Here we provide one of the first detailed studies of lichen and allied fungi diversity in a continental North American city (Edmonton, Alberta, Canada), including an annotated checklist, images of all species, dichotomous keys, and local distribution maps. Edmonton is the northernmost city in North America with a population of over one million, and an industrial and transportation gateway for much of northern Canada. Lichen-based biomonitoring could be a tool to track airborne pollutants resulting from Edmonton’s growing populace and industrial activity. The first step towards such a program is documenting the diversity and distribution of lichens in the city. To accomplish this, we conducted a city-wide, systematic survey of 191 sites focused on epiphytes growing on deciduous boulevard trees. We augmented that survey with surveys of rare trees, opportunistic collections from river valley and ravine habitats, herbarium collections, phylogenetic analyses of a subset of collections, and observations submitted to online nature-reporting applications. We present ITS sequence barcode data for 33 species, phylogenetic analyses for Candelariaceae, Endocarpon, Flavopunctelia, the Lecanora dispersa group, Lecidella, Peltigera, Physconia, and Punctelia, and detailed descriptions of 114 species in 47 genera and 23 families. Two species are hypothesized to be new to North America (Endocarpon aff. unifoliatum, Lecidella albida), twelve more are new to Alberta (Amandinea dakotensis, Bacidia circumspecta, Candelaria pacifica, Candelariella antennaria, Heterodermia japonica, Lecania naegelii, Lecanora sambuci, Lecanora stanislai, Lecidea erythrophaea, Peltigera islandica, Phaeocalicium aff. tremulicola, and the introduced Xanthoria parietina), and five are putative new species to science (Physcia aff. dimidiata, Physcia aff. stellaris, Phaeocalicium sp., Phaeocalicium aff. tremulicola, Lichenaceae sp.). Illustrations are provided for all species to aid in verification and public outreach. Species richness was highest in foliose lichens (48), followed by crustose and calicioid lichens and allied fungi (41), with the lowest richness in fruticose lichens (25). We did a preliminary assessment of the suitability of species for citizen-science biomonitoring by assessing their distribution across the city, perceptibility to the public, identification accuracy, and, for a subset, how consistently species were surveyed by trained novices. Compared to other urban areas where lichen diversity has been studied, Edmonton is relatively species-rich in calicioids and Peltigera. Promising bioindicators may be limited to chlorolichens, including Caloplaca spp., Evernia mesomorpha, Flavopunctelia spp., Phaeophyscia orbicularis, Physcia adscendens, Physcia aipolia group, Physcia aff. stellaris, Usnea spp., and Xanthomendoza fallax. Other genera that may be responsive to pollutants such as Cladonia and Peltigera were almost exclusively restricted to river valley and ravine ecosystems, limiting their application as bioindicators. Some species commonly used as biomonitors elsewhere were too rare, small, poorly developed, or obscured by more common species locally (e.g., Candelaria concolor s.l., Xanthomendoza hasseana). The low overlap with lists of biomonitoring species from other regions of North America illustrates the necessity of grounding monitoring in knowledge of local diversity. Future augmentation of this list should focus on enhanced sampling of downed wood-, conifer-, and rock-dwelling lichens, particularly crustose species. The next step in developing a biomonitoring program will require modelling species’ responses to known air quality and climatic gradients.

Molecular methods. – To verify or aid in the identification of a subset of collections, the internal transcribed spacer (ITS ribosomal DNA; internal transcribed spacer regions 1 and 2 as well as the embedded 5.8S region of the ribosomal rDNA and adjacent sections of the large and small ribosomal subunits, LSU and SSU) was Sanger sequenced by T. Spribille’s lab at the U of A. ITS is the single most sequenced locus in fungi and widely used as a barcode (Hoffman & Lendemer 2018, Schoch et al. 2012,). DNA was extracted using the Qiagen DNeasy Plant Mini Kit following the manufacturer’s instructions, or, in the case of sparse material, the QIAmp DNA Investigator Kit. PCR was performed using ITS1-F (Gardes & Bruns 1993) and ITS4 primers (White et al. 1990), and the KAPA 3G Plant PCR Kit (KAPA Biosystems). The PCR cycle used was: pre-denaturation for 5 min at 95°C, 35 cycles of amplification, each cycle 30 sec at 95°C, 30 sec at 57°C, and 30 sec at 72°C. After the 35 cycles, extension occurred over 7 min of 72°C and then samples were stored at 4°C. PCR products were visualized on agarose gel after electrophoresis and sent for sequencing if a product was seen. Samples with multiple bands were not sent for sequencing due to poor chance of a clear sequence. Prior to sequencing, samples were purified using standard ExoSap protocol. PCR products were sequenced by Psomagen, Inc., USA, and forward sequences were visually examined for errors or ambiguities prior to screening.

Phylogenetic analyses. – We screened sequences with BLAST searches against the NCBI nucleotide database to identify sequences that may represent non-target organisms (NCBI Resource Coordinators 2018). The sequences generated for this study were complemented with sequences from GenBank representing additional species and specimens, as well as a small number of sequences from the senior author. For queried sequences of species adequately represented in GenBank, we report similarity metrics with accessioned sequences in the annotated species list. Further analyses including de novo tree construction were conducted for Candelariaceae, Endocarpon, Lecidella, Flavopunctelia, Lecanora dispersa group, Peltigera, Physconia, and Punctelia, as BLAST results were insufficient.

For genera requiring phylogenetic analyses, the following steps were common across analyses; specifics for each phylogeny are provided below. Sequences for each analysis were aligned with our query sequence(s) using MAFFT via a web platform (MAFFT ver. 7.49, Katoh et al. 2002, Katoh & Standley 2013, Katoh et al. 2019) or in MegAlign Pro v. 17 (DNASTAR 2021), and visually inspected in BioEdit 7.7.1 (Hall 1999). We used ITSx 1.1 (Bengtsson-Palme et al. 2013) to split sequences into ITS, small subunit, and large subunit files to aid in sequence vetting and where appropriate create partitions for nucleotide substitution model fitting. We visually examined final alignments in BioEdit and trimmed all sites from the alignment present in ≤10% of sequences. Alignments were screened using GUIDANCE2 for ambiguous sites, and analyses were completed with and without ambiguous regions and the resultant trees visually compared. Original fasta files and final alignments are deposited in Dryad, and sequence data are provided in Supplementary Appendix 2. We generated maximum likelihood phylogenetic trees in W-IQ-TREE 1.6.12 (Nguyen et al. 2015, Trifinopoulos et al. 2016) via http://iqtree.cibiv.univie.ac, specifying partitions (partition model: Chernomor et al. 2016), linked branch lengths, automatic model selection (ModelFinder: Kalyaanamoorthy et al. 2017), and free rate heterogeneity. Branch support was analyzed by 1,000 ultrafast bootstraps (UFBoot: Hoang et al. 2018) as well as SH-aLRT single branch tests with 1,000 replicates. Trees were visualized and organized in Dendroscope 3.7.6 (Huson & Scornavacca 2012) and/or MegAlign Pro, and exported to Microsoft Office Professional Plus Powerpoint 2016 for editing.

The Candelariaceae phylogeny was generated de novo with seven new sequences from the senior author, GenBank sequences with high BLAST similarity to our new sequences, and sequences from Westberg et al. (2011), Liu & Hur (2018), and Liu et al. (2019). Additional sequences for Candelaria were added from GenBank to increase taxon sampling in that clade. The Endocarpon phylogeny was generated de novo using two new sequences from the senior author, GenBank sequences with high BLAST similarity to our new sequences, and sequences from Zhang et al. (2017). The Flavopunctelia phylogeny was constructed using all accessioned sequences of Flavopunctelia in GenBank, sequences from this study, and GenBank sequences with high BLAST similarity to our new sequences, regardless of determination. The Lecanora dispersa group phylogeny was created by adding new sequences from this study, their top-scoring megablast GenBank sequences, and the ITS of the type of L. lendemeri E. Tripp & C.A. Morse (Tripp et al. 2019) to the multiple sequence alignment from Śliwa et al. 2012 (Treebase study #12681, using ‘mafft—add’ (https://mafft.cbrc.jp/alignment/server/add.html, Katoh & Frith 2012). Similarly, the Physconia phylogeny was compiled using the 60 sequences from Esslinger et al. (2017, deposited in Dryad as https://doi.org/10.5061/dryad.bh7mc), and additional sequences from the senior author, GenBank, and this study. Finally, we aligned five new Punctelia sequences to the ITS portions of the concatenated alignment of Alors et al. (2016), and the ITS alignment of Lendemer & Hodkinson (2010) using ‘mafft—add’. ITS was concatenated with the other loci in Mesquite, and the new multiple sequence alignments were reanalyzed with partitions.

For Peltigera sequences, we also used NCBI BLAST with megablast to check the percent of our sequence that was identical to sequences published by F. Lutzoni and J. Miadlikowska Peltigera projects, which we mapped to currently undescribed molecular species delimited by Pardo-De la Hoz et al. (2018) and Magain et al. (2018). For Peltigera section Peltigera we also checked for the presence of species-specific hypervariable region sequences described in Magain et al. (2018).

A sterile crust that could not be assigned to genus on morphology or chemistry alone (Haughland 2020-28) was analyzed using the workflow in Hodkinson & Lendemer (2012). We first queried the sequence in NCBI BLAST with megablast. Based on the combination of the best BLAST hits, as well as possible species matches from the literature based on TLC results (Lendemer 2010, 2013; Malíček et al. 2017), a multi-genus dataset was created to show the placement of the sequence within the potential genera Buellia, Lecanora, Lecidella and Lepraria. Based on those results, we used the most recent Lecidella phylogeny from Zhao et al. (2015; TreeBase Study ID #17997), downloaded the seven loci dataset, added selected ITS sequences from GenBank and this study to the existing alignment (mafft--add, Katoh & Frith 2012), manually edited and trimmed the new alignment, and generated a new phylogenetic tree using methods described above, with partitions for each locus. For the multi-genus tree, we post-hoc graphically simplified and collapsed clades within Lecidella to focus on the broader, genus-level placement of our sequence.

Notepad++

Dendroscope 3.7.6  (https://uni-tuebingen.de/fakultaeten/mathematisch-naturwissenschaftliche-fakultaet/fachbereiche/informatik/lehrstuehle/algorithms-in-bioinformatics/software/dendroscope/)