Novel biodiversity and biogeography insights of filamentous Bangiales (Rhodophyta) from China
Article information
Abstract
The taxonomy of filamentous Bangiales was challenging due to its plastic and simple morphology. Molecular approaches play a significant role in revealing cryptic biodiversity and resolving phylogenetic relationships. Nonetheless, the biodiversity of filamentous Bangiales in many regions, including China, has not been thoroughly investigated yet using updated methods. In this study, we investigated the biodiversity of filamentous Bangiales in China, employing a more comprehensive and detailed sampling approach. The rbcL and 18S gene sequences (partial or complete) of 49 specimens collected from 37 sites were obtained. Phylogenetic analysis resolved these specimens into four clades, including Bangia, ‘Bangia’ 1, ‘Bangia’ 2, and Pseudoganglia. Seven putative species were delineated using empirical methods, including Automatic Barcode Gap Discovery (ABGD), General Mixed Yule Coalescent (GMYC), and Bayesian Poisson tree processes (bPTP), as well as genealogical analysis. Among them, four were species previously reported, two were cryptic species, and the other was a new record in China. The distribution of these molecularly delineated species was mapped along China’s coastline. Similar to the bladed Bangiales, antitropical, transoceanic, and thermal-gradient distribution patterns were also observed for filamentous Bangiales on a global scale. The ancestral region was retraced by incorporating species data from China into existing datasets using BioGeoBEARS with BAYAREALIKE + J as the best-supported model. Under this model, East Asia shows the highest marginal probability as the ancestral area for the crown clade, though uncertainty remains across alternative regions. A most parsimonious expansion scenario of Bangiales from east Gondwana to other regions of the world was proposed.
INTRODUCTION
The “filamentous Bangiales” was a term used to refer to the species in the order Bangiales with a filamentous thallus, which were previously classified in the genus Bangia. The taxonomy of Bangia sensu lato began with the description of Conferva atropurpurea Mertens & Roth (Roth 1806) and its marine counterpart, Conferva fuscopurpurea Dillwyn (Dillwyn 1809), both of which were moved to Bangia when Lyngbye established the genus (Lyngbye 1819). Several other names, including Girardia, Bangiella, and Diadenus (reviewed by Kuntze 1891), have also been applied to this group of algae. However, Bangia was conserved in the 1910 Brussels Code, while others were rejected (Briquet 1912), making Bangia the only legitimate name. Thus, Bangia sensu lato comprises 16 species (listed in AlgaeBase based on morphology, Guiry and Guiry 2025), characterized by red/purple, simple, thin, straight, unbranched, and segmented filaments (Li et al. 2025).
It was well established that the taxonomy of filamentous Bangiales (Bangia sensu lato) was extremely challenging due to its simple and variable morphology. Thanks to molecular approaches, such as DNA barcoding, filamentous Bangiales has been verified as paraphyletic (Oliveira et al. 1995). New genera, including Dione W. A. Nelson, Minerva W. A. Nelson and Pseudobangia Müller & Sheath, were separated from Bangia (Müller et al. 2005, Nelson et al. 2005). A more extensive study using two-gene phylogeny (rbcL and 18S) revealed that 31 specimens of Bangia sensu lato could be classified into seven genera (Sutherland et al. 2011). The analyzed specimens were mainly collected from New Zealand, the United States, Canada, and Japan. Compared with its global distribution from subpolar to subtropical coastlines, the studied area suggests that the biodiversity of filamentous Bangiales remains poorly understood. A piece of evidence for this was the proposal of a new monotypic genus, Kuwaitiella A. H. Hasan, P. Van der Aa, F. C. Küpper, D. Al-Bader & A. F. Peters, from Kuwait (Hasan et al. 2022). All the aforementioned genera are monotypic, except for Pseudobangia, which comprises two species (Müller et al. 2005, Dumilag et al. 2023). Three other putative genera are temporarily named as ‘Bangia’ 1, ‘Bangia’ 2, and ‘Bangia’ 3, composed of 24 undescribed taxonomic units (Sutherland et al. 2011).
In China, we were the first team trying to study the taxonomy of filamentous Bangiales using molecular approaches. We found that the freshwater specimens of filamentous Bangiales from China are likely the same species as those from Europe and North America, and two marine species were detected (our unpublished results). Deng et al. (2015) detected another marine species beyond our species, belonging to ‘Bangia’ 1 and ‘Bangia’ 2, respectively. Subsequent studies have yielded no novel discoveries regarding taxonomy (Chou et al. 2015, Tian et al. 2015, Xu et al. 2016). In summary, previous studies have identified four species from specimens collected at 15 sites, suggesting that the biodiversity of filamentous Bangiales in China remains largely uninvestigated, a situation similar to the global situation. In the present study, we investigated the biodiversity of filamentous Bangiales in China by collecting specimens from a greater number of sites.
MATERIALS AND METHODS
Materials
The filamentous Bangiales specimens were collected along China’s coastline during winter and spring (December to May) since 2009. We designated those that differed from each other at least in terms of collection site, collection time, or morphology as a specimen. All specimens were dried in a cool, well-ventilated environment and then brought back to the laboratory. Randomly chosen specimens were rehydrated with marine or freshwater, cleaned of epiphytes using a soft brush and given a unique identifier. Specimens from each site were stored in the herbarium at the Jiangsu Marine Fisheries Research Institute (JSMI). The detailed specimen information was provided in Supplementary Table S1.
Sequence data obtaining
Approximately 20 mg of the thallus was used to extract the genomic DNA, and the extraction methods were conducted according to Yang et al. (2013). The primer pair, rbcL-Rh1 and rbcS, used to amplify 1,374 bp of the large subunit of ribulose biphosphate carboxylase/oxygenase (rbcL), was synthesized by Sagon (Shanghai, China) according to Hanyuda et al. (2004). The primer pair, G01.1 and G15.1, from Müller et al. (1998) was used to amplify the complete gene of the small subunit of ribosomal rRNA (18S rRNA). For the specimens that failed to produce the complete 18S gene, the primer pair G06 (Saunders and Kraft 1996) and G15.1 was used to amplify the V9 region. The polymerase chain reaction (PCR) amplification system and procedure were conducted as described by Yang et al. (2018). The PCR products were bidirectionally sequenced by Azenta (Suzhou, China) after being verified by electrophoresis in a 1.0% agarose gel. Results from the bidirectional sequencing were combined using Vector NTI v.11 and manually corrected.
Species delineation
The obtained sequences were aligned using MEGA v.11 (Tamura et al. 2021) and ClustalW (Thompson et al. 1994). All sequences with one or more base pair variations were designated as unique and used as queries to blast the National Center for Biotechnology Information (NCBI) non-redundant database using the Blastn algorithm (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The top-hit sequences were used in the subsequent species delineation analysis. Three empirical methods, the Automatic Barcode Gap Discovery (ABGD) (Puillandre et al. 2012), the General Mixed Yule Coalescent (GMYC) (Pons et al. 2006), and the Bayesian Poisson tree processes (bPTP) method, were employed for species delineation. The ABGD analysis was performed online at https://spartexplorer.mnhn.fr/delimitation. The JC69 distance and an X of 1.0 were selected, with other parameters set to their default values (Pmin = 0.001, Pmax = 0.100, Steps = 10, Number of bins = 20). The GMYC method delineates species by identifying the transition between inter- and intraspecific evolutionary processes through the analysis of differences in branching patterns in a gene tree. The Bayesian tree used for the GMYC analysis was reconstructed using BEAST v. 10.5.0 (Drummond and Rambaut 2007) with the same parameters as those used by Yang et al. (2018). The tree convergence was diagnosed using Tracer v. 1.6 (Rambaut et al. 2014). The GMYC analysis was performed on maximum credibility trees calculated with Treeannotator v. 1.8.3 (Drummond and Rambaut 2007) using the SPLITS package for R (http://r-forge.r-project.org/projects/splits/). The single threshold was added to the bifurcation ultrametric tree. The bPTP analysis was also performed online (https://species.h-its.org/ptp/) on a consensus tree calculated with MrBayes in Phylosuite. Sensitivity analysis of bPTP using different parameters was conducted.
Phylogenetic analysis
To ascertain which genera the Chinese specimens belong to, we reconstructed the phylogenetic trees based on concatenated sequences of rbcL and 18S. Representative species from different genera of Bangiales are selected and used for the phylogenetic analysis. Their sequences were obtained from the NCBI database and incorporated with the data obtained in the present study. Chlidophyllon kaspar (W. A. Nelson & N. M. Adams) W. A. Nelson and Smithora naiadum (C. L. Anderson) Hollenberg were used as outgroups. The species list and accession numbers of the gene sequences for these selected species are provided in Supplementary Table S2. The rbcL and 18S sequences were multi-aligned using MEGA v.11. The unpaired end sequences were removed before being concatenated using Phylosuite v.1.2.3 (Zhang et al. 2020). The best-fit evolutionary model was assessed using ModelFinder (Kalyaanamoorthy et al. 2017), incorporated into Phylosuite, based on the Akaike information criterion (AIC). The GTR + I model was identified as the best fit. Bayesian inference was performed using MrBayes v.3.2.2 (Ronquist et al. 2011) with the GTR + I model. Two independent analyses were conducted with four chains each, for a total of two million generations (five million generations for Supplementary Fig. S1) with a sampling frequency of every 1,000 generations. The first 25% sampled trees were discarded as “burn-in” to ensure stabilization. The maximum likelihood analysis was performed using IQ-TREE (Nguyen et al. 2015) with the GTR + I model and tested with 1,000 replicates of the bootstrap value. The flow chart and configuration parameters were provided in Supplementary Fig. S2. The online tool iTOL (https://itol.embl.de/) was employed to visualize the phylogenetic trees.
Distribution analysis
To understand the distribution range of filamentous Bangiales in China, the distribution of each putative species was mapped along China’s coastline based on the collection site information (Supplementary Table S1) and molecular data obtained in the present study.
To ascertain whether filamentous Bangiales has a similar distribution pattern to Pyropia sensu lato found by Yang et al. (2018), the species distribution information of filamentous Bangiales species verified by the molecular evidence was retrieved from GenBank (Supplementary Table S3). The phylogenetic distribution of filamentous Bangiales was mapped globally based on the aforementioned species distribution information.
Ancestor region reconstruction
The rbcL and 18S sequences of 205 species (taxonomic units) were retrieved from GenBank (Supplementary Table S4) and incorporated into the dataset obtained in the present study. Chlidophyllon kaspar and Smithora naiadum were used as outgroups. A Bayesian tree was also reconstructed using BEAST with the same methods of species delineation, and a consensus tree was retrieved using Treeannotator. Twelve biogeographic regions were defined according to the distribution information obtained above: (A) East Asia, (B) Australia and New Zealand, (C) India, (D) Middle East, (E) Europe, (F) South Africa, (G) Greenland, (H) Eastern North America, (I) Eastern South America, (J) Southern South America and Antarctica, (K) Western South America, and (L) Western North America. To estimate the ancestral distribution pattern, diffusion-isolation analysis (S-DIVA) was performed using RASP v4 (Yu et al. 2020), with a maximum of two ancestral regions per node (Xu et al. 2018). The model comparison was performed using BioGeoBEARS with AIC values (Supplementary Table S5). BAYAREALIKE + J was chosen to be the best model, and the ancestral distribution pattern was also traced based on this model.
RESULTS
We collected 49 specimens from 37 sites along China’s coastline since 2009, with approximately half of the specimens collected in the last five years. Forty-eight rbcL (1,374 bp) and seventeen 18S (1,980–2,103 bp) sequences were obtained in the present study, among which 17 rbcL and 12 18S sequences were unique. The V9 region of the 18S gene was obtained from 19 specimens that failed to produce the complete 18S gene. Subsequent analyses were based on these data and those retrieved from GenBank.
Phylogenetic relationships
Phylogenetic analysis based on rbcL, 18S, or concatenated datasets demonstrated that specimens from China belong to four of the eight clades (genera) of filamentous Bangiales. To be specific, three specimens from China were resolved in the clades of Bangia, three in ‘Bangia’ 1, 40 in ‘Bangia’ 2, and three in Pseudobangia (Figs 1 & 2, Supplementary Table S1).
Phylogenetic relationships inferred from the large subunit of ribulose biphosphate carboxylase/oxygenase (rbcL) gene (1,374 bp). The dataset comprised unique rbcL sequences. Bayesian posterior probabilities (BPP) and maximum likelihood (ML) bootstrap values (BPP/ML) are shown at each node (only values >60 are displayed). Dione was used as the outgroup.
Bayesian phylogenetic tree of the order Bangiales based on concatenated sequences of large subunit of ribulose biphosphate carboxylase/oxygenase (rbcL) and 18S. Values near nodes represent Bayesian posterior probabilities/maximum likelihood support (values below 50 are not shown). Chlidophyllon kaspar and Smithora naiadum were used as outgroups. Species from China analyzed in this study are marked with black dots.
Species delineation
All 17 unique rbcL sequences were used to analyze the relationship among our specimens. Forty-eight specimens were resolved into six well-supported clades in the Bayesian tree based on these unique rbcL sequences (Fig. 2). The detailed specimens represented by these unique sequences were listed in Supplementary Table S1. These clades were tested using three empirical methods, ABGD, GMYC and bPTP, based on 18S and rbcL sequences, respectively. The summary of the results of ABGD and GMYC analyses is presented in Supplementary Table S6. The species delineation results of ABGD, GMYC, and bPTP were mapped on the Bayesian tree (Fig. 3). The thalli scan and voucher number were also incorporated in Fig. 3.
Species tree inferred using a single representative large subunit of ribulose biphosphate carboxylase/oxygenase (rbcL) sequence for each putative species. The numbers on the nodes represent the Bayesian posterior probability (BPP) and maximum likelihood (ML) analyses in the order of BPP/ML. For the seven Chinese filamentous Bangiales species, results are presented based on seven different species delimitation methods: the General Mixed Yule Coalescent (GMYC) method using the rbcL gene, the Automatic Barcode Gap Discovery (ABGD) method using the rbcL gene, the GMYC method using the 18S gene, the ABGD method using the 18S gene, the Bayesian Poisson tree processes (bPTP) method using the rbcL gene, the bPTP method using the 18S gene, and the genealogy.
Putative species are accepted when a barcode gap exists without overlapping and at least four independent delimitation frameworks converge under stated parameters. The supported monophyly was also taken into consideration. For rbcL, the specimens were assigned to six taxa by the ABGD and GMYC methods, but to seven taxa by the bPTP method. Three specimens assigned to Pseudobangia were lumped. For 18S, the specimens were assigned to seven taxa by all methods (Fig. 3). A singleton represented the putative species Pseudobangia sp. 2. ABGD and GMYC can recognize Pseudobangia sp. 1 based on 18S and genealogical analysis. Five other putative species were supported by all the species delineation methods tested: ‘Bangia’ 1 sp. 1, ‘Bangia’ 2 sp. 1, ‘Bangia’ 2 sp. 2, ‘Bangia’ 2 sp. 3, and Bangia. A clear intra- vs. interspecies barcode gap was observed in the present study based on rbcL (Supplementary Fig. S3). The larger interspecies gaps (>0.06) indicated the gaps between species resolved in different clades in Fig. 2. In contrast, the smaller ones (0.0066–0.0131) indicated those between species determined in the same clade. The intraspecies distance varied from 0 to 0.0044. The sensitivity analysis results of bPTP showed a basically stable mean number of species of 7, excluding the outgroup in the tree, with a relatively low acceptance rate (Supplementary Table S7). A plate of light microscopy images was provided for each putative species (Supplementary Figs S4–S10) solely to document voucher specimens and does not imply morphological diagnosability or taxonomic conclusions.
To further ascertain the identity of each clade, a representative sequence of each clade was searched against the GenBank database. The hit results were screened based on the percentage of identity. 18S sequences were also downloaded and aligned with the sequence obtained in the present study to find the best hit (Table 1). The rbcL sequence of B. atropurpurea obtained in the present study matched those from the exact location in previous reports with complete sequence identity. For 18S, the sequence obtained in the present study matched the previous ones completely in the exon region, while no introns were detected in these sequences (Deng et al. 2015, Tian et al. 2015). However, further alignment revealed that the same intron exists in this population (Xu et al. 2016). The top hits of rbcL and 18S sequence representing ‘Bangia’ 1 sp. 1, ‘Bangia’ 2 sp. 1, ‘Bangia’ 2 sp. 2, and ‘Bangia’ 2 sp. 3 were unnamed species. The hits of the rbcL sequence representing Pseudobangia sp. 1 and sp. 2 were Pseudobangia corderoi Gamus & Dumilag. Based on the sequence variation, ‘Bangia’ 2 sp. 2 was a new record for China, and Pseudobangia sp. 1 and Pseudobangia sp. 2 were identified as two cryptic species.
The probable species identity inferred by blasting against NCBI or aligning with downloaded sequences
Species distribution
The distribution of each putative species was mapped along the Chinese coastline (Fig. 4) based on the molecular data and available collection information, except B. atropurpurea, which is a freshwater species.
Distribution of filamentous Bangiales at 33 out of 37 sampling sites along the China coast (see Supplementary Table S1 for details of sites not included here). Sites harboring the same species and close to each other are omitted from the map.
‘Bangia’ 2 sp. 1 and ‘Bangia’ 2 sp. 3 were two extensively distributed species, but mainly in the temperate zone. The distribution of these two species overlapped almost entirely, with ‘Bangia’ 2 sp. 1 slightly to the south. ‘Bangia’ 2 sp. 3 was distributed from Dalian to Nanao, while ‘Bangia’ 2 sp. 1 was distributed from Dongying to Keniaowei. These two species concurred in most sites in Fujian and Guangdong Provinces, but not in other provinces. The other four marine species were distributed relatively narrowly over a small area. ‘Bangia’ 2 sp. 2, Pseudobangia sp. 1, and Pseudobangia sp. 2 were southern species, with Pseudobangia species found in subtropical regions in Guangdong and Hainan Provinces. ‘Bangia’ 2 sp. 2 was mainly found in Fujian Province. Contrarily, ‘Bangia’ 1 sp. 1 was only found in northern China, specifically in Shandong Province.
Biogeographic pattern
Distribution information for 63 taxa was retrieved from references or GenBank and mapped onto the global coastline based on phylogenetic clades (Fig. 5). We excluded the freshwater genus Bangia from this analysis. Among the current accepted marine clades/genera, Dione, Minerva, and Kuwaitiella are monotypic, with the former two distributed in New Zealand and the latter found only in Kuwait to date. All the other clades with more than one species, including ‘Bangia’ 1, ‘Bangia’ 2, ‘Bangia’ 3, and Pseudobangia, were transoceanic in distribution. Apparently, we lack data from the Southern Hemisphere and the Indian Ocean. The data from the Southern Hemisphere were primarily collected from New Zealand and a few locations in Australia. It is noteworthy that species in ‘Bangia’ 1, ‘Bangia’ 2, and ‘Bangia’ 3 were distributed in each of the main oceans that have been investigated in both hemispheres, a phenomenon referred to as an “anti-tropical distribution” in previous studies. This kind of distribution in different oceans is also known as “intercontinental disjunctions.”
Distribution map of filamentous Bangiales. Solid circles represent species distributions based on both large subunit of ribulose biphosphate carboxylase/oxygenase (rbcL) and 18S gene data, or on either rbcL or 18S data alone. Black lines indicate the biogeographic boundaries proposed by Lüning (1990) for August (these boundaries may shift in January when filamentous Bangiales flourish along the coast). A, arctic; C, cold temperate; W, warm temperate; T, tropical; ANT, antarctic.
A similar distribution pattern to Pyropia sensu lato, known as the “thermal-gradient” distribution, was also observed for the Bangia sensu lato species. In other words, species in different clades tend to be distributed in different temperature zones. Although there are overlaps and exceptions, the trend is clear. For example, species from Pseudobangia are primarily distributed in subtropical regions (Fig. 5, purple dots), whereas those of ‘Bangia’ 3 are mainly found in subpolar and cold temperate regions (Fig. 5, green dots). Species in ‘Bangia’ 2 can extend their distribution from cold-temperate to warm-temperate regions (Fig. 5, blue dots). In contrast, ‘Bangia’ 1 primarily inhabits cold temperate regions (Fig. 5, red dots), with a few exceptions (Fig. 5, red square). The pattern becomes more evident when mapping the species in ‘Bangia’ 1 based on two phylogenetic clades. Two species of Bangia 1 distributed in warm temperate regions in Australia (Fig. 5, red square) were resolved into a different clade from other species in ‘Bangia’ 1.
Ancestral region
The ancestral region of Bangiales was traced using a phylogenetic tree constructed from rbcL and 18S sequences available for 205 taxonomic units. The eastern Gondwana region, encompassing New Zealand and Australia, was the oldest ancestral origin of Bangiales. The order Bangiales was divided into two main clades. The basal one comprises the clades Bangia, ‘Bangia’ 1, Neothemis, Clymene, and Porphyra, while the crown one comprises the clades Miuraea, Lysithea, Pseudobangia, Fuscifolium, ‘Bangia’ 2, Boreophyllum, ‘Bangia’ 3, Wildemania, Phycocalidia, Pyropia, Neoporphyra, and Neopyropia. Though the tree topology between S-DIVA and BioGeoBEARS analysis was slightly different, they are consistent with each other on the nodes we are concerned about. The marginal probabilities on the node of the crown clade were 0.3018 for East Asia, 0.2604 for East Asia/Australia and New Zealand, 0.1789 for Australia and New Zealand, 0.1120 for East Asia/Western North America, indicating that East Asia is the most probable ancestral region of the crown clade. Besides, Europe and western northern America were also probable ancestral regions, which were later than the previous two. It is noteworthy that the ancestors of Bangiales in Europe and East Asia independently originated from eastern Gondwana, and those of western northern America originated from East Asia subsequently.
DISCUSSION
The present study analyzed the biodiversity of filamentous Bangiales along China’s coastline through a broader and finer sampling. Previous studies analyzed specimens from restricted sites. Specifically, we conducted the first taxonomic study of filamentous Bangiales in China using molecular approaches by analyzing specimens from five sites (unpublished data). Deng et al. (2015) collected specimens from nine sites, four of which were identical to our sampling sites. Xu et al. (2016) examined specimens from eight sites, six of which were the same as or near those studied by Deng et al. (2015), taking into account the inaccuracy of the coordinates. Tian et al. (2015) added three more sites to the list. Collectively, previous studies analyzed specimens collected from fifteen sites. In the present study, we analyzed 49 historical and contemporary herbaria collected from 37 sites across China’s coastline from Liaoning to Hainan (Fig. 3, Supplementary Table S1).
The broader and finer sampling reveals a greater diversity of filamentous Bangiales in China. All the studies mentioned above have detected the freshwater species B. atropurpurea in China. Although the number of sampling sites increased from one to three, no additional freshwater species were identified since our first molecular study was conducted (Deng et al. 2015, Xu et al. 2016). In contrast, more marine species were identified by increasing the sampling sites. Two marine species, ‘Bangia’ 2 sp. 1 (FJ769174) and ‘Bangia’ 2 sp. 3 (FJ769171), were detected in our unpublished results, corresponding to the two extensively distributed species verified in the present study. Deng et al. (2015) identified three marine species that were identical to ‘Bangia’ 1 sp. 1 (KJ023702), ‘Bangia’ 2 sp. 1 (KJ023693), and ‘Bangia’ 2 sp. 3 (KJ023687) in the present study. Xu et al. (2016) revealed ‘Bangia’ 1 sp. 1 (KP279672) and ‘Bangia’ 2 sp. 3 (KP279677). Tian et al. (2015) only detected ‘Bangia’ 2 sp. 1 (KM386512) from two sites. We covered all the above-mentioned species and identified three new cryptic species, ‘Bangia’ 2 sp. 2, Pseudobangia sp. 1, and Pseudobangia sp. 2 in China in the present study. Counting the species of bladed Bangiales identified previously (Yang et al. 2018), the results demonstrated that there are at least 20 species in Bangiales along China’s coastline. A much broader and finer sampling should be able to identify more cryptic species and reveal much greater biodiversity in the future.
Six species of filamentous Bangiales have been recorded in China, based on morphology, comprising five marine species and one freshwater species (Zheng and Li 2009). Among them, four species were recorded from mainland China, including Bangia breviarticulata Tseng (Tseng 1948), Bangia fuscopurpurea (Dillwyn) Lyngbye (Tseng 1936), Bangia gloiopeltidicola Tanaka, and Bangia radicula B. F. Zheng & J. Li (Zheng and Li 2009). The relationship between these morphology-based species and the species identified in the present study requires clarification based on integrative data from morphology, distribution, and molecular biology, which will be the focus of our subsequent work.
It was established that the freshwater specimens of filamentous Bangiales from Europe, North America, and East Asia were B. atropurpurea (Müller et al. 1998, Hanyuda et al. 2004, Van Beveren et al. 2022). In the marine environment, two species, ‘Bangia’ 2 sp. 1 and ‘Bangia’ 2 sp. 3, were extensively distributed in China. Molecular data have shown that ‘Bangia’ 2 sp. 3 was also distributed in Japan (AB053489 and AB053488) (Niwa 2003) and New Zealand (AY184341) (Broom et al. 2004). Unexpectedly, ‘Bangia’ 2 sp. 1 seems to be endemic to China. Consistent with previous studies, ‘Bangia’ 1 sp. 1 can only be detected in northern China, specifically in Shandong Province (Deng et al. 2015, Xu et al. 2016), which has also been reported in the USA (Müller et al. 1998, Kucera and Saunders 2012), France, Spain, Greece (Müller et al. 2001), Mexico (Müller et al. 2003), New Zealand (Sutherland et al. 2011), and Australia (OQ706577), most of which under the name Bangia fuscopurpurea. However, the identity of this name requires clarification. The three new records found in the present study were distributed in southern China, a region under investigation. ‘Bangia’ 2 sp. 2, identified from Fujian Province in China, was also reported from New Zealand as Bangia sp. BPL (Sutherland et al. 2011). The two subtropical species in Pseudobangia were closely related to Pseudobangia corderoi in the Philippines (Dumilag et al. 2023). It is noteworthy that four of the seven species identified from China have transoceanic distributions. The results demonstrated that transoceanic distribution of filamentous Bangiales is much more common than previously believed, a phenomenon also observed in eleven species of bladed Bangiales (Yang et al. 2018). Most transoceanic distributions of bladed Bangiales were presumed to have been introduced by anthropogenic activities relatively recently (Broom et al. 2002, Neefus et al. 2008, Milstein et al. 2015). However, most available theories explaining the antitropical and transoceanic distribution suggest that it is a paleontological event that occurred around the early Cenozoic (Briggs 1987). Therefore, it is also probable that these disjunct and transoceanic distributions of extensive species are relics of an ancient, wide distribution. Population genetic research might shed light on this question.
It has long been a consensus that taxonomic regimes based on molecular data differ from those based on morphology (Matsuyama-Serisawa et al. 2004), and genera can only be distinguished by monophyletic clades using molecular sequences (Sutherland et al. 2011). Yang et al. (2018) found a coincidence between the phylogenetic clades and their distribution. In other words, species from different clades tend to be distributed in different temperate regions. Although there were exceptions and overlaps, the trend is evident on a global scale and was later referred to as the “thermal-gradient distribution pattern” (Li et al. 2025). This pattern is the first and only characteristic found to be consistent with molecular phylogeny until now. Based on this discovery, the largest genus of Bangiales, Pyropia, was divided into four genera (Yang et al. 2018, 2020, Li et al. 2025). The phylogeny reconstructed based on organellar genomes provides more solid evidence for this division (Lee et al. 2023, Li et al. 2025, Tang et al. 2025). This “thermal-gradient distribution pattern” was also observed for filamentous Bangiales, especially those clades with more than one species. The results are another piece of evidence that the “thermal-gradient distribution pattern” could be a pointer for the taxonomy of filamentous Bangiales and even for other intertidal algae. The few exceptions occurred primarily along the northeast Atlantic coast. The distribution of Neoporphyra perforata in this region was also an exception to the “thermal-gradient distribution pattern” (Yang et al. 2018). These exceptions in this region were indicators that the biota boundaries of the northeast Atlantic coast (Fig. 5), drawn by Lüning (1991), might need reexamination.
The fossil record indicates that the origin of Bangiales can be dated back to ca. 1,087 million years ago (Mya) (Butterfield 2000, Gibson et al. 2018), suggesting that this group of red algae has undergone incomparable paleogeographic events, such as continental drift, glacial and interglacial periods, and the opening and closing of straits. Tracing the ancestral region based on modern distribution may provide insights into what they went through. Based on the presence of higher diversity and the retention of basal taxa of Bangiales in New Zealand, the Southern Hemisphere (eastern Gondwana, approximately 300 Mya) was believed to be the origin center of Bangiales (Broom et al. 2004). Yang et al. (2018) postulated theoretically that the Northwest Pacific might be the origin center of Pyropia sensu lato since the early Cenozoic (approximately 46–50 Mya), based on its antitropical distribution and higher diversity in the Northwest Pacific. Meanwhile, Xu et al. (2018) inferred that Pyropia sensu lato originated in East Asia during the Cretaceous period (approximately 121.37 Mya) by calculating the divergence time and tracing the ancestral region using empirical bioinformatic methods. However, data on filamentous Bangiales from China (which is in East Asia) were not available at that time. The species of filamentous Bangiales from China were resolved into different phylogenetic clades, including Bangia, ‘Bangia’ 1, ‘Bangia’ 2, and Pseudobangia, in the present study. If we incorporated this data and recalculated the ancestral region, the results should change. Expectedly, East Asia was not only the ancestral region of Pyropia sensu lato, but was resolved to be the most probable ancestral area for the crown clade. Meanwhile, East Asia, rather than eastern or western North America as Xu et al. (2018) estimated, was the ancestral region of the crown clade. According to Xu et al.’s (2018) divergence time, the basal and crown clades diverge at around 239.78 Mya, suggesting a time when Bangiales expanded from eastern Gondwana to East Asia. The expansion of Bangiales from eastern Gondwana to Europe may have occurred around 194.51 Mya. The expansion from East Asia to Western North America probably occurred at the same age. We propose a most parsimonious scenario for these expansions (Fig. 6). East Asia was relatively close to Eastern Gondwana around 250 Mya, which may have facilitated the expansion of the ancestors to the crown clade. The expansion of ancestors for the basal clade from Eastern Gondwana to Europe, which likely occurred around 190 Mya, possibly expanded through the shoreline of Neotethys. In contrast, the expansion from East Asia to northwestern America might have taken a route via the land bridge connecting Asia and America.
The most parsimonious routes and times of the three main expansions of Bangiales ancestors. The maps of the Earth show the position of Eastern Gondwana, East Asia, Europe, and northeastern America. The red arrow indicates the most parsimonious route of expansion from Eastern Gondwana to East Asia. The yellow line indicates the most parsimonious route of expansion from Eastern Gondwana to Europe, and the green line indicates that of expansion from East Asia to Western North America. The maps of Earth were obtained from the Wikipedia website (https://en.wikipedia.org/wiki/Pangaea), contributed by Scotese et al. (2024). We used the maps under the Commons Attribution 4.0 International License. We did not make any changes except adding the expansion routes, and we do not suggest that the authors of the maps endorse or support the scenarios. Mya, million years ago.
Notes
ACKNOWLEDGEMENTS
The authors extend sincere thanks to Prof. Qi-Jun Luo, Prof. Wei-Zhou Chen, and Prof. Da-Hai Gao for the collection of some specimens. This work was supported by the Excellent Young Talents Project, sponsored by 333 high-Level Talents in Jiangsu Province [grant No. SRC20222102], the National Key R&D Program of China [grant No. 2023YFD2400102], the China Agricultural Research System [grant No. CARS-50] and the “JBGS” Project of Seed Industry Revitalization in Jiangsu Province [(2021)033].
CONFLICTS OF INTEREST
The authors declare that they have no potential conflicts of interest.
DATA AVAILABILITY
The rbcL and 18S gene sequences obtained in the present study have been deposited in GenBank, and their accession numbers were provided in Supplementary Table S1. The sequence alignment, trees, configuration files, and scripts were deposited in Zenodo (doi.org/10.5281/zenodo.18411894).
SUPPLEMENTARY MATERIALS
Supplementary Table S1. Sampling information including sample identifier, collection location, putative species, collection date (https://www.e-algae.org).
Supplementary Table S2. Species used for the phylogenetic tree (https://www.e-algae.org).
Supplementary Table S3. Species distribution and sequence information used to infer phylogeny and map biogeography (https://www.e-algae.org).
Supplementary Table S4. Specimen collection information and GenBank accession numbers of sequences used for species distribution and historical biogeography analyses (https://www.e-algae.org).
Supplementary Table S5. The model test used in ancestral region analysis using BioGeoBEARS (https://www.e-algae.org).
Supplementary Table S6. Sequence matrix information and results of the GMYC and ABGD analyses (https://www.e-algae.org).
Supplementary Table S7. The sensitivity analysis of bPTP (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Tables.xlsxSupplementary Fig. S1. Ancestral area reconstruction of Bangiales as inferred from S-DIVA and BioGeoBEARS (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S1.pdfSupplementary Fig. S2. The flow chart and parameter configuration of phylogenetic analysis (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S2.pdfSupplementary Fig. S3. The barcode-gap plot (A) and distance distribution (B) based on rbcL within and between species (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S3.pdfSupplementary Fig. S4. The microscopic morphology of Bangia atropurpurea (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S4.pdfSupplementary Fig. S5. The microscopic morphology of 'Bangia' 1 sp. 1 (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S5.pdfSupplementary Fig. S6. The microscopic morphology of 'Bangia' 2 sp. 1 (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S6.pdfSupplementary Fig. S7. The microscopic morphology of 'Bangia' 2 sp. 2 (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S7.pdfSupplementary Fig. S8. The microscopic morphology of 'Bangia' 2 sp. 3 (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S8.pdfSupplementary Fig. S9. The microscopic morphology of Pseudobangia sp. 1 (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S9.pdfSupplementary Fig. S10. The microscopic morphology of Pseudobangia sp. 2 (https://www.e-algae.org).
algae-2026-41-2-27-Supplementary-Fig-S10.pdf