequences of L.ochrophaea were found to be intermixed within the otherwise monophyletic L.elatina s.str. These two taxa, which differ in contrasting reproductive mode and overall geographic distributions, are maintained as distinct, pending further studies with additional molecular loci. Lectotypes are selected for Lecanoraelatina, Pertusariachloropolia and P.chloropoliaf.cana. The latter is a synonym of Loxosporachloropolia. New primers for the amplification of mtSSU are also presented.

Some species that produce lichenised vegetative diaspores are morphologically (except for the development of such diaspores) and chemically almost identical to the taxa that lack those structures and such cases are referred to as species pairs (Poelt 1970;Crespo and Pérez-Ortega 2009).Molecular phylogenetic studies of such pairs and of species with lichenised vegetative diaspores generally, however, suggest that the situation is more complex and nuanced than binary pairs of species that either lack vegetative diaspores and are sexually reproducing or produce vegetative diaspores and are only infrequently sexually reproducing.In some cases, neither species delimited by the presence or absence of vegetative diaspores was found to be monophyletic and, instead, representatives of each were intermingled suggesting that independent lineages do not correspond to reproductive mode (e.g.Lohtander et al. (1998); Buschbom and Mueller (2006); Myllys et al. (2011); Tehler et al. (2013); Ertz et al. (2018)).In other cases, such pairs of species have been recovered as reciprocally monophyletic and sister (e.g.Miadlikowska et al. (2011); Lendemer and Harris (2014); Yakovchenko et al. (2017); Ohmura (2020)).Further, there are recent examples where next generation sequence data have provided support for species pair delimitations that lacked support from analyses of traditionally used loci that are typically more conserved and fewer in number (e.g.Grewe et al. (2018)).
In summer 2021, while performing field lichen studies in northern Poland, we collected specimens resembling Loxospora elatina growing on bark of Alnus glutinosa in black alder forest.They contained thamnolic acid as the main secondary metabolite; however, the thallus was continuous to areolate, in contrast to the tuberculate thalli typically found in L. elatina (e.g.Stenroos et al. (2016)).Molecular analyses showed that these specimens and some other samples published by Kelly et al. (2011) formed a group distinct from samples of L. elatina with typical tuberculate thalli.Recognising the need to re-evaluate the delimitation of L. elatina based on this material, we analysed additional sequences and specimens of other Loxospora species to confirm the relationships amongst currently recognised species, especially L. ochrophaea (Tuck.)R.C.Harris, which has been presumed to be the strictly sexual, esorediate counterpart to L. elatina (Brodo et al. 2001;Guzow-Krzemińska et al. 2018).Based on these analyses, we recognise the material of L. elatina with continuous to areolate thalli as distinct and introduce a new combination for it, discuss the status of L. elatina s.str.and L. ochrophaea (Tuck.)R.C. Harris and introduce the genus Chicitaea for the clade of Loxospora species producing 2'-O-methylperlatolic acid, which necessitates four new combinations.

DNA extraction, PCR amplification and DNA sequencing
Small pieces of thalli (approx. 2 mm 2 ) were put into Eppendorf tubes.Then DNA was extracted using a GeneMATRIX Plant & Fungi DNA Purification Kit (EURX) or a modified CTAB method (Guzow-Krzemińska and Węgrzyn 2000).Sequences of three molecular markers were amplified: nuITS rDNA using ITS1F (Gardes and Bruns 1993) or ITS5 (White et al. 1990) and ITS4 (White et al. 1990) primers, RPB1 using g-RPB1-A for (Stiller and Hall 1997) and f-RPB1-C rev (Matheny et al. 2002) primers and mtSSU using mrSSU1 (Zoller et al. 1999) and mrSSU3R (Zoller et al. 1999) primers.Due to difficulties in mtSSU amplification, new primers were designed by one of the authors (Beata Guzow-Krzemińska; primers here referred to as "Lox_mtSSU620_For": 5'-TTTACCTATATGTCTTGACCAA-3' and "Lox_mtSSU620_Rev": 5'-CTCTTATCATATTCCAATATAATG-3').PCR settings for each set of primers are shown in Suppl.material 1. Electrophoresis was performed on a 1% agarose gel to determine whether amplification of target molecular markers was successful.PCR products were purified using Clean-Up Concentrator (A&A Biotechnology).Sequencing was performed by Macrogen (The Netherlands).All newly-generated sequences were deposited in GenBank and their GenBank Acc.Numbers are presented in Table 1.

Sequence alignments and phylogenetic analyses
The newly-obtained sequences were trimmed using the Chromas programme (http://technelysium.com.au/wp/).All sequences were analysed using BLASTn search (Altschul et al. 1990).Independent alignments of nuITS, mtSSU rDNA and RPB1 markers were prepared using Seaview software (Galtier et al. 1996;Gouy et al. 2010) employing muscle option and guidance2 software implemented on an online website (Sela et al. 2015; https://guidance.tau.ac.il/).Single locus alignments consisted of 68 nuITS rDNA sequences with 548 sites, 47 mtSSU rDNA sequences with 635 sites and 13 RPB1 sequences with 562 sites.Then, datasets were concatenated into one matrix which consisted of 83 terminals with 1745 positions.The concatenated dataset was subjected to IQ-TREE analysis to find best-fitting nucleotide substitution models for each partition (Nguyen et al. 2015;Chernomor et al. 2016;Kalyaanamoorthy et al. 2017;Hoang et al. 2018).The model selection was restricted to models implemented in MrBayes and the following nucleotide substitution models for the three predefined subsets were selected: HKY+F+I for mtSSU rDNA, K2P+F+G4 for nuITS and K2P+F+I for RPB1.The search for the Maximum Likelihood tree was performed in IQ-TREE and followed with 1000 bootstrap replicates (Nguyen et al. 2015;Chernomor et al. 2016;Kalyaanamoorthy et al. 2017;Hoang et al. 2018).
The Bayesian analysis was conducted using MrBayes 3.2.7a(Huelsenbeck and Ronquist 2001;Ronquist and Huelsenbeck 2003) on the CIPRES Science Gateway (Miller et al. 2010).The analyses were conducted by running 10,000,000 generations.The chain was sampled every 1000 th generation.Posterior probabilities (PP) were determined by calculating a majority-rule consensus tree after discarding the initial 25% trees of each chain as the burn-in.All trees were visualised in FigTree v.1.4.(Rambaut 2009) and further modified in Inkscape (https://inkscape.org/).Bootstrap support (BS values ≥ 75) and PP values (values ≥ 0.95) are given near the branches on the phylogenetic tree.
Sequences obtained from GenBank and used in phylogenetic analyses are listed in Suppl.material 2.

Preparation of haplotype networks
Moreover, independent alignments of each marker for specimens of L. elatina, L. ochrophaea and L. chloropolia were prepared using Seaview software (Galtier et al. 1996;Gouy et al. 2010) employing muscle option and followed with manual correction.The final nuITS rDNA alignment consisted of 46 sequences with 443 sites, while RPB1 alignment consisted of 11 sequences with 723 sites.Haplotype analyses were performed using PopART software (https://popart.maths.otago.ac.nz) employing TCS network option (Clement et al. 2002).Moreover, variable sites that distinguish these taxa were identified.Similar analyses were done for specimens of L. assateaguensis, L. confusa and L. lecanoriformis.The final alignment of nuITS rDNA consisted of 11 sequences with 534 sites, while mtSSU rDNA alignment consisted of eight sequences and 613 sites.

Results and discussion
The representatives of the genus Loxospora s.l. are split into two highly-supported major clades (Fig. 1).The larger clade corresponds to Loxospora s.str.
(type: L. elatina), all containing thamnolic acid as the main secondary lichen substance and having asci with a uniformly amyloid apical dome and ascospores that are septate, fusiform to ellipsoidal and somewhat curved or twisted (Tønsberg 1992;Brodo et al. 2001;Sanderson et al. 2008).This clade is divided into two subclades.The smaller one consists of representatives of L. cismonica (Beltr.)Hafellner, while the larger subclade consists of two poorly-supported lineages, which might be the result of uneven coverage of sequences for each species in this subclade (see Table 1, Suppl.material 2).However, the phylogenetic analyses, based only on nuITS (not shown here) and the nuITS haplotype network analysis (Fig. 2), recovered these two groups as different and with high confidence.In the nuITS rDNA haplotype network analysis, these groups differ from each other in 21 nucleotide positions and the variability within the groups is up to three substitutions.Moreover, RPB1 haplotype network analysis also supports distinction of these two groups as they differ in 10 positions (Fig. 3), while the mtSSU rDNA marker showed very low variation (data not shown).The larger group includes sequences of specimens with at least partly tuberculate thalli with soralia, which are often fusing (i.e.corresponding to L. elatina s.str.) and thalli that uniformly lack soralia, but are typically fertile (i.e.corresponding to L. ochrophaea).The smaller group consists of sequences of samples in which the thalli are continuous to slightly cracked-areolate, but never tuberculate and soralia are usually discrete, rarely fusing and, if so, then only in older parts of the thallus.The specimens whose sequences were recovered in this latter group correspond morphologically to the type material of Pertusaria chloropolia Erichsen (≡ Lecanora chloropolia (Erichsen) Almb.), not to the type of Lecanora elatina Ach.(basionym of Loxospora elatina).Pertusaria chloropolia was synonymised with Loxospora elatina by Laundon (1963), a treatment followed subsequently by Hafellner and Türk (2016) and Westberg et al. (2021).All of the existing herbarium specimens corresponding to the type of Pertusaria chloropolia and presented in this present paper were initially identified as L. elatina and filed under that name in herbaria.However, as the molecular data show, this material corresponds to a phenotypically distinct monophyletic group for which the name P. chloropolia is available.The name is resurrected from synonymy and a new combination is proposed below.The revised circumscriptions of both Loxospora chloropolia and L. elatina are presented below and lectotypes are selected for both names.Moreover, in addition to morphology, their nuITS rDNA and RPB1 sequences differ in numerous positions of which several may be used as diagnostic characters to distinguish these taxa (Tables 2, 3).
The smaller clade of Loxospora s.l. is represented by L. assateaguensis Lendemer, L. confusa Lendemer, L. cristinae Guzow-Krzem., Łubek, Kubiak & Figure 1.IQ-tree based on a combined nuITS rDNA, mtSSU and RPB1 dataset for Loxospora s.l.The names of species are followed with sample number (see Table 1, Suppl.material 2).Bootstrap supports from IQ-tree analysis ≥ 70 (first value) and posterior probabilities from BA ≥ 0.95 (second value) are indicated near the branches.Umbilicaria spp.were used as outgroup.Loxospora chloropolia clade is marked with blue box and Chicitaea gen.nov. is marked with green box.
Kukwa and L. lecanoriformis (Fig. 1).All these species produce 2'-O-methylperlatolic acid and it has been repeatedly suggested that they represent a group distinct from the thamnolic acid producing species of Loxospora s. str.which likely merits recognition as a distinct genus (Lumbsch et al. 2007;Lendemer 2013;Guzow-Krzemińska et al. 2018).While apothecia are known only in L. lecanoriformis, in that species, the asci lack an amyloid apical dome, unlike in Loxospora s.str.and the ascospores are simple, ellipsoidal, straight or slightly bent (Lumbsch et al. 2007;Papong et al. 2009).Due to the consistent differences from Loxospora s.str. in secondary lichen substances, the differences in ascus amyloidy and the strongly-supported monophyly of this group in molecular phylogenetic analyses, we recognise it as a distinct genus under the name Chicitaea below.Four new combinations are proposed for the species currently known to belong to this clade.Chicitaea cristinae was recovered as monophyletic and sister to the rest of the species, which form a well-supported clade, but with poorly resolved relationships between Ch. confusa and Ch.lecanoriformis.The fertile Ch. lecanoriformis, known from Australia and Thailand (Lumbsch et al. 2007;Papong et al. 2009), is nested within a subclade of sequences of Ch. confusa, an isidioid species which occurs in North America and is not known to occur in the Southern Hemisphere or Australasia (Lendemer 2013).Due to the lack of nuITS rDNA sequence for Ch.lecanoriformis and very low variation found in mtSSU sequences (Fig. 4), the relationship between these species cannot be resolved.Nevertheless, both species clearly differ morphologically and have disjunctive distributions (Lumbsch et al. 2007;Papong et al. 2009;Lendemer 2013).Chicitaea confusa seems to be paraphyletic and may represent two cryptic species (Fig. 1).This conclusion is also supported by the haplotype analyses of mtSSU and nuITS sequences (Figs 4,5) which also show that two specimens (Ch.confusa 1 and 2) significantly differ from all the newly-sequenced representatives of Ch. confusa, but more material is needed to solve this problem.The sequences of one specimen, initially determined as Ch.confusa (Ch.aff.confusa 6; Figs 1, 4, 5), is identical in mtSSU and nuITS sequences with Ch. assateaguensis.This suggests that Ch. assateaguensis can represent a cryptic species, even though, as stated by Lendemer (2013), the species differed from Ch. confusa, but more material is necessary before final conclusions.
Loxospora elatina s.str.and L. ochrophaea are morphologically similar in terms of thallus and apothecia and both produce thamnolic acid often with elatinic acid and trace amounts of squamatic acid (Tønsberg 1992;Brodo et al. 2001;Sanderson et al. 2008).The only difference between L. elatina s.str.and L. ochrophaea is the consistent presence of soralia in L. elatina (apothecia are very rare) and the absence of soralia in L. ochrophaea which is, instead, consistently fertile and routinely produces apothecia (Kalb and Hafellner 1992;Tønsberg 1992;Brodo et al. 2001;Sanderson et al. 2008).From a phenotypic perspective, these two taxa can be considered a species pair (cf.Poelt (1970); Crespo and Pérez-Ortega (2009)).
Although both species are frequently found on the acidic bark of trees and both are distributed in the Northern Hemisphere, their distributions are divergent and not entirely sympatric.Loxospora elatina is widely distributed in boreal and northern temperate areas of the Northern Hemisphere with oceanic climates (e.g.Sanderson et al. (2008); Urbanavichus (2010); Stenroos et al. (2016)).ochrophaea.The names of species are followed with sample numbers (see Table 1, Suppl.material 2).Newly-sequenced samples are marked in bold.Mutational changes are presented as numbers in brackets near lines between haplotypes.
In contrast, L. ochrophaea has a narrower, disjunct distribution between the Appalachian-Great Lakes regions of eastern North America and north-eastern Asia (Japan and the Russian Far East) (e.g.Tuckerman (1848); Brodo et al. (2001); Urbanavichus (2010); Ohmura and Kashiwadani (2018)).Indeed, the distributions of these two taxa follow the predictions of the species pair hypothesis, wherein the species with vegetative diaspores has a much larger range compared to that of the strictly sexual species that lacks vegetative diaspores (Poelt 1970;Mattsson and Lumbsch 1989).
Table 2. Variable positions in the alignment of nuITS rDNA marker of Loxospora chloropolia, L. elatina and L. ochrophaea.Variable characters are marked in bold, while diagnostic nucleotide position characters to distinguish L. chloropolia from both L. elatina and L. ochrophaea are marked with a gray background, including indels.

Position in alignment
L. chloropolia 1   The names of species are followed with sample numbers (see Table 1, Suppl.material 2).Newly-sequenced samples are marked in bold.Mutational changes are presented as numbers in brackets near lines between haplotypes.

C C C T T T C T A T A ----
In our analyses, sequences of Loxospora elatina s.str.were intermingled with L. ochrophaea within the same clade (Fig. 1).Six different nuITS haplotypes were found in these species which differed up to three nucleotide substitutions between each other (Fig. 2).The most common haplotype was found in 20 specimens of L. elatina collected in Poland, Switzerland and two geographically distant locations in Appalachian eastern North America (sample L. elatina 22 is from New York, U.S.A. and sample L. elatina 23 is from North Carolina, U.S.A.; Table 1).Moreover, in the nuITS haplotype network, four samples of L. elatina and four samples of L. ochrophaea share the same haplotype (Fig. 2).While these samples were all collected in eastern North America, they include samples of each species that were collected at very distant locations (e.g.sample L. ochrophaea 3 is from coastal Maine, U.S.A., while samples L. ochrophaea 5, 6 and 7 are from Appalachian North Carolina and Tennessee, U.S.A.; sample L. elatina 20 is from coastal Maine, U.S.A, sample L. elatina 21 is from the Great Lakes of Michigan, U.S.A., while samples L. elatina 26 and L. elatina 27 are from Newfoundland, Canada; Table 1).Interestingly, a sample of each species was collected in close proximity at the same locality (samples L. ochrophaea 3 and L. elatina 20, both from the same location on Roque Island in Maine, U.S.A.; Table 1).Given their phenotypic similarity and the lack of resolution using nuITS rDNA, the molecular barcoding marker for fungi, it is possible that L. elatina and L. ochrophaea may represent variants of a single species.On the other hand, it is also possible that our data were insufficient to distinguish between two closely-related species and more detailed study would allow to find differences between them.Recently, in the case of Usnea antarctica Du Rietz and U. aurantiacoatra (Jacq.)Bory, RADseq and comparative genomics supported recognition of a species pair that had previously been proposed to be synonyms (Grewe et al. 2018).Given that the species have strongly divergent distributions and that they are morphologically distinct when they co-occur, we refrain from synonymising them at this time.Diagnosis.Differs from Loxospora s.str. in the presence of 2'-O-methylperlatolic acid (vs.thamnolic acid), asci without an amyloid apical dome (vs.asci with a uniformly amyloid apical dome) and simple, broadly ellipsoid, straight or slightly bent ascospores (known only in the type species; vs. transversely septate ascospores).
Generic type.Chicitaea lecanoriformis (Lumbsch, A.W. Archer & Elix) Guzow-Krzem., Kukwa & Lendemer.Etymology.The generic epithet honours Chicita F. Culberson (1931Culberson ( -2023)), Senior Research Scientist at Duke University, U.S.A., for her foundational, pioneering and lifelong contributions to the fields of lichen chemistry and lichen taxonomy.In addition to establishing standardised protocols to study lichen secondary chemistry that have been routinely used by workers worldwide for more than half a century, she was an influence for generations of lichenologists with whom she generously shared her knowledge and experience.

Chicitaea assateaguensis (Lendemer
Typifications.The type specimen of Pertusaria chloropolia consists of thin, continuous thallus with discrete soralia forming from flat parts of thalli or from slightly convex areoles and contains thamnolic acid (detected by I. M. Brodo).
In the type specimen of P. chloropolia f. cana, soralia are partly damaged, but, similarly to the type of P. chloropolia, the type consists of thin, continuous thallus with discrete soralia and contains thamnolic acid (detected by I. M. Brodo).In the protologue of P. chloropolia f. cana, the type locality was cited as 'Tschechoslowakei: Karpathoruβland, Lopusanka' (Erichsen 1935), but to our knowledge, it is now located in western Ukraine.The name 'Lopusanka' is a spelling error as, on the label, it is 'Lopušanka'.Erichsen (1935) cited only one locality for both names.However, the lectotypes are selected, because it is not known if, at the time of describing both taxa, C. F. E. Erichsen used only one element upon which the validating descriptions were based (Art. 9.3;Turland et al. (2018); see also McNeill (2014)).
Notes.Loxospora chloropolia differs from L. elatina in having a thin, continuous to cracked-areolate thallus with mostly regular soralia, which are discrete at least in young parts of thalli (Fig. 7).Areoles in the central parts of larger thalli may become convex (in few specimens; Fig. 7E), but are never tuberculate or isidia-like as in L. elatina (Fig. 8).Soralia develop by breaking the cortex and are mostly regular, discrete and convex, rarely flat.Sometimes the neighbouring soralia are fused; however it is still possible to detect the boundaries between individual soralia in most cases.Loxospora elatina, in contrast, has thalli which are, in most cases, tuberculate (sometimes only locally) or with areoles that resemble coarse isidia (Fig. 8).Tuberculate areoles are grouped or dispersed and constricted at the base.Soralia develop from the top of the tuberculate or pustulate areoles and are never regular as in L. chloropolia and, in most thalli, form granular-sorediate patches covering large areas (sometimes almost the entire thallus is covered with soredia; Fig. 8D).Moreover, these species differ in several nucleotide positions in both nuITS rDNA and RPB1 markers (Tables 2, 3).
Habitat and distribution.Typification.In the protologue of Lecanora elatina, Acharius (1810) cited the locality as "Habitat in cortice Pini Abietis Silesiae.Mosig".The type collection in H-ACH consists of four pieces of bark covered with thalli of Loxospora elatina.Three (H-ACH 1199A, 1199B and 1199C) are annotated "Lusatia" with a very faint pencil note next to H-ACH 1199A deciphered as possibly "Mosig" (this note probably not added by Acharius himself as the handwriting in pencil differs from all notes made in ink).The fourth specimen, H-ACH 1199D is annotated "Germania.Schrader".According to the label added in modern times and attached to the type collection, Lusatia was part of Silesia, therefore, the three specimens annotated "Lusatia" can be considered original material; however, it is impossible to verify whether all three were collected by Mosig.Nevertheless, the largest sample (H-ACH 1199A) is fertile and apothecia were mentioned in the diagnosis, therefore it is selected as lectotype.The Acharius collection in BM also contains a specimen of Lecanora elatina, however without any locality details; therefore, it cannot be considered as an isolectotype.
Description.Thallus crustose, grey, matt, thin (at the margin) or more usually thick, continuous or cracked, slightly folded at least the margins, later areolate-verrucose to tuberculate (sometimes only part of the thallus tuberculate).Areoles usually strongly convex, tuberculate and constricted at the base or resembling coarse isidia, sometimes pustulate, dispersed or aggregated.Soralia whitish to greenish-grey, flat or more often convex, rounded or more often irregular, bursting from the top of areoles, often fused and tending to coalesce locally on the thallus or covering most parts of the thallus, sometimes developing from irregular cracks of the thallus.Soredia up to 60 µm in diam., often in consoredia up to 120 µm wide.Apothecia rare, up to 1.2 mm in diam., single or grouped up to five apothecia.Thalline margin present in young apothecia, smooth to flexuose, verrucose or dentate, sometimes with small soralia, later excluded.Excipulum proprium thin, flesh-coloured to white grey in surface view, orange-brown in section, smooth or more often flexuous, up to 100 µm wide in section.Disc reddish-brown, thinly white pruinose.Hymenium up to 125 µm high.Epihymenium straw-brown (K+ pale reddish-brown), with dense granules dissolving in K. Paraphyses not capitate, sometimes anastomosing.Asci 8-spored, with uniformly KI+ blue apical dome.Ascospores 0-5-septate, spiralled in asci, hyaline, fusiform, curved, 35-53(-64) × 4.5-6.5(-7)µm.Pycnidia not known.Photobiont chlorococcoid, cells up to 12 µm in diam.
Notes.Loxospora elatina is similar to L. chloropolia; for differences, see under that species.The name (often as Haematomma elatinum (Ach.) A. Massal.) was often used in the past for the non-sorediate specimens currently referred to as L. ochrophaea.Both species, as mentioned above, are indeed morphologically (except for the production of soralia) and chemically almost identical and may represent the same species.
Loxospora ochrophaeoides, when described, was compared with L. ochrophaea and characterised as differing only in the presence of semi-globose soralia (Kalb and Hafellner 1992).Whether this taxon is distinct or synonymous with L. elatina or L. chloropolia, needs further studies using molecular techniques.Some specimens of L. elatina were found to be determined as Ochrolechia androgyna (Hoffm.)Arnold, but that species and the recently segregated O. bahusiensis H. Magn.and O. mahluensis Räsänen differ in the production of gyrophoric acid and simple, larger ascospores (Tønsberg 1992;Kukwa 2011).
Habitat and distribution.The species is corticolous or lignicolous and grows on bark of various coniferous and deciduous tree in forests.The species was reported from many countries in the Northern Hemisphere; however, as some records may belong to L. chloropolia, its distribution needs revision.In the course of this study, we examined specimens from Austria, Czechia, Estonia, Finland, Latvia, Lithuania, Poland, Slovakia, United Kingdom, Ukraine and USA.

Figure 2 .
Figure2.Haplotype network showing relationships between nuITS rDNA sequences from Loxospora chloropolia, L. elatina and L. ochrophaea.The names of species are followed with sample numbers (see Table1, Suppl.material 2).Newly-sequenced samples are marked in bold.Mutational changes are presented as numbers in brackets near lines between haplotypes.

Figure 3 .
Figure 3. Haplotype network showing relationships between RPB1 sequences from Loxospora chloropolia, L. elatina and L.ochrophaea.The names of species are followed with sample numbers (see Table1, Suppl.material 2).Newly-sequenced samples are marked in bold.Mutational changes are presented as numbers in brackets near lines between haplotypes.

Figure 4 .
Figure 4. Haplotype network showing relationships between mtSSU rDNA sequences from Chicitaea assateaguensis, Ch. confusa and Ch.lecanoriformis.The names of species are followed with sample numbers (seeTable 1, Suppl.material 2).Newly-sequenced samples are marked in bold.Mutational changes are presented as numbers in brackets near lines between haplotypes.

Figure 5 .
Figure5.Haplotype network showing relationships between nuITS rDNA sequences from Chicitaea assateaguensis and Ch.confusa.The names of species are followed with sample numbers (see Table1, Suppl.material 2).Newly-sequenced samples are marked in bold.Mutational changes are presented as numbers in brackets near lines between haplotypes.

Figure 6 .
Figure 6.Morphology of two species of Chicitaea A Thallus of Ch. confusa on tree trunk (taken by J. Hollinger in the field) B thallus of Ch. cristinae on tree trunk (taken by D. Kubiak in the field) C, D Thalli of Ch. cristinae showing soralia (paratypes of L. cristinae C UGDA L-22396 D UGDA L-20385).Scale bars: 1 mm (C, D).

Table 1 .
Specimen data and the GenBank accession numbers of newly-obtained sequences of the taxa used in the phylogenetic analyses.A dash provides information about lack of DNA sequence.For sequences obtained from GenBank, see Suppl.material 2.

Table 3 .
Variable positions in the alignment of RPB1 marker of Loxospora chloropolia, L. elatina and L. ochrophaea.Variable characters are marked in bold, while diagnostic nucleotide position characters to distinguish L. chloropolia from both L. elatina and L. ochrophaea are marked with a gray background.