Usnea nipparensis and U. sinensis form a ‘species pair ’ presuming morphological, chemical and molecular phylogenetic data

Phylogenetic relationships between Usnea nipparensis and U. sinensis, caperatic acid containing Usnea species, were examined based on ITS rDNA, and the phylogenetic position of U. nipparensis was inferred based on multi-locus gene analysis using ITS rDNA, nuLSU, and MCM7. Although U. nipparensis and U. sinensis have a sorediate and an esorediate shrubby thallus, respectively, and in general look quite different, other detailed morphological and chemical features are similar. Analysis of the ITS rDNA sequences suggests their close relationship, but also confirms the independence of both species, and that they most likely form a ‘species pair’ based on morphological, chemical and molecular phylogenetic data. Phylogenetic trees based on both multi-locus gene and ITS rDNA alone strongly support that U. nipparensis and U. angulata belong to the same clade.


Introduction
The genus Usnea (Parmeliaceae) is known as one of the most difficult genera to identify due to the high morphological variability within species (Clerc 1998), while recent studies with molecular data using correctly identified specimens made big progress to understand the species concept and phylogeny of this difficult group (Ohmura 2002(Ohmura , 2008Ohmura & Kanda 2004;Wirtz et al. 2008;Kelly et al. 2011;Lumbsch & Wirtz 2011;Saag et al. 2011;Truong et al. 2013;Truong & Clerc 2016;Clerc & Otte 2018;Gerlach et al. 2017Gerlach et al. , 2019Ohmura & Clerc 2019).
Among c. 350 species of Usnea worldwide (Lücking et al. 2017), clarifying 'species pairs' would be an interesting matter to discuss, considering their distribution, dispersal strategy and evolution. The 'species pair' concept (Poelt 1970(Poelt , 1972Tehler 1982;Mattsson & Lumbsch 1989) is generally applied to a pair of taxa morphologically, anatomically and chemically similar, but that can be distinguished by their sexual vs. asexual reproductive strategies. The 'primary species' produces fruiting bodies and sexual spores, while its counterpart, the 'secondary species' is vegetatively dispersed by soredia, isidia, or fragmentation.
Regarding the genus Usnea, U. florida (L.) F.H. Wigg. and U. subfloridana Stirt. are a good example of 'species pair', being the primary and secondary species respectively (Clerc 1984). Several other species pairs were also proposed in the genus Usnea by Walker (1985) (i.e., U. aurantiacoatra -U. antarctica; U. perpusilla -U. sphacelata; U. trachycarpa -U. subantarctica) and by Shen et al. (2012) (i.e., U. orientalis Motyka -U. pygmoidea). However, molecular phylogenetic analyses using single-or multi-locus genetic data have not supported most of these relationships and they considered that they are conspecific (Articus et al. 2002;Seymour et al. 2007;Saag et al. 2011;Wirtz et al. 2012;Mark et al. 2016) except U. aurantiacoatra -U. antarctica that were revealed as independent species by using microsatellite analysis  and RADseq (Grewe et al. 2018). Since the relationship for U. orientalis -U. pygmoidea was not tested by phylogenetic analysis in Shen et al. (2012), these ITS rDNA sequences were also incorporated into the analysis in this present study.

Materials and methods
This study is based on the examinations of herbarium specimens housed in the National Museum of Nature and Science (TNS), Tsukuba, Japan (File S1).
Morphological observations for identification were made using a dissecting microscope and a bright field microscope. The ratios of thickness of the cortex, medulla, and axis for the branch were measured following the method of Clerc (1984Clerc ( , 1987. Cross sections of thallus were cut by hand with a razor blade, and observed after mounting in GAW (glycerin: ethanol: water, 1: 1: 1).
The sequences were aligned in MAFFT Version 7 (Katoh et al. 2019) using the default settings. Each data set (ITS rDNA, nuLSU, and MCM7) was separately aligned. After removing sites with gaps, missing data and ambiguous data, the data were concatenated. The resulting alignment of 1,512 sites for the multi-locus data set (File S2) or 457 sites for ITS rDNA (File S3) was used for the molecular phylogenetic analyses.
The maximum likelihood (ML) (Felsenstein 1981) and neighbor-joining (NJ) (Saitou & Nei 1987) analyses were performed with the best nucleotide substitution model [TN93+G model (Tamura-Nei 1993) for multi-locus analysis and K2+G model (Kimura 1980) for ITS rDNA analysis]. The bootstrap values (Felsenstein 1985) with 1,000 replicates for ML and NJ were shown on the branches only when both were ≥ 50% simultaneously. All calculations were conducteed in MEGA 10.1.8 (Kumar et al. 2018).
The sample data for molecular analyses and their Gen-Bank accession numbers for the obtained sequences are shown in Table 1.

Results and discussion
Phylogenetic position of Usnea nipparensis in the subgenus Usnea The topology of the molecular phylogenetic tree based on the multi-locus dataset of ITS rDNA, nu LSU and MCM7 obtained in this study ( Fig. 1) is not in conflict with the one shown in Truong & Clerc (2016). The clades or nodes of NEUROPOGON and USNEA-1 to USNEA-4 were formed in the same order as in Truong & Clerc (2016) but USNEA-3 clade was not formed in this tree even with weak support value. Within the USNEA-4 clade, some branches in the tree were insufficiently supported by the bootstrap values. This is, because unlike Truong & Clerc (2016), the tree was calculated with less alignment data, removing the sites with gaps and missing data. Such treatment for alignment is generally desirable for phylogenetic analysis, because different regions of DNA or amino acid sequences evolve under different evolutionary forces (Kumar et al. 2018).
Usnea nipparensis formed a monophyletic clade with U. angulata Ach. with high support values (ML/NJ = 95/90). The phylogenetic position of U. nipparensis -U. angulata clade within the USNEA-4 clade could not be inferred from the current data.
Ohmura (2002) showed a weak relationship (< 50% support value) in the NJ tree based only on ITS rDNA between U. nipparensis and U. mutabilis Stirt., which contains murolic acid complex (fatty acids). However, the tree in this study based on multi-locus gene analyses with ML and NJ methods was also unable to improve the weak support value for the relationship.
Phylogenetic relationship of Usnea nipparensis, U. sinensis, and the related species Six sequences of ITS rDNA for U. nipparensis and ten sequences for U. sinensis were analyzed within the subgenus Usnea using the same dataset of ITS rDNA sequences used in the multi-locus analysis and sequences of U. orientalis, U. pygmoidea, and the related taxa in order to test the hypotheses of species pair relationships. The samples of U. nipparensis consist of two chemotypes: chemotype 1 (usnic and caperatic acid) for Ohmura 3825, 6274, 6282, 12248, and 12249; and chemotype 2 (usnic, norstictic, caperatic, and stictic acids) for Ohmura 9054. They form a monophyletic clade with high support values (ML/NJ = 100/100) (Fig. 2). Therefore, the chemical difference seen in the U. nipparensis morphotype is certainly confirmed as a variation within a single species. All samples of U. sinensis examined by the author contain usnic, norstictic, and caperatic acids as major substances except Y. Ohmura 7408 (TNS) in which norstictic acid appeared as a faint trace in TLC. The amount of norstictic acid in U. sinensis is variable and sometimes not detected by TLC (see Ohmura 2002). In contrast, caperatic acid was not reported from the voucher specimen of the GenBank accession number FJ494953 (Shen et al. 2012). This is because it was examined by means of High Performance Liquid Chromatography (HPLC) that is generally difficult to detect fatty acids lacking benzene rings in the structure (Huneck et al. 1994). All samples of U. sinensis form a monophyletic clade with high support value (94/99) (Fig. 2). The U. nipparensis and U. sinensis clades form a monophyletic clade together with support values (60/73). This U. nipparensis -U. sinensis clade forms a monophyletic clade with U. angulata with high support values (91/92). The chemistry of U. angulata is fundamentally the same as U. sinensis, e.g., containing usnic, norstictic and caperatic acids (Ohmura 2001), although the presence of caperatic acid in U. angulata was not confirmed in some studies (e.g., Awasthi 1986;Stevens 1999;Truong et al. 2013). Caperatic acid, a fatty acid, is usually detected on TLC with water, but it is sometimes ambiguous. It would be easily detected and identified by a microcrystal test in addition to TLC (Yoshimura & Kurokawa 1976). In fact, caperatic acid was detected from U. angulata collected in South America (specimens housed in TNS) (Fig. S1), although Truong et al. (2013) did not detect it from the South American materials. Morphology of U. angulata is distinctively different from U. nipparensis and U. sinensis in having a pendulous thallus with ridged to alate branches, the presence of punctiform soralia, and the %C/%M/%A [(5.6-)9.2-15(-17)/(6.7-)8.6-19(-28)/ (25-)40-57(-61)] (Ohmura 2001(Ohmura , 2012. The tree suggests these three species might have evolved from a common ancestor, and the evolutionary order is supposed to be U. angulata, U. sinensis, and U. nipparensis from oldest to most recent (Fig. 2). Multi-locus analysis also supported the order for U. angulata and U. nipparensis (Fig. 1). Since the multi-locus genes, except ITS rDNA, were not available for U. sinensis in this study, the species was not included in the tree (Fig. 1).
Based on morphological, chemical, and molecular phylogenetic data, U. nipparensis and U. sinensis could be assumed as a 'species pair', and they are supposed to have evolved from the common ancestor of U. angulata. The 'species pair' concept is generally applied to two taxa having the same chemistry, but different reproductive strategies, one being esorediate and the other sorediate (Poelt 1972). Both U. nipparensis and U. sinensis have caperatic acid as the major compound. However, in addition to caperatic acid, U. nipparensis has ± stictic acid group and U. sinensis has norstictic and ± salazinic acids. In a strict sense, these species do not have exactly the same chemistry, but all of these additional substances are β-orcinol depsidones (Culberson 1969). The fact that the norstictic ± salazinic acids chemotype and the stictic acid group chemotype can occur within a single species, e.g., U. glabrescens var. glabrescens (Clerc & Figure 1. Phylogenetic position of Usnea nipparensis in the subgenus Usnea inferred from the multi-locus dataset of ITS rDNA, nu LSU, and MCM7 genes. The tree was constructed using the NJ method, and the reliability of each branch was tested by ML and NJ methods. The bootstrap values for ML/NJ were generated from 1,000 replicates and shown on the thick branches only when both were ≥50% simultaneously. The OTU names indicated the taxon epithet, the voucher number, and location (see Table 1). The clade names are identical to those in Truong & Clerc (2016). All positions containing gaps and missing data were eliminated. There were a total of 1,512 positions in the final dataset.
of U. nipparensis and U. sinensis might be caused by a small evolutionary event or an unknown factor.
In the species pair concept, secondary species (vegetative lineage) are assumed to have arisen from primary species (sexual lineage) through a rare transition event, and the vegetative lineage is thought to be successful due to its superior ability to colonize and survive in marginal habitats (Buschbom & Mueller 2005). This idea could be also applied to the case of U. sinensis (primary species) and U. nipparensis (secondary species) from the phylogenetic result in this study. The known distribution of U. sinensis is narrower than that of U. nipparensis: i.e., U. sinensis is collected from Yunnan in the mainland of China and Taiwan, while U. nipparensis is recorded from Japan, Taiwan, Korea, India, and Nepal (Ohmura 2001). Although the distribution of U. nipparensis is currently restricted in South and East Asia, it could be much wider because of the vegetative dispersal strategy. Indeed, U. boomiana P. Clerc, collected from the Canary Islands (van den Boom et al. 2015;G -holotype!), and the caperatic acid chemotype of U. subeciliata (Motyka) Swinscow & Krog, collected from Australia ( Fig. 67 in Stevens 1999; specimens not seen), resemble to U. nipparensis both in morphology and chemistry. Further research using molecular phylogenetic analyses may solve the relationship between U. nipparensis and these species.
Although single-or multi-locus genetic data were usually not able to resolve species pair relationships in the genus Usnea, this study shows a presumable species pair relationship between U. nipparensis and U. sinensis based on ITS rDNA. In addition, these two species were suggested to have speciated from the common ancestor U. angulata. There should be many species pairs in the genus Usnea with different evolutionary histories. Species pairs having an old evolutionary history could be clarified by single-or multi-locus genetic data. However, recently . Species pair relationship for Usnea nipparensis and U. sinensis inferred from ITS rDNA. The tree was constructed using the NJ method, and the reliability of each branch was tested by ML and NJ methods. The bootstrap values for ML/NJ were generated from 1,000 replicates and shown on the thick branches only when both were ≥50% simultaneously. The OTU names indicated the taxon epithet, the voucher number, and location (see Table 1). The clade names are identical to those in Truong & Clerc (2016). All positions containing gaps and missing data were eliminated. There were a total of 457 positions in the final dataset. speciated taxa representing a species pair should be analyzed using fine scale markers, such as microsatellite and RADseq (Grewe et al. 2018;Lagostina et al. 2018).
This study also confirmed the close relationship between U. orientalis and U. pygmoidea [as a species pair hypothesized by Shen et al. (2012)], forming a monophyletic clade with support value (62/64) (Fig. 2). However, the independency of each species is needed to be examined with further data.
Insufficient resolution in molecular phylogenetic analysis using single-or multi-locus genetic data can cause incorrect interpretations, especially when it comes to test the conspecificity as pointed out by Grewe et al. (2018), but also while testing higher taxonomic groups in which results would vary depending on the analysis performed (Truong et al. 2013;Divakar et al. 2017). Future integrated studies with traditional careful α-taxonomy and fine-scale or genomic mega data for molecular analyses may solve difficult taxonomic problems that remain among the genus Usnea.

Acknowledgements
I would like to express my sincere gratitude to P. Clerc for his kind hospitality, help and discussion of Usnea when I visited G in 1998, 2012, and 2019; to G. Kokubugata for providing a fresh specimen of Usnea; and to anonymous reviewers for their valuable constructive comments and linguistically revising this manuscript. Figure S1. Caperatic acid crystalized by microcrystal test with GE (glycerin: acetic acid, 1:3). Download file File S1. Detailed information of specimens examined and housed in TNS and NIPR. Download file File S2. Final alignment for the multi-locus data set removing sites with gaps, missing data and ambiguous data. Download file File S3. Final alignment for the ITS rDNA data removing sites with gaps, missing data and ambiguous data. Download file