﻿The phylogeny and taxonomy of Glypholecia (Acarosporaceae, lichenized Ascomycota), including a new species from northwestern China

﻿Abstract Glypholeciaqinghaiensis An C. Yin, Q. Y. Zhong & Li S. Wang is described as new to science. It is characterized by its squamulose thallus, compound apothecia, ellipsoid ascospores, and the presence of rhizines on the lower surface of the thallus. A phylogenetic tree of Glypholecia species was constructed based on nrITS and mtSSU sequences. Two species G.qinghaiensis and G.scabra are confirmed in China.

The first known specimen of the genus now designated Glypholecia, was collected in the summer of 1810, by Balbis G. B., who was working at the Botanical Gardens of Turin, Italy. Balbis sent this lichen specimen to Germany, where it was received by Funck H. C., who labelled the specimen as Balbis. This name was never published. Later, Persoon H. C. examined this same specimen in France, and described it as a new species Urceolaria scabra Pers. (Persoon 1810).
Accordingly, most lichenologists accept only one species G. scabra (Pers.) Müll. Arg. as belonging to the genus Glypholecia. The distribution of G. scabra is largely disjunct, with intercontinental populations. It occurs on siliceous rocks in desert, alpine regions, including Europe, Africa, Asia, and North America (Magnusson 1940;Thomson 1979Thomson , 1984Ryan 2002). It exhibits extensive morphological diversity across its range.
Previously, a second species, G. tibetanica H. Magn., was described as endemic to China. It has only been recorded from Xizang province, China (Zahlbruckner 1933). However, Obermayer (2004) suggested that G. tibetanica might belong to Acarospora nodulosa var. reagens (Zahlbr.) Clauzade and Cl. Roux, which has also been reported from China, based on the K reaction of the cortex. But Magnusson H. appears to have incorrectly reported a 'C' reaction (to calcium hypochlorite) as a KOH reaction (potassium hydroxide). According to Zahlbruckner's (1933) description, the cortex of G. tibetanica should show a reddish C+ reaction due to gyrophoric acid, whereas A. nodulosa var. reagens shows a red K+ reaction caused by norstictic acid (Magnusson 1940;Zahlbruckner 1933;Cao and Wei 2009). Unfortunately, the holotype of G. tibetanica has been lost, so there are no materials available for further research regarding this putative species.
During the Second Tibetan Plateau Scientific Expedition and Research Program (STEP), we collected numerous lichen specimens from across northwestern China. The expedition included many field surveys across the type locality of G. tibetanica, during which we discovered some specimens that conformed to Zahlbruckner's (1932) description. Our new specimens differ from A. nodulosa var. reagens in the characteristics of their upper cortex, hymenium chemical reaction and their number of ascospores. The molecular sequences obtained from these new specimens demonstrate that they belong to the genus Glypholecia, rather than to Acarospora (Knudsen et al. 2020). Of these, some samples had squamulose thalli, compound apothecia and white rhizines. These were confirmed as belonging to a new species of the genus Glypholecia. In this paper, we describe a new species, Glypholecia qinghaiensis, from northwestern China.

Morphological and chemical study
We examined materials of Glypholecia from the lichen herbaria of the Kunming Institute of Botany (KUN-L), the College of Life Science and Technology, Xinjiang University (XJU), and the Swiss Federal Institute for Forest, Snow and Landscape Research, Switzerland (WSL). We made morphological observations of the specimens using a Nikon SMZ 745T (Nikon Corp., Tokyo, Japan) dissecting microscope. We cut vertical sections of apothecia and thalli using a razor blade, mounted sections in GAW (glycerol: ethanol: water = 1: 1: 1), then examined them under a Nikon Eclipse 50i stereomicroscope. We measured the average spore size and described sections under both the microscope and stereomicroscope. We photographed all specimens with a Nikon digital camera head DS-Fi2. We identified secondary metabolites by their color reaction coupled with thin-layer chromatography (TLC), using solvent system C (toluene: acetic acid = 85:15), following the methods of Culberson (1970) and Orange et al. (2001).

Phylogenetic analysis
We aligned DNA sequences using the program MAFFT v. 7.107 in GENEIOUS v. 8.0.2, setting the following parameters: algorithm = auto; scoring matrix = 200 PAM / k=2; gap open penalty = 1.53; offset value = 0.123 (Katoh et al. 2005). We conducted single-gene analyses to test for potential incongruence among the two-gene fragments, using maximum likelihood (ML) analyses and Bayesian inference (BI). We generated a matrix of Glypholecia and its related genera using GENEIOUS v. 8.0.2. In addition to the DNA sequences obtained from our own material, we also downloaded all available sequences of Glypholecia from GenBank at the National Center for Biotechnology Information (NCBI, https://www.ncbi. nlm.nih.gov/), and added these to the matrix. We performed ML analyses within RaxML v. 8.2.12 (Stamatakis 2014), using the General Time Reversible model of nucleotide substitution with the gamma model of rate heterogeneity (GTRGA-MMA). We selected best partitioning scheme and evolutionary models for two pre-defined partitions using PartitionFinder2 (Lanfear et al. 2016), with greedy algorithm and AICc criterion. We used PhyloSuite (Zhang et al. 2020) inferred Bayesian Inference phylogenies using MrBayes 3.2.6 (Ronquist et al. 2012) under a partition model (2 parallel runs, 10 million generations), for which the initial 25% of sampled data were discarded as burn-in. We inferred support values from the 70% majority-rule tree of all saved trees obtained from 1000 non-parametric bootstrap replicates. We obtained Posterior Probabilities (PPs) from the 95% majority rule consensus tree of all saved trees. We visualized tree files using FigTree 1.4.4.

Results
In the present study we generated twenty new nrITS and eighteen new mtSSU sequences. We constructed ML and BI topologies based on these nrITS and mtSSU sequences, and nine additional sequences downloaded from GenBank (Table 1). We used Pleopsidium as the outgroup (Crewe et al. 2006;Reeb et al. 2007;Schmull et al. 2011). In the phylogenetic tree, the Glypholecia specimens formed a monophyletic lineage, which was divided into two clades, representing G. scabra in Clade 2 and the new species G. qinghaiensis in Clade 1. The result showed high support for new specimens in this study being assigned to Glypholecia (99% ML and 1.00 PP, Fig. 3). Our phylogenetic results indicated that Glypholecia is monophyletic in China. The genus Glypholecia is characterized by its squamulose to crustose thallus, compound apothecia forming multiple structures, multi-spored asci (usually exceeding 30 spores per ascus), spherical small ascospores and C+ red reaction of the cortex due to the presence of gyrophoric acid.
Species of Glypholecia were separated into two main clades, as inferred from the phylogenetic tree with strong support. Based on the combination of morphological characters and phylogenetic analysis, we propose a new species in Glypholecia. We divide the specimens of the genus Glypholecia collected from China into two clades, which correspond to differences in the lower surface morphology of the thallus: the specimens in Clade 1 have ellipsoid ascospores (2.5-3 × 4-6.5 μm), and rhizines on the lower surface of the thallus, whereas Clade 2 contains specimens with spherical ascospores (3-4 µm), and a central holdfast, called an umbilicus. Within Clade 2, our samples of G. scabra collected from China were clustered with the European samples, but with some genetic divergence. All Glypholecia specimens which we collected from the type locality of "G. tibetanica" were monophyletic with G. scabra, with a high support value. Within Clade 1, those specimens assigned as G. qinghaiensis lack genetic variability within the gene regions included in this study. This might be due to either the sparse population or the shared geography and environment. All of our G. qinghaiensis specimens were collected from arid to semi-arid areas of northwestern China, usually growing on sandy rock or sandy soil.
Our phylogenetic analyses showed that a specimen from Gansu province in China, which was identified as G. scabra by Cao and Wei (2009), was clustered with Acarospora schleicheri (the type species of Acarospora) in Clade 3. Therefore, this specimen's previous identification as G. scabra was incorrect.
Ecology and distribution. Usually on sandy rocks or rarely on soil over rocks; so far only known from Gansu, Ningxia, Qinghai and Xizang provinces in China; growing in arid to semi-arid areas; distributed from 1600 to 4700 m altitude.
Etymology. The epithet "qinghaiensis" refers to the holotype locality of the species.
Notes. Glypholecia qinghaiensis can be distinguished from G. scabra by having ellipsoid ascospores and rhizines. Phylogenetic analysis in this study supports the separate classification of these two species. This species has only been reported from northwestern China, including Gansu, Ningxia, Qinghai and Xizang provinces.
Glypholecia scabra is characterized by its squamulose thallus, abundant and compound apothecia, and umbilicate lower surface, as shown in Fig. 2A-F, J-L. In the Qinghai-Tibetan Plateau, these specimens have an umbilicus at the center of the lower surface, formed by fasciculate white rhizines, which are single or branched, dense or loose. It has a global distribution (see citations above). Within China it has been reported from Gansu (Magnusson 1940;Cao and Wei 2009), Xinjiang (Wang 1985;Abbas et al. 1993;Wu 1994 1998), Ningxia (Liu and Wei 2013) and Xizang (Wei and Jiang 1986) provinces. Delimiting species boundaries can be complicated by the potential role of the environment in shaping morphology. Our phylogenetic study showed that a specimen previously collected and reported from Gansu (GenBank number: FJ919810) did not belong to the genus Glypholecia, but should rather be placed within Acarospora. We also report a new record for G. scabra in Neimenggu province, northern China. For further synonyms and detailed descriptions of G. scabra, see Thomson (1979) and Ryan (2002).
We found some Glypholecia specimens with numerous pycnidia, as shown in Fig. 2G-H, but rarely with apothecia. Their lower surface usually had fasciculate rhizines aggregated into an umbilicate. Although morphological characters, including the shape (bacilliform), size (3.5-4.5 × 1 µm) of the conidia, and spot reaction (cortex and medulla C+ red) are consistent with those of Zahlbruckner's G. tibetanica, the molecular data show that these specimens instead belong to G. scabra. In 2019, we also searched the herbarium of the Zurich Botanical Garden for the holotype of G. tibetanica which had been deposited by Walter Bosshard (Zahlbruckner 1933). The holotype could not be located. Therefore, we propose that G. tibetanica could potentially be treated as a synonym of G. scabra. Further research is required to determine their synonymy. At present, as the holotype could not be examined, the species name G. tibetanica should be maintained.