Introduction
Delimitation of species in lichen-forming fungi has changed dramatically with the availability of DNA sequence data (reviewed in Crespo & Lumbsch Reference Crespo and Lumbsch2010; Lumbsch & Leavitt Reference Lumbsch and Leavitt2011; Leavitt et al. Reference Leavitt, Moreau, Lumbsch, Upreti, Divakar, Shukla and Bajpai2015). Within the Parmeliaceae, the largest family of lichen-forming fungi that currently includes c. 2800 species worldwide (Kraichak et al. Reference Kraichak, Huang, Nelsen, Leavitt and Lumbsch2018), numerous cryptic lineages have been detected. In fact, the estimate by Crespo & Lumbsch (Reference Crespo and Lumbsch2010) of 80 cryptic lineages in parmelioid lichens hidden under widely distributed species seems, about a decade later, too conservative. Recently there has been an increased interest in improving the understanding of species delimitation in tropical lineages, resulting in the discovery and description of new clades, primarily based on molecular data (Parnmen et al. Reference Parnmen, Rangsiruji, Mongkolsuk, Boonpragob, Nutakki and Lumbsch2012; Moncada et al. Reference Moncada, Lücking and Coca2013; Kirika et al. Reference Kirika, Divakar, Crespo, Mugambi, Orock, Leavitt, Gatheri and Lumbsch2016a, Reference Kirika, Divakar, Crespo, Gatheri, Mugambi, Leavitt, Moncada and Lumbschb, Reference Kirika, Divakar, Buaruang, Leavitt, Crespo, Gatheri, Mugambi, Benatti and Lumbsch2017, Reference Kirika, Divakar, Crespo and Lumbsch2019; Singh et al. Reference Singh, Aptroot, Rico, Otte, Divakar, Crespo, Caceres, Lumbsch and Schmitt2018). Given that tropical regions are biodiversity hot spots and are among the most species-rich areas for lichenized fungi, a better understanding of the delimitation of species in the tropics is crucial for gaining insight into global fungal diversity (Hawksworth Reference Hawksworth2012; Hawksworth & Lücking Reference Hawksworth and Lücking2017).
Canoparmelia Elix & Hale is a medium-sized genus consisting of c. 40 species in the parmelioid group, belonging to the parmotremoid clade (Crespo et al. Reference Crespo, Kauff, Divakar, Del-Prado, Pérez-Ortega, Amo de Paz, Ferencova, Blanco, Roca-Valiente and Núñez-Zapata2010b). Species in the genus are characterized by having relatively narrow, subirregular lobes with rounded or subrounded eciliate margins, a pored epicortex, the presence of isolichenan in the cell walls, bifusiform conidia and simple rhizines (Elix Reference Elix1993; Crespo et al. Reference Crespo, Kauff, Divakar, Del-Prado, Pérez-Ortega, Amo de Paz, Ferencova, Blanco, Roca-Valiente and Núñez-Zapata2010b). Canoparmelia is widely distributed throughout the tropical and subtropical regions of the Old and New Worlds. In its original circumscription (Elix et al. Reference Elix, Johnston and Verdon1986), Canoparmelia was found to be highly polyphyletic with species transferred to other genera, including Austroparmelina A. Crespo et al. (Crespo et al. Reference Crespo, Ferencova, Pérez-Ortega, Argüello, Elix and Divakar2010a), Parmotrema A. Massal. and Crespoa (D. Hawksw.) Lendemer & B. P. Hodk. (Crespo et al. Reference Crespo, Kauff, Divakar, Del-Prado, Pérez-Ortega, Amo de Paz, Ferencova, Blanco, Roca-Valiente and Núñez-Zapata2010b; Hawksworth Reference Hawksworth2011; Lendemer & Hodkinson Reference Lendemer and Hodkinson2012; Kirika et al. Reference Kirika, Divakar, Crespo, Mugambi, Orock, Leavitt, Gatheri and Lumbsch2016a; Divakar et al. Reference Divakar, Crespo, Kraichak, Leavitt, Singh, Schmitt and Lumbsch2017). Kirika et al. (Reference Kirika, Divakar, Crespo, Mugambi, Orock, Leavitt, Gatheri and Lumbsch2016a) identified a core group of Canoparmelia, which formed a sister relationship to the rest of the genera included in the parmotremoid clade. Canoparmelia s. str. is sister to the Xanthoparmelia clade and diverged c. 48 million years ago (Divakar et al. Reference Divakar, Crespo, Wedin, Leavitt, Hawksworth, Myllys, McCune, Randlane, Bjerke and Ohmura2015, Reference Divakar, Wei, McCune, Cubas, Boluda, Leavitt, Crespo, Tchabanenko and Lumbsch2019). Canoparmelia texana (Tuck.) Elix & Hale is the type species of the genus and is common throughout the tropics extending into the temperate zone, and is common in Kenya, where many samples for this study originated. It is characterized by having a sorediate upper surface, eciliate lobe margins and containing divaricatic and nordivaricatic acids. Morphological variations in lobe configuration, thallus size and fertility have been noted in a previous study (Divakar & Upreti Reference Divakar and Upreti2005). Given the wide distribution of this taxon, the high level phenotypic variation across its range and previous studies in other clades where cryptic lineages were found, we sampled material of C. texana in order to examine the species delimitation of this widespread tropical to warm-temperate species using a three-locus data set.
Materials and Methods
Taxon sampling
The analyzed data matrices included 30 samples comprising six species of Canoparmelia and four outgroup taxa, focusing on recently collected samples from East Africa. A DNA data matrix of nuLSU, ITS and mtSSU rDNA sequences was used to infer evolutionary relationships. Thirty sequences were newly generated for this study. Five samples were used as outgroup taxa, including two samples of Nesolechia oxyspora (Tul.) A. Massal. and three of Xanthoparmelia (Vain.) Hale (X. chlorochroa (Tuck.) Hale, X. exornata (Zahlbr.) Brusse & M. D. E. Knox and X. saxeti (Stizenb.) G. Amo et al.). Information on studied materials, including GenBank Accession numbers, is provided in Table 1.
Specimens of Canoparmelia, and other Parmeliaceae species used in this study, with voucher information and GenBank Accession numbers. Newly obtained sequences for this study are in bold and missing data are indicated with a dash (—).
DNA extraction and PCR amplification
Total genomic DNA was extracted from small pieces of thallus devoid of any visible damage or contamination using the USB PrepEase Genomic DNA Isolation Kit (USB, Cleveland, OH, USA). We generated sequence data from three nuclear ribosomal markers: the ITS region, a fragment of nuLSU, and a fragment of the mtSSU. Polymerase chain reaction (PCR) amplifications were performed using Ready-To-Go PCR Beads (GE Healthcare, Pittsburgh, PA, USA) using dilutions of total DNA. Fungal ITS rDNA was amplified using primers ITS1F (Gardes & Bruns Reference Gardes and Bruns1993), ITS4 and ITS4A (White et al. Reference White, Bruns, Lee, Taylor, Innis, Gelfand, Sninsky and White1990; Larena et al. Reference Larena, Salazar, González, Julián and Rubio1999), nuLSU rDNA was amplified using LR0R and LR5 (Vilgalys & Hester Reference Vilgalys and Hester1990), and mtSSU rDNA was amplified using the primers mrSSU1, mrSSU3R and mrSSU2R (Zoller et al. Reference Zoller, Scheidegger and Sperisen1999). The primer combination ITS1F and ITS4A was used when the universal primer ITS4 failed to amplify the ITS region. Polymerase chain reaction products were visualized on 1% agarose gel and cleaned using ExoSAP-IT (USB, Cleveland, OH, USA). Cycle sequencing of complementary strands was performed using BigDye v. 3.1 (Applied Biosystems, Foster City, CA, USA) and the same primers as used for PCR amplifications. Sequenced PCR products were run on an ABI 3730 automated sequencer (Applied Biosystems) at the Pritzker Laboratory for Molecular Systematics and Evolution at the Field Museum, Chicago, and at the Unidad de Genómica (Parque Científico de Madrid).
Sequence editing and alignment
New sequences were assembled and edited using Geneious v. 8.1.9 (Kearse et al. Reference Kearse, Moir, Wilson, Stones-Havas, Cheung, Sturrock, Buxton, Cooper, Markowitz and Duran2012). Multiple sequence alignments for each locus were performed using the program MAFFT v. 7 (Katoh & Standley Reference Katoh and Standley2013). For the ITS and nuLSU sequences, we used the G-INS-i alignment algorithm and ‘20PAM/K = 2’ scoring matrix, with an offset value of 0.3 and the remaining parameters set to default values. We used the E-INS-i alignment algorithm and ‘20PAM/K = 2’ scoring matrix, with the remaining parameters set to default values, for the mtSSU sequences. The program Gblocks v. 0.91b (Talavera & Castresana Reference Talavera and Castresana2007) was used to delimit and remove ambiguous nucleotide positions from the final alignments using the online web server (http://molevol.cmima.csic.es/castresana/Gblocks_server.html), implementing the options for a less stringent selection of ambiguous nucleotide positions, including the ‘Allow smaller final blocks’, ‘Allow gap positions within the final blocks’, and ‘Allow less strict flanking positions’ options.
Phylogenetic analyses
Phylogenetic relationships were inferred using maximum likelihood (ML) and Bayesian inference (BI). Exploratory phylogenetic analyses of individual gene topologies showed no evidence of well-supported (≥ 70% bootstrap values) topological conflict, so relationships were estimated from a concatenated, three-locus (ITS, nuLSU and mtSSU) data matrix using a total-evidence approach (Wiens Reference Wiens1998). RAxML v. 8.1.11 (Stamatakis Reference Stamatakis2014) was implemented to reconstruct the concatenated ML gene tree using the CIPRES Science Gateway server (http://www.phylo.org/portal2/) and the ‘GTRGAMMA’ model was used, with locus-specific model partitions treating all loci as separate partitions, and evaluated nodal support using 1000 bootstrap pseudoreplicates. Exploratory analyses using alternative partitioning schemes resulted in identical topologies and similar bootstrap support values. We also reconstructed phylogenetic relationships from the concatenated multilocus data matrix under BI using the program BEAST v. 1.8.2 (Drummond & Rambaut Reference Drummond and Rambaut2007). We ran two independent Markov chain Monte Carlo (MCMC) chains for 20 million generations, implementing a relaxed lognormal clock, with a birth-death speciation process prior. The most appropriate model of DNA sequence evolution was selected for each marker using PartitionFinder v. 1.1.1 (Lanfear et al. Reference Lanfear, Calcott, Ho and Guindon2012), treating the ITS1, 5.8S, ITS2, nuLSU and mtSSU as separate partitions. The first two million generations were discarded as burn-in. Chain mixing and convergence were evaluated using the effective sample size (ESS) values > 200 as a good indicator. Posterior trees from the two independent runs were combined using LogCombiner v. 1.8.0 (Drummond et al. Reference Drummond, Suchard, Xie and Rambaut2012), and the final maximum clade credibility (MCC) tree was estimated from the combined posterior distribution of trees.
Morphological and chemical studies
Morphological and anatomical characters were studied using a Leica Wild M8 dissecting and Leica Leitz DM RB compound microscope. Chemical constituents were identified by high performance thin-layer chromatography (HPTLC) using standard methods (Arup et al. Reference Arup, Ekman, Lindblom and Mattson1993) with a Camag horizontal developing chamber (Oleico Laboratory, Stockholm) using solvent system C.
Results and Discussion
A total of 30 sequences, including 13 nuclear ITS, 9 nuLSU and 8 mitochondrial SSU rDNA from 14 samples of Canoparmelia, were generated in this study and uploaded to GenBank (Table 1). The aligned data matrix contained 444 unambiguously aligned nucleotide position characters in ITS, 741 in nuLSU and 752 in mtSSU. The final alignment of the three-locus concatenated data set was 1937 positions in length, with 383 variable characters. TNe + G4, TNe + I and HKY + F + G4 were selected as the best fit models of evolution for the ITS, nuLSU and mtSSU data sets, respectively.
The single locus trees demonstrated no supported conflicts (results not shown) and therefore the concatenated three-locus data matrix (ITS, nuLSU and mtSSU) was analyzed. The partitioned ML analysis of the concatenated data matrix resulted in an optimal tree with ln likelihood value = −6485.867 (Fig. 1). Maximum likelihood and Bayesian topologies were largely similar and did not show any supported conflict (e.g. PP ≥ 0.95 and ML bootstrap ≥ 70%), and therefore the ML tree topology is depicted here with the Bayesian posterior probabilities added (Fig. 1). We consider PP ≥ 0.95 and ML bootstrap ≥ 70% as strong support for nodes.
Phylogenetic relationships of Canoparmelia species based on maximum likelihood (ML) and Bayesian analyses of a concatenated, three-locus data set (ITS, nuLSU and mtSSU rDNA). The ML tree is shown here. Posterior probabilities ≥ 0.95 from the Bayesian analysis and ML bootstrap values ≥ 70% are given above branches. Information for the specimens used in this analysis are given in Table 1.
Canoparmelia albaniensis, habit (P. Kirika 4649). In colour online
Our samples of Canoparmelia texana do not form a monophyletic group, but cluster into two well-supported clades (clades 1 and 2 in Fig. 1). Clade 2 forms a sister-group relationship with the apotheciate C. nairobiensis (J. Steiner & Zahlbr.) Elix & Hale. However, this relationship lacks strong support. The African endemic C. nairobiensis has been hypothesized to be the esorediate progenitor of C. texana (Hale Reference Hale1976). Clades 1 and 2 together with C. nairobiensis form a supported monophyletic group and this clade forms a strongly supported sister group with isidiate C. ecaperata (Müll. Arg.) Elix & Hale and one sample of C. caroliniana (Nyl.) Elix & Hale from Kenya. The latter species is also polyphyletic with the other two samples of C. caroliniana from the USA, forming a well-supported sister group with C. austroamericana Adler. Canoparmelia eruptens (Kurok.) Elix & Hale is the earliest diverging clade within the strongly supported, monophyletic genus Canoparmelia, but this relationship is supported only in the ML analysis.
The present investigation supports a previous study (Kirika et al. Reference Kirika, Divakar, Crespo, Mugambi, Orock, Leavitt, Gatheri and Lumbsch2016a) indicating that the species delimitation in Canoparmelia requires revision. We have re-examined the secondary chemistry and morphology of the samples of both major clades found in C. texana. The chemistry of all samples was similar, with atranorin, chloroatranorin and divaricatic acid present in all specimens, whereas the presence of nordivaricatic acid differed. Specimens in both major clades could have or lack the latter substance, which is closely related to divaricatic acid, and its absence from TLC plates might also be due to a lack of sensitivity of the analytical methods.
A re-examination of phenotypic features, including substratum specificity of samples from both Canoparmelia texana clades, revealed subtle morphological differences. The samples of clade 1 had a smaller ascospore size (7.5–10 μm long), which fits well within the ascospore range of C. texana (9–11 μm in length; Hale Reference Hale1976), and conspicuous maculae on the upper thallus surface. Furthermore, as the sample from the type locality (Texas), belongs to clade 1 we here consider this clade to be C. texana s. str. The samples grouped in clade 2 had a relatively larger ascospore size (11–14.5 μm long) and inconspicuous maculae on the upper thallus surface. However, as we have examined only a small number of samples, a larger sampling effort will be needed to evaluate whether or not these phenotypic differences are consistent between the two clades. All other characters showed no significant differences.
Subsequently, we investigated available names that are currently considered synonyms of Canoparmelia texana. In most cases the ascospore size of the types suggested that these names are indeed synonyms of C. texana, with the exception of Parmelia albaniensis C. W. Dodge which has ascospores 11–13.0 μm in length. Therefore, we propose to use this name to accommodate specimens of clade 2, and the name is transferred to the genus Canoparmelia below.
Taxonomic Treatment
Canoparmelia albaniensis (C. W. Dodge) Divakar & Kirika comb. nov.
MycoBank No.: MB 841885
Parmelia albaniensis C. W. Dodge, Ann. Miss. Bot. Gard. 46, 121 (1959); type: South Africa, Cape of Good Hope, forests of Albany, Zeyher 3 (FH (Taylor Herbarium)—holotype!).
Thallus foliose, adnate, ash grey or grey-green, lobe margin often tinged with brown. Lobes 3–7 mm wide, crenate or deeply incised, eciliate, sometimes imbricate or lobulate, margins usually turned down. Upper cortex pitted, maculate, and rugose. Medulla white. Lower cortex black, with narrow, brown, naked marginal zone, rhizines simple, black, often tipped with brown or white. Soralia laminal, punctiform or originating from low pustules, coalescing in older parts of the thallus.
Apothecia rare, laminal, thalline margin sorediate; asci 8-spored; ascospores 11.0–14.5 × 6.0–7.5 μm, rarely biguttulate.
Conidia weakly bifusiform, 6–8 μm long.
Secondary chemistry
Divaricatic acid, nordivaricatic acid (medulla C+ pale rose, KC+ purple), atranorin and chloroatranorin.
Ecology and distribution
Corticolous, rarely saxicolous, common in urban habitats and well-lit sites in dry, lowland forested areas to lower montane forests (1100–2600 m). Currently known from Argentina, China, Kenya and south-eastern United States (see clade 2 of Supplementary Material Fig. S1, available online), but it is probably overlooked and has been confused with C. texana s. str.
Notes
Canoparmelia albaniensis can easily be confused with C. texana in the field, but the former differs in having larger ascospores (11.0–14.5 μm long) and inconspicuous maculae on the upper surface. Furthermore, in molecular phylogenetic reconstruction C. albaniensis does not form a sister relationship with C. texana but with a non-sorediate African species, C. nairobiensis (Fig. 1). It is also morphologically similar to C. aptata (Kremp.) Elix & Hale, which differs in containing perlatolic acid.
Although Dodge (Reference Dodge1959) reported the medulla C−, KC− on the type material of Parmelia albaniensis C. W. Dodge, in the re-examination we found it C+ rose, KC+ purple.
Additional specimens examined
Kenya: Kakamega Co.: Kakamega Forest, Isecheno Forest Station, tropical rainforest, 1760 m, 0°14ʹN, 34°52ʹE, on bark, 2013, P. Kirika 3424 (EA). Nyeri Co.: Mt Kenya, Naro Moru route, 4 km from Park gate towards Met. station, Podocarpus-bamboo forest, 2561 m, 0°10ʹS, 37°09ʹE, on bark, 2014, P. Kirika 4391 (EA). Kajiado Co.: Karen, Ololua Forest, disturbed dry upland forest with Olea, Croton, Calodendrum, Schrebera, 1800 m, 1°21ʹS, 36°41ʹE, on bark, 2018, P. Kirika 5335 (EA). Baringo Co.: Rift Valley, Eldama Ravine, Lembus Forest, off Eldama Ravine-Eldoret Road, remnant montane forest, 2137 m, 0°16ʹN, 35°75ʹE, on bark, 2013, P. Kirika, G. Mugambi & H. T. Lumbsch 2817 (EA, F). Makueni Co.: Utu, Chyulu Hills National Reserve, dry rocky woodland, 1150 m, 2°40ʹS, 37°57ʹE, on bark, 2014, P. Kirika 4617 (EA); Chyulu Hills National Reserve, Chyulu-2 near ranger's post, woodland with Erythrina abyssinica and Olea europaea, 1430 m, 2°44ʹS, 37°56ʹE, on bark, 2014, P. Kirika 4649 (EA).
Acknowledgements
Extracted DNA sequences were generated in the Pritzker Laboratory for Molecular Systematics and Evolution at the Field Museum and at the Molecular Laboratory, Department of Pharmacology, Pharmacognosy and Botany, Faculty of Pharmacy, Complutense University of Madrid. We thank Isabel DiStefano (Chicago) for laboratory work in Chicago and Todd Widhelm (Chicago) for support with obtaining fresh material from Texas. We thank Steven Leavitt for generating sequences of C. texana from the type locality. We are grateful to Genevieve Tocci (Cambridge, Massachusetts) for sending us the type of Parmelia albaniensis on loan for examination. Jason Singhurst and Colton Nolen (Texas Parks and Wildlife Division) provided access and permission to collect specimens at Fawcett Wildlife Management Area in Palo Pinto County, Texas. This study was supported by grants from the IDP/Africa Training Fund and the Bill Stanley Memorial Fund at the Field Museum, the Spanish Ministerio de Ciencia e Innovación (PID2019-105312GB-I00), and the Santander-Universidad Complutense de Madrid (PR87/19-22637 and G/6400100/3000). The herbarium MAF–Lich is sincerely thanked for providing herbarium numbers and samples for this study.
Author ORCIDs
Pradeep K. Divakar, 0000-0002-0300-0124; H. Thorsten Lumbsch, 0000-0003-1512-835X.
Supplementary Material
To view Supplementary Material for this article, please visit https://doi.org/10.1017/S0024282922000135.
Open access
