Southeast Asia harbors a unique and diverse freshwater fauna of Mesozoic origin, which is under severe threat of extinction because of rapid economic development and urbanization. The largest freshwater basins of the region are certainly the primary evolutionary hotspots and they attract the most attention as key biodiversity areas for conservation. In contrast, medium-sized rivers are considered low-importance areas with secondary biodiversity, whose faunas originated via founder events from larger basins during the Pleistocene, although such a scenario has never been tested by using a phylogenetic approach. In this investigation, we used freshwater mussels (Unionidae) as a model to estimate the levels of endemism within the Sittaung, a little-known remote basin in Myanmar, compared with the surrounding larger rivers (Irrawaddy, Salween and Mekong). We discovered that the Sittaung represents an exceptional evolutionary hotspot with numerous endemic taxa of freshwater mussels. On the basis of our extensive dataset, we describe two new tribes, two genera, seven species and a subspecies of Unionidae. Our results highlight that medium-sized basins may represent separate evolutionary hotspots that harbor a number of endemic lineages. These basins should therefore be a focus of special conservation efforts alongside the largest Southeast Asian rivers.
In the modern period of the sixth mass extinction (Anthropocene), freshwater biodiversity is under severe threat because of increasing anthropogenic pressure, which leads to habitat degradation, water pollution, keystone species declines, and the homogenization of faunas1,2,3,4,5,6. Climate change may increase the effects of human impacts, especially for taxa with low abundance and restricted ranges, and may trigger multiple local extinctions7, 8. Our understanding of spatial biodiversity patterns across freshwater basins is very limited because the systematics of many groups are not developed, including the important invertebrates such as bivalves and gastropods9,10,11,12,13. In Southeast Asia, the lack of reliable taxonomic information precludes producing the national maps of freshwater biodiversity hotspots, which is a task of great importance for conservation planning9.
Here, we use mussels in the family Unionidae, or naiads, as a model group for the assessment of spatial patterns of freshwater biodiversity across western Indo-China. This is the most species-rich bivalve family, with ~620–680 extant species14,15,16,17. The Unionidae most likely originated in Southeast and East Asia in the Jurassic, with subsequent expansions into other landmasses9. In several major Asian river systems (e.g., Mekong and Yangtze), exceptional intra-basin radiations of the Unionidae were discovered, which suggests that these basins may be considered ancient (long-lived) rivers that have existed throughout the Cenozoic9, 18. However, the freshwater mussel faunas of Asia have attracted little attention from scientists compared with those from Europe and North America10, 17. Although the importance of freshwater mussels in tropical ecosystems is still poorly known, they could play an essential role as biofilters in polluted water bodies19. Several species are successful invaders, and have spread beyond their native ranges together with the introduction of their host fishes and may threaten native communities20, 21. Finally, freshwater mussels are important objects for the ornamental pet trade, pearl cultivation and food markets across Asian countries22,23,24.
The taxonomy of freshwater mussels of Southeast Asia is complicated, because many nominal taxa were described from this region on the basis of conchological features, including small differences in shell shape9, 10, 21, 25. In western Indo-China, most historical samples were collected from the Irrawaddy, Salween, Pegu, Tavoy and Great Tenasserim river catchments10, 24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48. In contrast, nominal taxa from the Sittaung, a medium-sized river basin in Myanmar, were not described. In the 1880s, Leonardo Fea, an Italian naturalist and traveler, collected mussels from several tributaries of the Sittaung49. Fea’s collection was deposited in the Museo Civico di Storia Naturale di Genova (MSNG, Genoa, Italy). Since then, none of the freshwater mussels have been collected from the basin. Moreover, for some enigmatic reason Fea’s samples from the Sittaung were not used for the description of any unionid taxon39, 40, 43, 50.
From the comprehensive phylogenetic study of the Unionidae across the primary basins of the Oriental Region, Bolotov et al.9 showed that each large river system of this region is a separate evolutionary hotspot harboring a unique endemic naiad fauna. In the present investigation, we expand this sample to estimate the levels of endemism within the Sittaung compared with the surrounding larger rivers (Irrawaddy, Salween, and Mekong). From the phylogenetic and morphological studies of the mussel samples, we describe seven new species and a single subspecies, which are ancient endemic lineages of the Sittaung. Additionally, we found two new tribes and two new genera, which are described herein. Our findings highlight that the medium-sized basins may represent separate evolutionary hotspots that should be considered in future conservation planning for freshwater biodiversity in Southeast Asia.
Phylogenetic and species delimitation analyses
Our multi-locus phylogeny (COI + 16S rRNA + 28S rDNA) contains 469 specimens of Unionidae, including 403 specimens from the Oriental Region (Supplementary Tables 1 and 2, Supplementary Figs 1 and 2). The Oriental sample includes 256 unique haplotypes belonging to 80 putative species (Fig. 1). The Bayesian Poisson tree processes (bPTP) model supports the majority of these possible species-level units (Fig. 1). The results inferred from the single-rate PTP (sPTP) model were generally similar to that of bPTP model, with a few exceptions (Fig. 1). The molecular operational taxonomic units (MOTUs) of the multi-rate PTP (mPTP) model were also comparable to those of the two other models, but in some cases, the mPTP revealed larger clusters, which occasionally joined all of the species within a certain genus (e.g., Indonaia and Radiatula) into a single MOTU (Fig. 1).
Searching with MrBayes, BEAST and RAxML returned a similar well-resolved topology (Fig. 2 and Supplementary Figs 1 and 2). These phylogenetic models support three subfamilies, Pseudodontinae, Rectidentinae and Parreysiinae, whose representatives are widely distributed across the Oriental Region. In addition to previously described tribes, we found two new distant clades: Leoparreysiini Vikhrev, Bolotov et Kondakov tribe nov. and Pilsbryoconchini Bolotov, Vikhrev et Tumpeesuwan tribe nov. Our results support the majority of previously designated genera of the Indo-Chinese Unionidae, but the new phylogeny reveals that three genera should be restored and that two genera are new to science (Fig. 2).
Eight species and a single subspecies from the Sittaung were considered new to science, because they represent distinct endemic lineages and reveal significant morphological and molecular differences from previously known taxa (Fig. 1, Tables 1 and 2). The majority of these taxa are sisters to species from the Irrawaddy, indicating multiple ancient connections between these basins since the Oligocene (Fig. 1 and Supplementary Fig. 2). In this study, we describe seven species and a subspecies (Figs 3 and 4) because our sample of Oxynaia sp. “Taungoo” is too small and needs to be expanded in the future. Additionally, Lamellidens savadiensisstat. res. inhabits both the Sittaung and Irrawaddy drainages (Table 3). However, the presence of a single haplotype in the Sittaung instead of multiple haplotypes in the Irrawaddy (Fig. 1) may indicate a recent natural or human-mediated invasion of this species into the Sittaung. The habitats of each taxon that we collected are listed in Supplementary Table 3. All of the type series are deposited in RMBH, Russian Museum of Biodiversity Hotspots, the Federal Center for Integrated Arctic Research, Russian Academy of Sciences (Arkhangelsk, Russia).
Family Unionidae Rafinesque, 1820
Subfamily Parreysiinae Henderson, 1935
Type Genus: Parreysia Conrad, 1853 (by original designation)
Comments: This subfamily includes at least five valid tribes: Parreysiini Henderson, 1935, Coelaturini Modell, 1942, Lamellidentini Modell, 1942, Oxynaiini Starobogatov, 197017 and Leoparreysiini Vikhrev, Bolotov et Kondakov, a new tribe described here.
Tribe Leoparreysiini Vikhrev, Bolotov et Kondakov tribe nov.
Type Genus: Leoparreysia Vikhrev, Bolotov et Aksenova gen. nov.
Comments: This tribe includes only the genus Leoparreysia Vikhrev, Bolotov et Aksenova gen. nov.
Diagnosis: Shell very thick, shape of round, umbo rather prominent and situated near the anterior end. Pseudocardinal teeth well developed; laterals convex and thick. Anterior adductor scar round and very deep.
Distribution: Western Indo-China.
Genus Leoparreysia Vikhrev, Bolotov et Aksenova gen. nov.
Type Species: Leoparreysia canefrii Vikhrev, Bolotov et Kondakov sp. nov.
Comments: We assigned eleven species to the genus, including a new species from the Sittaung River basin (Table 3).
Etymology: This genus is named after Leonardo Fea (1852–1903), an adventurous Italian naturalist, who sampled freshwater mussels from unexplored areas of British Burma (Myanmar).
Diagnosis: The same as for the tribe.
Distribution: Western Indo-China.
Leoparreysia canefrii Vikhrev, Bolotov et Kondakov sp. nov.
Type material: Holotype RMBH biv254_4: Myanmar: Sittaung near Taungoo, 26.xi.2016, Vikhrev leg. Paratypes: the type locality, 6 specimens (RMBH biv254_6, biv254_2, biv249, biv252_1, biv252_3, and biv252_2), 25–26.xi.2016, Vikhrev leg.
Etymology: This species is named after Cesare Maria Tapparone-Canefri (1838–1891), an Italian malacologist, who described numerous mussel taxa from British Burma (Myanmar).
Diagnosis: The new species is similar to L. olivaceacomb. nov. and L. tavoyensiscomb. nov. but differs in a more inequilateral shell shape, specific shell sculpture with zigzag ridges, deeper umbo cavity, and fixed nucleotide substitutions (Table 2).
Description: Shell shape oval to slightly elliptic, not very inequilateral, inflated, heavy and thick with zigzag ridges ranging from umbo, obscured in several adults. Posterior ridge somewhat oblique. Shell length (SL) 20.0–76.0 mm, height (SH) 14.6–56.4 mm, width (SW) 9.2–33.6 mm. Shell sculpture strong. Periostracum dark-olive to yellowish-brown; nacre white, sometimes with yellow spots. Umbo pronounced; beak sculpture strong. Left valve with two short curved lateral teeth and two striate pseudocardinals. Right valve with a single curved lateral tooth and two distinct pseudocardinals; anterior tooth strongly indented, massive, posterior tooth small and lamellar. Umbo cavity deep. Anterior adductor scar well-marked, very deep, funneled; posterior scar shallow.
Distribution: Sittaung River.
Tribe Oxynaiini Starobogatov, 1970
Type Genus: Oxynaia Haas, 1913 (designated in this study)
Comments: This tribe includes at least three phylogenetically distant genus-level clades: Radiatula Simpson, 1900 (see below), Indonaia Prashad, 1918 stat. res. (type species: Unio caeruleus Lea, 1831, by original designation; type locality: Bengal, India) and Oxynaia Haas, 1913 (type species: Unio jourdyi Morlet, 1886, by original designation; type locality: Tonkin, environs de Dang-son). The latter clade may belong to another, possibly undescribed genus, but the Oxynaia jourdyi sequences are not available. A single Indonaia species and three Oxynaia taxa were found in western Indo-China (Table 3).
Genus Radiatula Simpson, 1900
Type Species: Unio crispisulcatus Benson, 1862 (by original designation; type locality: rivulo Bangong, prope Thyet-Myo, regionis Burmanicae)
Comments: Five species of the genus are recorded in western Indo-China, including a new species from Sittaung (Table 3).
Radiatula mouhoti Vikhrev, Bolotov et Konopleva sp. nov.
Type material: Holotype RMBH biv256: Myanmar: Sittaung near Taungoo, 26.xi.2016, Vikhrev leg. Paratypes: Myanmar: the type locality, 5 specimens (RMBH biv253_1, biv253_6, biv248_4, biv248_1, and biv248_3), 25–26.xi.2016, Vikhrev leg.
Etymology: This species is named after Henri Mouhot (1826–1861), a French naturalist and explorer, who travelled across Siam, Cambodia and Laos.
Diagnosis: The new species is similar to R. aff. bonneaudii sp.1, but differs by a less deep and rounded anterior adductor scar, a well-marked posterior scar, and fixed nucleotide substitutions (Table 2).
Description: Shell elliptical, elongate, inequilateral, somewhat inflated, ventral margin straight; with thin wrinkles from umbo along dorsal margin; sculptured in the umbo area. Posterior ridge oblique, narrow. SL 32.2–48.7 mm, SH 24.0–28.7 mm, SW 11.6–19.7 mm. Periostracum smooth, olive, slightly greenish, some shells with dark-brown ventral margin, nacre yellowish with visible pallial line. Umbo not prominent, beak sculpture not well developed. Left valve with two short lateral teeth and two ribbed pseudocardinals. Right valve with a single lateral tooth and two distant pseudocardinals; anterior tooth strong, indented; posterior tooth slender. Umbo cavity rather deep. Anterior adductor scar pronounced, oval; posterior scar well-marked, rounded.
Distribution: Sittaung River.
Tribe Lamellidentini Modell, 1942
Type Genus: Lamellidens Simpson, 1900 (by original designation)
Comments: This tribe includes at least two distant genus-level clades: Lamellidens Simpson, 1900 and Trapezidens Bolotov, Vikhrev et Konopleva gen. nov.
Genus Lamellidens Simpson, 1900
Type Species: Unio marginalis Lamarck, 1819 (by original designation)
Comments: A large genus, which includes numerous nominal taxa24, 40, 43,44,45, 51, 52 and multiple cryptic DNA lineages, particularly from India9. In contrast, the species diversity of Lamellidens in western Indo-China was largely overestimated, as we were able to find only four species there (Table 3). Based on the DNA sequences and conchological features of the topotypes, Physunio ferrugineussyn. nov. and P. micropteroidessyn. nov., two nominal taxa from Lake Inle, are synonyms of Lamellidens generosus, which is the only member of the genus in the Salween River basin (Supplementary Fig. 3 and Table 3). It is another example of incorrect placement of Lamellidentini taxa within the Contradentini because of the convergent similarity in the shell shape25. In the Irrawaddy, we recorded only Lamellidens savadiensisstat. res. (Table 3). This species was also found in the Sittaung, where it lives together with a species new to science, which is described here.
Lamellidens brandti Bolotov, Konopleva et Vikhrev sp. nov.
Type material: Holotype RMBH biv243_14: Myanmar: Sittaung, Pathi River, 23.xi.2016, Vikhrev leg. Paratypes: the type locality, 3 specimens (RMBH biv242_3, biv243_17 and biv243_10), Myanmar: Sittaung, reservoir at Yetho River, 3 specimens (RMBH biv244_5, biv244_2 and biv244_3), Myanmar: Sittaung, fishing ponds near Taungoo, 1 specimen (RMBH biv247_10), Myanmar: Sittaung, Myit Kyi Pauk Stream, 1 specimen (RMBH no. biv250_13), 23–26.xi.2016, Vikhrev leg.
Etymology: This species is named after Dr. Rolf Arthur Max Brandt (1917–1989), a famous German malacologist, who studied the mussels of Southeast Asia.
Diagnosis: The new species is similar to L. savadiensis, but differs by a smaller shell, shallow anterior and posterior adductor scars, and fixed nucleotide substitutions (Table 2).
Description: Shell oval or trapezoidal, elongated, with more or less pronounced postdorsal wing, thin, inequilateral, not inflated. Posterior ridge broad, angular. SL 48.6–62.6 mm, SH 28.5–35.5 mm, SW 15.2–21.1 mm. Shell sculpture not strong. Periostracum smooth, brown, with yellowish concentric bands, mainly along the ventral margin; nacre white-bluish, shining. Umbo not prominent, beak sculpture slightly prominent. Left valve with two parallel lateral teeth and one lamellar pseudocardinal. Right valve with a single straight lateral tooth and two thin lamellar pseudocardinals. Umbo cavity not deep. Muscle adductor scars are shallow.
Distribution: Sittaung River Basin.
Genus Trapezidens Bolotov, Vikhrev et Konopleva gen. nov.
Type Species: Unio exolescens Gould, 1843 (type locality: Tavoy, Burmah)
Comments: A small remarkable genus with three species: T. exolescenscomb. nov., T. obesastat. res. et comb. nov. (with three subspecies) and T. scutumcomb. nov. (Table 3).
Diagnosis: Shell slightly thickened, of trapezoidal or elliptical shape, umbo indistinct and situated near the anterior end. Pseudocardinal teeth well developed, triangular; laterals strait and thick. Anterior adductor scar round and rather deep.
Etymology: This genus is named after its type species, Unio exolescens, which was erroneously considered a member of another genus, Trapezoideus Simpson, 1900, due to the convergent similarity in the shell shape25.
Distribution: Western Indo-China9.
Trapezidens obesa feae Kondakov, Konopleva et Vikhrev ssp. nov.
Type material: Holotype RMBH biv250_4: Myanmar: Sittaung, Myit Kyi Pauk Stream, 26.xi.2016, Vikhrev leg. Paratypes: the type locality, 3 specimens (RMBH biv250_7, biv250_8 and biv250_3), Myanmar: Sittaung near Taungoo, 1 specimen (RMBH biv255_1), 26.xi.2016, Vikhrev leg.
Etymology: This subspecies is named after Leonardo Fea (1852–1903), an Italian naturalist.
Diagnosis: The new subspecies differs by an almost straight dorsal margin without a wing, weakly developed adductor scars, a more oblong shell shape, and fixed nucleotide substitutions (Table 2).
Description: Shell elliptic, slightly inflated, very inequilateral, rather strong. Posterior ridge narrow, roundly angular. SL 70.2–95.2 mm, SH 35.1–45.6 mm, SW 19.6–28.6 mm. Periostracum blackish-brown with light-brown or sandy border along the ventral margin; nacre whitish. Left valve with two parallel straight lateral teeth and two pseudocardinals, anterior tooth well developed, posterior tooth small, undeveloped. Right valve with a single blade-shaped straight lateral tooth and two distant pseudocardinals, the lower tooth better developed, with small scratches. Anterior adductor scar not developed, of oval shape; posterior scar shallow.
Distribution: Sittaung River Basin.
Subfamily Pseudodontinae Frierson, 1927
Type Genus: Pseudodon Gould, 1844 (by original designation)
Comments: This remarkable subfamily, which was restored by Bolotov et al.,9 includes the two valid tribes: Pseudodontini Frierson, 1927 and Pilsbryoconchini Bolotov, Vikhrev et Tumpeesuwan tribe nov. Frierson55 (p. 67) did not provide any description of the subfamily, but noted that “The above species [Gonidea angulata (Lea, 1838)] is usually without ‘teeth’, but sometimes bears large high teeth, one in each valve, strikingly alike to the Pseudodon cambodgensis [sic.], and the outward facies of the two species are quite similar, so that it might be that the species in fact a member of the Pseudodontinae.”
Tribe Pseudodontini Frierson, 1927
Type Genus: Pseudodon Gould, 1844 (by original designation)
Comments: This tribe includes the single genus Pseudodon Gould, 1844. The genera Trigonodon Conrad, 1865 [type species: Monocondyloea crebristriata Anthony, 1865, by original designation; type locality: Pegu, British Burmah] and Indopseudodon Prashad, 1922 [type species: Anodon salweniana Gould, 1844; type locality: Salwen River, British Burmah] are considered synonyms of Pseudodon on the basis of conchological features.
Distribution: Western Indo-China.
Genus Pseudodon Gould, 1844
Type Species: Anodon inoscularis Gould, 1844 (by original designation; type locality: River Salwen, Tavoy, Brit. Burmah)
Comments: This genus contains seven species, including two new species from Sittaung (Table 3).
Diagnosis: Shell rather thick, of elliptical or round shape, umbo slightly prominent and situated near the anterior end. Pseudocardinal teeth tubercle-like, rather strong and prominent.
Distribution: Western Indo-China.
Pseudodon bogani Bolotov, Kondakov et Konopleva sp. nov.
Type material: Holotype RMBH biv241_5: Myanmar: Sittaung, Kanni River, 23.xi.2016, Vikhrev leg. Paratypes: the type locality, 4 specimens (RMBH biv241_8, biv241_4, biv241_7 and biv241_6), 23.xi.2016, Vikhrev leg.
Etymology: This new species is dedicated to Dr. Arthur Bogan, a famous American malacologist, who developed the basic principles for the conservation of freshwater mussels worldwide.
Diagnosis: The new species is similar to P. avae and P. manuelisp. nov., but differs by a straighter ventral margin, massive hinge plate, and fixed nucleotide substitutions (Table 2).
Description: Shell ovate to elliptical, thick, very inequilateral, moderately inflated, umbo area ornamented by W-shaped wrinkles. Posterior ridge broader than the anterior ridge. SL 56.0–62.6 mm, SH 32.3–40.0 mm, SW 17.0–21.8 mm. Shell sculpture strong. Periostracum not smooth, with visible concentric bands, dark-brown; nacre peach-colored, yellowish. Umbo not projecting, corrugated, beak sculpture not prominent. A massive tubercle-like pseudocardinal tooth in each valve. Umbo cavity shallow. Anterior adductor scar well-marked, bean-like, rather deep; posterior scar rounded, visible.
Distribution: Kanni River.
Pseudodon manueli Konopleva, Kondakov et Vikhrev sp. nov.
Type material: Holotype RMBH biv246_3: Myanmar: Sittaung, Pyowne Stream, 25.xi.2016, Vikhrev leg. Paratypes: the type locality, 4 specimens (RMBH biv246_1, biv246_8, biv246_6 and biv246_2), 25.xi.2016, Vikhrev leg.
Etymology: This species is dedicated to Dr. Manuel Lopes-Lima, a malacologist from Portugal, who developed a recent integrative system for the Unionidae.
Diagnosis: The new species is similar to P. boganisp. nov. and P. avae, but differs by less pronounced pseudocardinals, adductor scars not produced, and fixed nucleotide substitutions (Table 2).
Description: Shell ovate to elliptical, not very thick, inequilateral, moderately inflated, dorsal margin ornamented by W-shaped wrinkles. Posterior ridge broader than the anterior ridge. SL 65.8–82.1 mm, SH 41.5–52.3 mm, SW 23.1–28.3 mm. Periostracum light brown; nacre white-yellowish. Umbo slightly projecting, somewhat corrugated, beak sculpture not strong. A small tubercle-like pseudocardinal in each valve. Umbo cavity shallow. Anterior adductor scar poorly visible; posterior scar shallow.
Distribution: Pyowne Stream.
Tribe Pilsbryoconchini Bolotov, Vikhrev et Tumpeesuwan tribe nov.
Type Genus: Pilsbryoconcha Simpson, 1900
Comments: This tribe includes at least three valid genera: Pilsbryoconcha Simpson, 1900 (type species: Anodonta exilis Lea, 1838, by original designation; type locality: unknown), Monodontina Conrad, 1853 stat. res. (type species: Margaritana vondembuschiana Lea, 1840, by original designation; type locality: Java), and Bineurus Simpson, 1900 stat. res. (type species: Monocondyloea mouhotii Lea, 1863, by original designation; type locality: Laos Mts., Cambodia, Siam). However, there are three distant, relatively well-supported clades, which may represent distinct genera (Fig. 2). An integrative revision of this large tribe should be undertaken in the future based on expanded sampling from the Mekong, as the phylogenetic relationships between these clades are still unresolved (Fig. 2).
Diagnosis: Shell rather thin, of elliptical or elongated shape, umbo slightly prominent and situated near the anterior end. Pseudocardinals reduced or lacking; laterals reduced.
Distribution: Paleo-Mekong basin9.
Subfamily Rectidentinae Modell, 1942
Comments: This subfamily includes two tribes: Rectidentini Modell, 1942 (type genus: Rectidens Simpson, 1900) and Contradentini Modell, 1942 (see below).
Tribe Contradentini Modell, 1942
Type Genus: Contradens Haas, 1911 (by original designation)
Comments: This tribe includes at least three genus-level phylogenetic clades: Contradens Haas, 1911, Trapezoideus Simpson 1900, and Physunio Simpson, 1900 (Fig. 2).
Genus Trapezoideus Simpson 1900
Type Species: Unio foliaceus Gould, 1843 (by original designation; type locality: Tavoy, Burmah)
Comments: The genus was established on the basis of conchological features39, 40. We consider this genus as a separate Contradentini clade, which includes taxa inhabiting the rivers of western Indo-China, although the molecular sequences of the type species are not available25. The genus contains at least six species, including two new taxa from the Sittaung (Table 3). Among the other taxa that were assigned to the genus, Trapezoideus peninsularis Simpson, 1900 [type locality: Sumatra] is surely a member of Contradens, but the position of Indian taxa such as T. prashadi Haas, 1922 [type locality: Mysore, Sudöstindien] and T. theca (Benson, 1862) [type locality: fluvio Cane, prope Banda, Bundelkhund] is still uncertain. The two latter taxa may belong to Lamellidens33, 46, but their DNA sequences are not available.
Diagnosis: Shell rather thin, from a trapezoidal to elongate-elliptic form, the anterior margin narrower than the posterior margin. One lateral tooth and two pseudocardinals in the right valve, two laterals and elongate lamellar cardinal tooth in the left valve. Pseudocardinals reduced or lacking. Anterior adductor scar more or less produced, posterior scar usually less marked. Umbo more or less projected. Older specimens with more elongate shape and flattened umbo.
Distribution: Western Indo-China.
Trapezoideus nesemanni Konopleva, Vikhrev et Bolotov sp. nov.
Type material: Holotype RMBH biv255_2: Myanmar: Sittaung, Tauk Ue Kupt River, 26.xi.2016, Vikhrev leg. Paratypes: the type locality, 4 specimens (RMBH biv255_3, biv_144_14, biv_144_25, and biv_144_19), 20.iv.2015 and 26.xi.2016, Bolotov, Vikhrev & locals leg.
Etymology: This new species is dedicated to Dr. Hasko Friedrich Nesemann, an Austrian malacologist, who made great contributions to the knowledge of molluscs from the Ganga River system.
Diagnosis: The new species is similar to T. subclathratus and T. panhaisp. nov., but differs in its very thin shell, shallow adductor scars, and fixed nucleotide substitutions (Table 2).
Description: Shell trapezoidal, thin, not inflated, anterior side rounded and narrow. Posterior ridge broad and sloped. SL 28.6–82.5 mm, SH 16.4–39.4 mm, SW 7.8–22.7 mm. Shell sculpture not strong. Periostracum brown to dark-brown; nacre yellowish. Umbo not prominent, with wrinkles in the umbo area, corrugated. Beak sculpture not strong, slightly pronounced. Right valve with two slightly beaked, flat pseudocardinals and one lateral tooth. Left valve with one elongate, flat pseudocardinal tooth and two laterals. Several specimens with a single thin lateral tooth in each valve and a single shallow, lamellar pseudocardinal in each valve. Umbo cavity shallow. Anterior adductor scar not deep, drop-like in shape. Posterior scar rounded, shallow.
Distribution: Tauk Ue Kupt River.
Trapezoideus panhai Konopleva, Bolotov et Kondakov sp. nov.
Type material: Holotype RMBH biv_138_4: Myanmar: Sittaung, Kyan Hone River, 17.iv.2015, Bolotov leg. Paratypes: the type locality, 5 specimens (RMBH biv_138_7, biv_155_4, biv_155_25, biv_138_12, biv_155_11), 17–18.iv.2015, Bolotov & locals leg.
Etymology: This new species is dedicated to Prof. Dr. Somsak Panha, a famous Thai zoologist, who described numerous molluscan taxa from Southeast Asia.
Diagnosis: The new species is similar to T. subclathratus and T. nesemannisp. nov., but differs by a thicker shell, more developed hinge and umbones, and fixed nucleotide substitutions (Table 2).
Description: Shell oval, elongated, with narrower anterior side and broader posterior end, rather thick, inequilateral. Ventral margin concave or straight. Posterior ridge broad. SL 31.7–47.2 mm, SH 18.2–26.1 mm, SW 10.9–17.9 mm. Shell sculpture rather strong. Periostracum brown to olive-brown; nacre whitish. Umbo somewhat prominent, beak sculpture not very strong, pronounced. Hinge plates with one lateral tooth and two pseudocardinals in the right valve, and two laterals and elongate lamellar cardinal in the left valve. Umbo cavity not very deep. Anterior adductor scar drop-like and deep; posterior scar somewhat oval-shaped and well-marked.
Distribution: Kyan Hone River.
The taxonomy of the Unionidae from western Indo-China is complicated9, 10, 25. Our study is the first attempt to establish an updated taxonomic scheme for the fauna of this region based on an integrative approach (Table 3). The results of this investigation indicate that the Irrawaddy represents the most species-rich basin (16 taxa, including 10 species sequenced), followed by the Sittaung (10 species, all sequenced), Salween (6 taxa, 2 species sequenced), and Tavoy (4 taxa, 3 species sequenced). The faunas of other basins are poorly known. Only five nominal taxa were described from the Pegu River, and none of them were sequenced. The small and medium sized basins of the western coast of Myanmar (Bay of Bengal and Andaman Sea) are almost unstudied.
There is a historical tradition to attribute numerous unionid species from western Indo-China to those described from Indian rivers, which began with the pioneering works36,37,38, 46 followed by the subsequent revisions39, 40, 43,44,45, 50, 51. This view was accepted until recently10, 24, but none of the true Indian taxa are actually distributed in western Indo-China9. For example, the records of Lamellidens corrianus (Lea, 1834), L. jenkinsianus (Benson, 1862), L. lamellatus (Lea, 1838), L. marginalis (Lamarck, 1819), Parreysia corrugata (Müller, 1774), P. favidens (Benson, 1862), P. smaragdites (Benson, 1862), Radiatula bonneaudii (Eydoux, 1838), R. caerulea (Lea, 1831), and R. pachysoma (Benson, 1862) in the Irrawaddy10, 24, 44 are erroneous. However, the occurrences mentioned above may correspond to morphologically similar but phylogenetically distinct, endemic taxa such as Lamellidens savadiensis, Trapezidens obesa, Leoparreysia spp., Radiatula aff. bonneaudii sp.1, and Indonaia andersoniana. The published records of Lamellidens consobrinus (Lea, 1860), L. generosus and Physunio ferrugineus from the Irrawaddy10 most likely refer to Lamellidens savadiensis.
Biogeography and conservation
Our modeling confirms that the primary Indo-Chinese Unionidae clades are of Mesozoic origin and that the most ancient intra-area radiations occurred within the putative paleo-Mekong basin. These results agree with the model of Bolotov et al.9, although the present study suggests that the two largest paleo-Mekong radiations may have had a pre-Cenozoic origin (mean age = 65–71 Ma). Based on our fossil-calibrated phylogeny, the fauna of the Sittaung is related to that of the Irrawaddy but represents a separate evolutionary entity harboring a number of endemic lineages. Our model suggests that the majority of the lineages were separated during the Miocene (mean age = 8.0–22.1 Ma), but the split between Pseudodon avae and the two sister taxa from the Sittaung appears to be more ancient (mean age = 35.3 Ma) (Supplementary Fig. 2). These results are consistent with the model indicating that the major rivers of Indo-China represent ancient evolutionary hotspots with almost 100% level of endemism in the Unionidae9. In contrast, our modeling contradicts the hypothesis of an Early Miocene paleo-drainage joining the Salween and Mekong56, 57 because the Salween fauna relates to those of the Sittaung and Irrawaddy but not the Mekong. The most important biogeographic boundary between the unionid faunas in Indo-China is situated along the Mekong – Salween watershed. The Sittaung is an example of the medium-sized drainages, which are less resistant to human activities, e.g., dam construction, channelization and water pollution58. This unique basin should therefore be a focus of international conservation efforts alongside the largest Southeast Asian rivers.
The electronic edition of this article conforms to the requirements of the amended International Code of Zoological Nomenclature (ICZN), and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank (http://zoobank.org), the online registration system for the ICZN. The LSID for this publication is: urn:lsid:zoobank.org:pub:1CC25D33–7DF8–41D5–9817–5285C01A7779. The electronic edition of this paper was published in a journal with an ISSN, and has been archived and is available from PubMed Central.
Studies of the type series of the Oriental taxa
The type specimens were studied in the malacological collections of the National Museum of Natural History, Smithsonian Institution, Washington, DC, USA (NMNH), the British Museum of Natural History, London, UK (NHMUK), and the Museo Civico di Storia Naturale di Genova, Genoa, Italy (MSNG). Additionally, we accessed the images of the types of several nominal taxa at the MUSSELp Database59.
Taxon sampling and laboratory protocols
Our phylogeny of the Unionidae was based on 469 representatives of 138 in-group species from the Oriental Region (Indo-China and India), East Asia, Europe, Africa and North America (Supplementary Table 1). This sample includes all primary clades of the Unionidae, which were determined in the recent studies9, 17. The majority of these sequences were sampled from our work on the biogeography of the Oriental freshwater mussels9. New sequences were obtained from 74 specimens belonging to 15 species that were collected from the Sittaung and Irrawaddy river basins, and from a river of the Malay Peninsula (Supplementary Tables 1 and 2). All of our voucher specimens are deposited in RMBH, Russian Museum of Biodiversity Hotspots, the Federal Center for Integrated Arctic Research, Russian Academy of Sciences (Arkhangelsk, Russia). Total genomic DNA was extracted from 95% ethanol-preserved tissue samples using the NucleoSpin® Tissue Kit (Macherey-Nagel GmbH & Co. KG, Germany), following the manufacturer’s protocol. For molecular analyses we obtained partial sequences of two mtDNA markers, i.e., the cytochrome c oxidase subunit I gene (COI) and the 16S ribosomal RNA (16S rRNA), and a fragment of the nuclear 28S ribosomal DNA (28S rDNA). Primer sequences for PCR are shown in Supplementary Table 4. Thermocycling was implemented with marker-specific PCR programs as follows: (i) COI: 95 °C (4 min), followed by 37 cycles at 94 °C (50 sec), 50 °C (50 sec), 72 °C (50 sec) and a final extension at 72 °C (5 min); (ii) 16S rRNA: 95 °C (4 min), followed by 33 cycles at 94 °C (50 sec), 47 °C (50 sec), 72 °C (50 sec) and a final extension at 72 °C (5 min); (iii) 28S rDNA: 95 °C (4 min), followed by 38 cycles at 94 °C (50 sec), 57 °C (50 sec), 72 °C (50 sec) and a final extension at 72 °C (5 min). Forward and reverse sequence reactions were performed directly on purified PCR products using the ABI PRISM® BigDye™ Terminator v. 3.1 reagents kit and run on an ABI PRISM® 3730 DNA analyzer (Thermo Fisher Scientific Inc., Waltham, MA, USA). The resulting sequences were checked by eye using a sequence alignment editor (BioEdit v. 7.2.5)60. A total of 25 mussel species were used as an out-group, including representatives of Margaritiferidae (10 species), Iridinidae (2 species), Etheriidae (1 species), Mycetopodidae (1 species), Hyriidae (9 species) and Trigoniidae (2 species) (Supplementary Table 1).
Sequence alignment, checking the congruence of phylogenetic signals and substitution saturation analyses
The sequence alignment of COI, 16S rRNA and 28S rDNA gene fragments was performed separately using the Muscle algorithm implemented in MEGA661. The aligned sequence data sets were checked through GBlocks v. 0.91b62 which allow to exclude hypervariable fragments from the sequence alignments using options for less stringent selection, enabling gap positions, smaller final blocks and less strict flanking positions. The resulting lengths of the sequence alignments are listed in Supplementary Table 5. To estimate each partition for evidence of substitution saturation, we performed the test of Xia et al.63 using DAMBE v. 5.3.10864, which showed little saturation even under the assumption of an asymmetrical tree (P < 0.001). A partition homogeneity test was calculated in PAUP* v. 4.0a151 to confirm the congruence of phylogenetic signals among sequence data sets65. This test revealed that the signals among the data sets are consistent (Supplementary Table 6).
Phylogenetic analyses and divergence time estimates
The alignment data sets were joined in combined multi-gene nucleotide sequence alignments and collapsed into unique haplotypes (Supplementary Table 1) using an online FASTA sequence toolbox (FaBox 1.41)66. Absent sites were treated as missing data. For phylogenetic analyses, we used the resulting combined data set with unique haplotypes, including those that were possibly identical but differed by the availability of gene partitions for certain specimens.
In phylogenetic analyses, we tested only the dataset with five partitions (3 codons of COI + 16S rRNA + 28S rDNA), because Bolotov et al.9 show that the combined phylogeny of the Indo-Chinese Unionidae corresponds to those obtained from separate partitions. The ML phylogenetic analysis was conducted using RAxML v. 8.2.6 HPC Black Box67 at the San Diego Supercomputer Center through the CIPRES Science Gateway68. A unique GTR model was applied for each partition with corrections for gamma distribution. Nodal support values were estimated using an automatic, rapid bootstrapping algorithm according to the developer’s recommendation67, and the majority rule consensus tree was constructed from the independent searches. Bayesian inference (BI) analyses were performed in MrBayes v. 3.2.669 at the San Diego Supercomputer Center through the CIPRES Science Gateway68. The data set tested was similar under the ML model. The best models of sequence evolution for each partition based on the corrected Akaike Information Criterion (AICc) of MEGA661 are presented in Supplementary Table 8. Two runs, each with three heated (temperature = 0.1) and one cold Markov chain, were conducted for 15 million generations. Trees were sampled every 1000th generation. After completion of the MCMC analysis, the first 15% of trees were discarded as burn-in (pre-convergence part), and the majority rule consensus tree was calculated from the remaining trees. Convergence of the MCMC chains to a stationary distribution was checked visually based on the plotted posterior estimates using an MCMC trace analysis tool (Tracer v. 1.6)70. The effective sample size (ESS) value for each parameter sampled from the MCMC analysis was always recorded as >1300.
We estimated the acceptance of a global molecular clock to our multi-gene data set using the maximum likelihood test of MEGA661, which revealed that the null hypothesis of equal evolutionary rate throughout the tree was rejected at a 5% significance level (p < 0.001). Thus, the time-calibrated haplotype-level Bayesian phylogeny was reconstructed in BEAST v. 1.8.4 based on multiple fossil calibration points using a lognormal relaxed clock algorithm with the Yule speciation process as the tree prior71,72,73. Calculations were performed at the San Diego Supercomputer Center through the CIPRES Science Gateway68. A fossil-calibrated ultrametric tree was obtained using BEAST v. 1.8.4. We specified similar settings to five partitions (3 codons of COI + 16S rRNA + 28S rDNA) as in the MrBayes analyses, but by using simplified evolutionary models. The HKY model was applied to each partition instead the GTR model, because the prior and posterior ESS values under the GTR model were recorded always <100. This indicates that the GTR model is likely overly complex for our data9. The eight published fossil calibrations were used for timing of the phylogeny9. We designated priors for out-group taxa using a “Monophyly” option of BEAUti v. 1.8.473 as follows: (Trigoniidae, (Unionida)). Four replicate BEAST searches were conducted, each with 50 million generations. The trees were sampled every 5,000th generation. The log files were checked visually with Tracer v. 1.6 for an assessment of the convergence of the MCMC chains and the effective sample size of parameters70. The first 10–52% of trees were discarded as an appropriate burn-in. All the ESS values were recorded as >250, with exception of two parameters with the ESS >120; the posterior distributions were similar to the prior distributions. The resulting tree files from four independent analyses were compiled with LogCombiner v. 1.8.473. The maximum clade credibility tree was obtained from 31,404 primary trees using TreeAnnotator v. 1.8.473.
Species delimitation analyses
The preliminary delimitation of biological species was based on a molecular approach using the MOTU concept74,75,76,77,78. MOTUs were separated based on the bPTP model78 to infer putative species boundaries on a phylogenetic input tree inferred from a maximum likelihood (ML) analysis. The ML analysis was conducted based on an alignment of the COI haplotype sequences of Parreysiinae and Pseudodontinae + Rectidentinae groups separately using RAxML v. 8.2.6 HPC Black Box67 at the San Diego Supercomputer Center through the CIPRES Science Gateway68. A unique GTR model was applied for each partition (three codons of COI) with corrections for gamma distribution. Margaritifera laosensis and M. dahurica were used as an out-group for the each group. We used an implementation of the bPTP model thorough online bPTP server (http://species.h-its.org/ptp) with 500,000 Markov Chain Monte Carlo (MCMC) generations and 10% burn-in78. All out-group taxa were removed from the input tree using an appropriate option of the server. The output parameters of the bPTP model for each clade under the highest Bayesian solution were as follows: (i) Parreysiinae: an estimated number of species = 40–88, mean value = 61.1, acceptance rate = 0.46; and (ii) Pseudodontinae + Rectidentinae: an estimated number of species = 45–69, mean value = 53.5, acceptance rate = 0.31. Additionally, MOTUs were obtained using the sPTP (p < 0.001) and mPTP models of Kapli et al.79 thorough online mPTP server (http://mptp.h-its.org). A phylogenetic input tree was obtained from ML analysis, which was conducted based on an alignment of the COI haplotype sequences of the Oriental Unionidae with the two haplotypes of Margaritifera laosensis and M. dahurica as an out-group using RAxML v. 8.2.6 HPC Black Box67.
In order to diagnose each new species, we followed a two-step procedure of Delić et al.80. Firstly, we estimate the morphological differences between a new species and closely related (congeneric) taxa. The comparative analysis of the shell morphology was carried out with attention to the structure of the pseudo-cardinal and lateral teeth, muscle attachment scars, shell shape and umbo position25. Secondly, the molecular diagnosis of each new species was provided using the fixed nucleotide differences80,81,82, which were estimated for each gene separately using a Toggle conserved sites tool of MEGA6 at 50% level61. For each species, an alignment of congeneric haplotype sequences was performed using the ClustalW algorithm implemented in MEGA661. All the deleterious mutations were retained for the analyses. Additionally, a COI mean p-distance to the nearest neighbor of each species was calculated in MEGA661.
The sequences generated under this study are available from GenBank. Accession numbers for each specimen are presented in Supplementary Table 1. The type specimens of the new species and voucher specimens of the other taxa that we studied are available in the RMBH, Russian Museum of Biodiversity Hotspots, the Federal Center for Integrated Arctic Research, Russian Academy of Sciences (Arkhangelsk, Russia).
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank the Associate Editor Dr. Ciro Rico and three anonymous reviewers for their helpful comments. This work was partly funded by grants from the Russian Ministry of Education and Science (project no. 6.2343.2017/4.6), the Federal Agency for Scientific Organizations (project nos. 0410-2014-0028 and 0409-2016-0022), the President of the Russia Grant Council (no. MD-7660.2016.5), the Russian Foundation for Basic Research (no. 16-34-00638), and Northern Arctic Federal University. We are grateful to Dr. Tony Whitten, Mr. Frank Momberg, Mr. Zau Lunn, and Mr. Nyein Chan (Fauna & Flora International, Myanmar), Prof. Dr. Myin Zu Minn and Prof. Dr. Thida Lay Thwe (University of Yangon, Myanmar), Prof. Maxim V. Vinarski (Saint Petersburg State University, Russia), Dr. Alexander P. Novoselov (PINRO, Russia), and Mr. ‘Theo’ Ko Htay Aung (Myanmar) for their great help during this study. We would like to express our sincerest gratitude to the Department of Fisheries of the Ministry of Livestock, Fisheries and Rural Development (Myanmar) and personally to Mr. Myint Than Soe for the permission of the field work and sampling in Myanmar (sampling permission no. 15/6000/MOAI-3103/2016 and export permission no. MOAI/2016-5856).
Electronic supplementary material
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.