Discovery of a cryptic species, Erysiphe salicina sp. nov., and reconstruction of the phylogeny of powdery mildews on Populus and Salix spp.

Recently performed phylogenetic-taxonomic analyses of species belonging to Erysiphe sect. Uncinula on willows (Salix spp.) demonstrated a much higher diversity than previously assumed. Phylogenetic analyses and morphological examinations of Chinese Erysiphe collections on Salix abscondita (= S. raddeana), S. sinica and S. taraikensis, all belonging to Salix subgen. Vetrix sect. Vetrix, revealed an additional cryptic species which is described as Erysiphe salicina. The new species is morphologically and phylogenetically distinguished from all other powdery mildew species. The phylogeny of the new species and closely related species on Populus and Salix spp. (Salicaceae) was reconstructed and discussed using a combined alignment of the internal transcribed spacer (ITS) regions and partial 28S rDNA sequences. The present phylogenetic analyses suggest that the recently described E. salicicola (on Salix gracilistyla in Republic of Korea) has to be reduced to synonymy with E. salicis-gracilistylae comb. nov. (≡ Uncinula salicis-gracilistylae).


Introduction
The genus Salix L. (Salicaceae), commonly known as willows, consists of~450 species, mainly distributed throughout the Northern Hemisphere (Skvortsov 1999;Fang et al. 1999;Gulyaev et al. 2022). Salix comprises valuable ecological species that are commonly used in conservation and environmental projects for habitat restoration and erosion control (Kuzovkina and Quigley 2005;Kuzovkina and Volk 2009). Subgeneric classification of Salix has remained in a process of endless revision (Gulyaev et al. 2022). Chen et al. (2010) proposed a infrageneric classification of Salix with four subgenera: Salix, Chosenia, Triandrae, and Vetrix. Species of the genus Salix are diverse in China (275 spp., Fang et al. 1999). Salix sinica (K. S. Hao ex C. F. Fang and Skvortsov) G. Zhu is a species originally described as S. caprea var. sinica K. S. Hao (Zhu 1998). It is considered a species of its own, according to the current taxonomic Section Editor: Roland Kirschner * Shu-Yan Liu liussyan@163.com concept recognized in Flora of China and belongs to Salix sect. Vetrix (http://www.iplant.cn/info/Salix%20sinica?t=z). Salix spp. are commonly infected by powdery mildews almost worldwide. Four genera, six species, and two varieties of powdery mildews have been recorded parasitizing Salix spp., viz., Podosphaera schlechtendalii Lév., Phyllactinia populi (Jacz.) Y.N. Yu, Pleochaeta salicicola R.Y. Zheng and G.Q. Chen, Erysiphe adunca (Wallr.) Link var. adunca and var. salicis-gracilistylae (Homma) U. Braun and S. Takam., E. capreae DC. ex Duby, and E. pseudoregularis U. Braun (Braun and Cook 2012). The three species of Erysiphe sect. Uncinula on willows are mainly distributed in Asia (China, India, Japan, Korea, Russia [Siberia and Far East]), Europe, and North America (Braun and Cook 2012). Recently, a comprehensive morphological and phylogenetic revision of the E. adunca complex has been conducted by Darsaraei et al. (2021). They reinstated E. salicis DC. (on Salix spp. belonging to various subgenera of Salix in Asia and Europe) as a recognized species of its own, introduced the new combination Erysiphe salicis var. salicis-gracilistylae and reduced Erysiphe pseudoregularis to synonymy with E. capreae. The morphological characteristics of a powdery mildew found in China on Salix sinica, S. taraikensis, and S. abscondita turned out to be different from all recognized Erysiphe spp. on Populus and Salix spp. in terms of morphology and phylogenetic position. Furthermore, the phylogenetic position and taxonomic status of Erysiphe salicicola, recently described from the Republic of Korea on Salix gracilistyla var. melanostachys (Boonmee et al. 2021), have been reassessed. Sequences retrieved from the new Chinese species and from E. salicicola have been added to the comprehensive phylogenetic analyses recently published by Darsaraei et al. (2021) in order to update the phylogeny of Erysiphe (sect. Uncinula) spp. on Salicaceae.

Specimens examined
Twenty-eight specimens of Erysiphe species on Salix and Populus were collected in China and Russia. They are now deposited in the Herbarium of Mycology of Jilin Agricultural University (HMJAU). Details of these specimens are provided in Table 1.

Morphological examinations
Anamorphs and teleomorphs of dried herbarium specimens were examined in lactic acid using a light microscope (Axio Scope A1, ZEISS, Germany). The sizes of diverse structures of the anamorphs and teleomorphs were measured with at least 30 repetitions. If possible, measurements of microscopic structures were made at a magnification of ×1000, and 95% confidence intervals were determined.

DNA extraction, PCR, and sequencing
Whole-cell DNA was extracted by the Chelex-100 method (Hirata and Takamatsu 1996;Walsh et al. 1991) from mycelia and chasmothecia. The complete internal transcribed spacer (ITS) region (including 5.8S rDNA) and 5′-end of the 28S rDNA (including D1 and D2 domains) of the fungi were amplified by polymerase chain reaction (PCR) using the primer sets ITS5/ITS4, PM10/PM2, PM5/PM6 for the ITS region and LSU1/LSU2, PM3/TW14 and PM28F/PM28R for the partial of 28S rDNA region. The details of primers are presented in Table 2. The reaction components were 2 μL of the total genomic DNA, 12.5 μL premixed PCR mixtures (TaKaRa, Tokyo, Japan), 1 μL of each primer (10 μM) and sterile ddH 2 O up to a final volume of 25 μL. The PCR reactions were conducted under the following thermal cycling conditions: an initial predenaturation step of 5 min at 95°C, 35 cycles of 30 s at 95°C, followed by 1 min at 56°C for annealing, and 30 s at 72°C for extension, and a final extension step of 8 min at 72°C. The PCR products were subjected to electrophoresis in a 1.2% agarose gel in 0.5 × TBE buffer. The amplicons were sent to Sangon Biotech (Shanghai, China) for direct sequencing in the forward and reverse direction using the same primers as for the PCR. The accession numbers are shown in Table 1. Details of the sequences retrieved from GenBank are shown in Supplementary Table 1.

Molecular phylogenetic analysis
The obtained sequences, including the complete ITS and partial 28S rDNA from twenty-eight herbarium specimens, were aligned with other sequences retrieved from GenBank. Multiple sequence alignments were conducted by MUSCLE implemented in MEGA-X (Kumar et al. 2018). Alignments were further manually refined and deposited in TreeBASE (http://www.treebase.org/) under the accession number S28710. Phylogenetic trees were obtained from the data using the maximum parsimony (MP) and maximum likelihood (ML) methods. MP analysis was performed in PAUP 4.0 (Swofford 2002) with the heuristic search option using the tree bisection reconnection (TBR) algorithm with 100 random sequence additions to find the global optimum tree. All sites were treated as unordered and unweighted, with gaps treated as missing data. The strength of internal branches of the resulting trees was tested with bootstrap (BS) analysis using 1000 replications with the step-wise addition option set as simple (Felsenstein 1985). Tree scores, including tree length, consistency index (CI), retention index (RI), and rescaled consistency index (RC), were also calculated. The ML analysis was performed using raxmlGUI 2.0 (Edler et al. 2020), under a GTRGAMMA
Diagnosis: Differs from all known Erysiphe species on hosts belonging to Salix in having hyphal appressoria in opposite pairs, short conidiophores only up to 87 μm, small conidia at most 30 μm in length and 14 μm in width, chasmothecia with few and short appendages and large ascospores exceeding a length of 30 μm, and by forming a monophyletic clade separate to all other Erysiphe species in phylogenetic analyses.
Etymology: Composed of the name of the host genus, Salix, + Latin adjectival suffix -ina = belonging to.

Phylogenetic analysis
Twenty-eight nucleotide sequences spanning the ITS1-5.8S rDNA-ITS2 and 28S rDNA D1/D2 domains were generated in this study. These sequences were aligned with 42 sequences of E. adunca s. lat. taken from Darsaraei et al. (2021) and 2 sequences of E. salicicola (Boonmee et al. 2021). A sequence Fig. 2 Phylogenetic analysis of combined data of the ITS and partial 28S rDNA region for 73 sequences from Erysiphe salicina, E. adunca s. lat. and an outgroup sequence from E. australiana. The tree was constructed using a heuristic search. Horizontal branch lengths are proportional to the number of substitutions that were inferred to have occurred along a particular branch of the tree. BS (≥ 70%) values by the maximum parsimony (MP) and maximum likelihood (ML) are shown on the branches. The GenBank accession numbers, scientific names of hosts and country codes are included in the tree. Sequences in bold are from type specimens. Sequences in red are from China in this study and two sequences in blue were from E. salicicola in Republic of Korea of E. australiana (AB022408) was used as an outgroup taxon in accordance with Takamatsu et al. (2015). The alignment consisted of 73 sequences and 1269 characters, of which 133 characters were variable and 141 were informative for parsimony analysis. The maximum parsimony tree (TL = 397, CI = 0.8615, RI = 0.9726, RC = 0.8378) with the highest likelihood value is shown in Fig. 2. The ML tree topology was almost identical to the MP tree and only bootstrap supports are shown in the MP tree.The 72 ingroup sequences are divided into five groups. Three groups (Groups I, II, and III) consist of sequences from powdery mildews on Salix spp. and the other two groups (Groups IV and V) are composed of sequences from powdery mildews on Populus spp. All five groups are supported by BS analyses with 70% or higher. Group I contains all Erysiphe salicina sequences obtained from powdery mildews on Salix spp. and forms an independent clade supported by a bootstrap value of 100%. Group II contains two sister clades, viz., E. salicis-gracilistylae (≡ E. salicis var. salicis-gracilistylae), including two sequences obtained from E. salicicola, and E. salicis supported by a bootstrap value of 100%.

Discussion
Salix is a common host genus for powdery mildews of various genera. Four genera, seven species and two varieties of powdery mildews are currently recognized on Salix spp. (Braun and Cook 2012;Yu and Lai 1979;Zheng and Chen 1978;Boonmee et al. 2021;Darsaraei et al. 2021). Powdery mildews in the genus Erysiphe are the predominant species infecting Salix hosts and are geographically widespread. Molecular phylogenetic analyses play an important role in the fungal taxonomy of the powdery mildews. Genetic evaluations are the main method to discover new species, above all cryptic species within complexes (Kirschner et al. 2020;Qiu et al. 2020;Yamaguchi et al. 2021). The recently published phylogenetic-taxonomic revision of the Erysiphe adunca complex (Darsaraei et al. 2021) is a case in point, and provides the basis for the present examinations and analyses. The comparative morphology and phylogenetic analysis from the current study elucidate several issues with this group of fungi including the recent publication of the species E. salicicola. The morphology of E. salicicola is consistent with E. salicisgracilistylae (≡ Uncinula salicis-gracilistylae), and sequences retrieved from E. salicicola cluster within the E. salicis-gracilistylae clade, suggesting that E. salicicola has to be reduced to synonymy with E. salicis-gracilistylae. This clade and corresponding taxon have been discussed in detail in Darsaraei et al. (2021). Owing to uncertainties in the interpretation of the cluster concerned, either as species clade or as subclade of the E. salicis clade led to the maintenance of this taxon as variety, as in previous treatments (Braun and Cook 2012). However, sequences retrieved from the new species recently described from Republic of Korea and the present new phylogenetic analysis confirm a robust, strongly supported cluster that can be treated as species clade. The host range within this clade is also different from that of E. salicis. However, the treatment of this taxon as a species of its own requires the introduction of the new combination Erysiphe salicis-gracilistylae. Powdery mildews found on multiple Salix spp. in China have been morphologically and phylogenetically examined. The species is genetically and morphologically distinct from all allied Erysiphe species on Populus and Salix spp. and is thus described as E. salicina sp. nov. This species is morphologically different from E. capreae and E. salicis in having shorter conidiophore foot-cells [17-59 × 4-7 μm vs (25-)35-90(-110) × 5-9 μm in E. capreae and 15-45 × 5-9 μm in E. salicis] and smaller conidia (20-29 × 8-14 μm vs 25-45 × 10-24 μm in E. capreae and 23-40 × 10-18 μm in E. salicis). The ascospores are larger (24-34 × 11-17 μm) compared to E. salicis (16-29 × 9-16 μm) and E. salicisgracilistylae (19-24 × 10-18 μm). In addition, Erysiphe salicina has chasmothecia with fewer appendages per chasmothecium [22-53 vs (33-)50-400 in E. capreae, 25-95 in E. salicis and E. salicis-gracilistylae], and appendages with enlarged uncinate-circinate apices (vs tightly uncinatecircinate, not enlarged apices in E. salicis and E. salicisgracilistylae). Furthermore, the appendages are regular, i.e., without constrictions and swellings, compared to E. salicisgracilistylae, and they arise equatorially, as in E. salicis and E. salicis-gracilistylae, compared to E. capreae that are not equatorial or only somewhat in the upper half, and have a tendency to point upwards at maturity. Erysiphe adunca s. str. on Populus spp. (Salicaceae) is genetically as well as morphologically also distinguished from E. salicina by having much longer conidiophores, 70-150 μm, and foot-cells, 40-110 μm, larger conidia, 28-38 × 14-18 μm, and much longer chasmothecial appendages, 1-3 times as long as the chasmothecial diam. (Darsaraei et al. 2021).
At first glance, the result was unexpected, because the most common host plant of the new species, S. sinica, was originally described as S. caprea var. sinica, i.e., this taxon is closely related to S. caprea and both species pertain to Salix subgen. Vetrix sect. Vetrix. Salix caprea is the principal host of E. capreae, which is common in the northern hemisphere. Before our analysis we expected E. capreae also on S. sinica. The occurrence of a separate species on two hosts belonging to Salix subgen. Vetrix sect. Vetrix, which is morphologically and phylogenetically close to E. salicis, suggests a higher and not yet fully elucidated diversity of Erysiphe on Salix spp. in Asia, and in particular China. We hypothesize that future research will reveal additional cryptic species, hidden under previous reports as Erysiphe (Uncinula) adunca.
There are additional open questions that will require analyses. For example, Zhu (1998) supposed that S. sinica collections from northern and northeastern China have often been misidentified as S. caprea. Erysiphe spp. on these hosts require further examinations in northeastern China. Erysiphe salicis has not yet been verified for China and future collection efforts are needed to elucidate whether E. salicis is distributed in China. Additionally, Erysiphe adunca on Populus spp. in China needs to be evaluated. It is currently known that E. adunca is a common and widespread species on a wide range of Populus spp. in Asia, Europe and North America (Darsaraei et al. 2021). Several Uncinula spp., based on Chinese specimens on poplars, have been described (Zheng and Chen 1977), but all names were reduced to synonymy with E. adunca (Braun and Cook 2012;Darsaraei et al. 2021). It is still unclear whether all Chinese Erysiphe collections on Populus spp. pertain to two Erysiphe species viz. E. adunca and E. mandshurica. It cannot be excluded that, comparable to Erysiphe on willows, the diversity of Erysiphe on poplars is much higher than previously assumed, most notably in China.