In this study, we summarized all available ITS barcoding sequences of “Sanghuang” from GenBank. A total of 271 ITS sequences related to “Sanghuang” including 31 newly generated sequences for this study were analyzed. More than half of these sequences, or say 142, were mislabeled. So many errors undoubtfully raised chaos when BLAST search, especially for non-taxonomists.
Comparing with specimens, much more mislabeled sequences came from strains. Most of these sequences were submitted by non-taxonomists. One typical case is a recently published paper on genome sequencing of “Sanghuang” that meanwhile submitted six ITS sequences to GenBank (Shao et al. 2020). In GenBank, all these six sequences were labeled as Inonotus sp. rather than certain species of Sanghuangporus (MN242716–MN242721), while the six strains generating these sequences were named as Sanghuangporus sanghuang in the paper submitting these sequences (Shao et al., 2020). However, five of the six strains including that subject to genome sequencing are actually Sanghuangporus vaninii (Fig. 1, Zhou et al., 2020). That is to say, five out of six strains were wrongly identified to a species level. Therefore, this incorrected species identification makes the whole genome sequence of “Sanghuang” misapplied to an inappropriate species. Even worse, Shao et al. (2020) stated that these six strains are commercially cultivated, which further results in the name chaos for commercial products of “Sanghuang”. Another case is a paper specially on the species identity of “Sanghuang” strains (Han et al. 2016). Thirty strains deposited in the Agricultural Sciences Institute culture collection (Mushroom Research Division, Rural Development Administration, Republic of Korea) were correctly identified as Sanghuangporus vaninii and S. sanghuang according to an ITS-based phylogenetic analysis; however, unfortunately, most of these ITS sequences were mislabeled when being submitted to GenBank.
Nine mislabeled sequences came from specimens. These errors were caused mainly by the update of taxonomic recognition. Six sequences of specimens originally labeled as Sanghuangporus sp. are accepted to represent S. quercicola (Table 1). In the paper submitting these six sequences, the specimens generating them were newly described as Sanghuangporus toxicodendri (Wu et al. 2019b). However, in that paper the separation of S. toxicodendri and S. quercicola was actually not supported from a phylogenetic perspective, and moreover, the morphological differences between these two species are not on the basis of stable characters (Wu et al. 2019b). In the current phylogenetic analyses, the six specimens of S. toxicodendri, three specimens of S. quercicola and additional four collections merged together in a fully supported clade (Additional file 1: Tree S1, Additional file 2: Tree S2, Fig. 2). Therefore, S. toxicodendri and S. quercicola are considered to be conspecific, and S. quercicola has priority over S. toxicodendri. Another mislabeled sequence was generated from a specimen originally described as Inonotus tenuicontextus L.W. Zhou & W.M. Qin (Zhou and Qin 2012). Although this species was online published earlier than Inonotus weigelae T. Hatt. & Sheng H. Wu, the basionym of Sanghuangporus weigelae (Wu et al. 2012a), its online date is before January 1st, 2012 and thus not effective. Soon, I. tenuicontextus was treated as a later synonym of I. weigelae (Tian et al. 2013). Therefore, this mislabeled sequence is accepted to represent S. weigelae (Table 1).
The independence of Sanghuangporus lonicericola was not well supported in the current phylogenetic analyses (Additional file 1: Tree S1, Additional file 2: Tree S2, Fig. 2). Similarly, Sanghuangporus alpinus and S. sanghuang were not strongly supported as monophyletic species by the ML algorithm (Fig. 2). However, the intraspecific difference of ITS sequences in each of the three species was quite low (0.10–0.49%, Table 2). So, we still accept S. alpinus, S. lonicericola and S. sanghuang as three independent species. Maybe a phylogenetic analysis employing more loci will improve the resolution. On the contrary, Sanghuangporus baumii, S. weirianus and S. zonatus are the only three species with more than 1.00% of intraspecific ITS differences (Table 2). However, these three species all received strong supports as independent lineages (Additional file 1: Tree S1, Additional file 2: Tree S2, Fig. 2). Noteworthily, Chinese collections of Sanghuangporus baumii formed three strongly supported subclades corresponding to geographic origins, viz. nine from Northeast China, two from Beijing and two from Shanxi; regarding S. zonatus, two collections of from Hainan, China grouped together with full statistical support, and then formed a fully supported clade with the collection from Yunnan, China (Table 1, Fig. 2). Moreover, branch lengths of the only two available collections of S. weirianus were extremely different (Fig. 2). A more comprehensive sampling of these three species in phylogenetic analyses will further clarify their intraspecific relationships. For now, we tentatively accept them as monophyletic species.
Although intact mature specimens of “Sanghuang” are not difficult to be morphologically identified to a species level in a short time, most of commercial products are chips and pieces or even powders. Normally, it is impossible to rapidly determine which species such kind of commercial products really represents. Like other medicinal mushrooms (Raja et al. 2017), species names of Sanghuangporus are sometimes misapplied to certain products of “Sanghuang” (Shao et al. 2020). This confused situation to some extent restricts the industrial development of “Sanghuang” (Zhou 2020). Therefore, to standardize the industry of “Sanghuang”, ten candidate sequences were provided for HRCA based on the accurate boundaries among three commonly studied and cultivated species, viz. Sanghuangporus baumii, S. sanghuang and S. vaninii (Lin et al. 2017; Zhou et al. 2020). HRCA is an isothermal amplification approach and thus provides a rapid, simple and low-cost detection of specific nucleic acid sequences (Nilsson et al. 1994; Lizardi et al. 1998). This approach has been widely used for clinic detection of human-pathogenic microfungi (Zhou et al. 2008; Trilles et al. 2014; Rodrigues et al. 2015), and recently, was also reported for rapid detection of poisonous macrofungi (He et al. 2019a, 2019b). Regarding lethal Amanita species, a more than two-nucleotide-long difference was evidenced to be valid for identification of α-amanitin gene (He et al. 2019a). Here, to provide more candidates, two and more nucleotide differences are given, because it was reported that this approach could reveal single nucleotide differences (Nilsson et al. 1997). Hopefully, certain candidates will work well in future experiments.