Vietnam, a Hotspot for Chromosomal Diversity and Cryptic Species in Black Flies (Diptera: Simuliidae)

The increasing attention on Vietnam as a biodiversity hotspot prompted an investigation of the potential for cryptic diversity in black flies, a group well known elsewhere for its high frequency of isomorphic species. We analyzed the banding structure of the larval polytene chromosomes in the Simulium tuberosum species group to probe for diversity beyond the morphological level. Among 272 larvae, 88 different chromosomal rearrangements, primarily paracentric inversions, were discovered in addition to 25 already known in the basic sequences of the group in Asia. Chromosomal diversity in Vietnam far exceeds that known for the group in Thailand, with only about 5% of the rearrangements shared between the two countries. Fifteen cytoforms and nine morphoforms were revealed among six nominal species in Vietnam. Chromosomal evidence, combined with available molecular and morphological evidence, conservatively suggests that at least five of the cytoforms are valid species, two of which require formal names. The total chromosomal rearrangements and species (15) now known from the group in Vietnam far exceed those of any other area of comparable size in the world, supporting the country’s status as a biodiversity hotspot. Phylogenetic inference based on uniquely shared, derived chromosomal rearrangements supports the clustering of cytoforms into two primary lineages, the Simulium tani complex and the Southeast Asian Simulium tuberosum subgroup. Some of these taxa could be threatened by habitat destruction, given their restricted geographical distributions and the expanding human population of Vietnam.


Introduction
Vietnam moved to center stage in the 1990s as a hotspot for biodiversity and endemism when new species of large mammals were discovered [1]. These discoveries were not isolated examples of Vietnam's remarkable biodiversity. New species have been discovered in nearly every investigated group of animals and plants in Vietnam, highlighting the country's standing as 25 th in the world in species richness [2] despite ranking 65 th in total area. The wealth of Vietnam's biodiversity derives from a complex climatic and geological history, significant elevational (0-3143 m asl) and latitudinal gradients (8.4-23.4°N), and a subtropical-tropical setting with diverse ecoregions [2][3][4].
Invertebrate diversity in Vietnam is rich but woefully underexplored. Two of the bestknown groups, butterflies and mosquitoes, provide a general index of richness in the country. More than 1100 species of butterflies [5] and about 226 species of mosquitoes [6] are known from Vietnam-more than 6% of each group's total world fauna. The insect diversity in Vietnam has been explored largely through conventional morphological approaches. The actual extent of biodiversity is probably far greater when cryptic species are considered [7]. Taking a lead from amphibian studies, which have recognized two to six times the number of each putative species of frog in Vietnam [8,9], biologists might expect comparable cryptic species richness in insects.
Among the insect groups best known for repetitive discoveries of cryptic species are the black flies (Simuliidae) and mosquitoes [10,11]. A cryptic species of mosquito, for example, was discovered in Vietnam when malarial vectors were investigated [12], and three additional species of black flies were revealed among two nominal species when molecular techniques were applied [13]. The intricate banding patterns of polytene (giant) chromosomes provide a time-tested means of revealing cryptic species of black flies through evidence of reproductive isolation [14]. The taxonomic framework for the family Simuliidae now rests in significant part on characters of the polytene chromosomes [11,14].
Indications that Vietnam is a hotspot for biodiversity in the Simuliidae are based on recent surveys in three of the 58 provinces, which increased the country's number of known species to 46, including 22 (48%) described as new [15][16][17][18]. Our objective was to explore the biodiversity that might further be revealed in the macrogenome of a single species group of black flies in Vietnam. We selected the Simulium tuberosum species group, based on an opportunity to compare our findings with those of a molecular study of two nominal species in the group in Vietnam [13] and with the extensive cryptic taxa discovered in the group in Thailand [19]. The Simulium tuberosum group is a well-defined clade [20] of more than 50 nominal species distributed across the Holarctic Region deep into the Oriental Region [21]. The Holarctic namesake (Simulium tuberosum sensu stricto) for the group provided one of the earliest examples of chromosomal discovery of cryptic species in the family Simuliidae [22].

Ethics statement
All samples were collected on public land with access from public roads. No permissions were required to access sites or collect material, and the collections did not involve endangered or protected species.

Collection and preparation of material
Larvae and pupae were collected with forceps from substrates in 16 streams in Vietnam, spanning more than 1150 km of the country's length, plus 2 streams in Malaysia to aid species identifications (Table 1). They were fixed in ethanol or (larvae only) in 1:3 glacial acetic acid:95% ethanol (Carnoy's fixative). Adults were allowed to emerge from additional pupae to facilitate morphological identifications. Larvae in Carnoy's fixative were sorted into morphotaxa, based on their key characters and those of associated life stages [18]. Polytene chromosomes were prepared according to standard Feulgen-staining procedures [23].
Carcasses of all chromosomally examined larvae and photographic negatives of chromosomes are deposited in the Clemson University Arthropod Collection. Additional larvae and associated life stages are deposited in the Institute of Biological Sciences, Faculty of Science, University of Malaya, Kuala Lumpur, Malaysia.

Chromosomal mapping and analyses
Chromosomal mapping procedures, conventions, and terminology follow established procedures [19,24]. Chromosomal banding patterns of larvae were compared with the standard banding sequence of the subgenus Simulium and the Simulium tuberosum group [19,24]. Section numbers on our maps follow those for the S. tuberosum group [19]. Inversions discovered in our material, which are shared with species previously known chromosomally in Southeast Asia [19], are given the same number. Newly discovered inversions are numbered to follow the last-used number in each chromosome arm in previous treatments [24] of the S. tuberosum species group. Fixed inversions within a cytoform are italicized; polymorphic inversions are not. Each heteroband (hb; thickened band relative to the standard), heterochromatic block (hc; insertion of heterochromatin between existing bands), and finer band insertion (in) is coded by the arm and section number in which it occurs (e.g., IS 13hb, IIIL 100hc, and IIIL 85in, respectively). All chromosomal rearrangements were indicated with precise locations and breakpoints on our maps. We use the following previously applied [25] definition of cytoform: a chromosomally distinct entity recognizable at an individual or a population level, without regard to whether the entity is part of a larger breeding population (cytotype) or is reproductively isolated (cytospecies). New cytoforms of the S. doipuiense and S. tani complexes were named to follow the lastrecognized cytoforms ('B' and 'L' , respectively) [21]. New cytoforms of the nominal species S. brevipar, S. rufibasis, and S. yuphae, not previously known to contain cytoforms, were each designated Cytoform 'B' (and Cytoform 'C' for S. brevipar), while the original chromosomally studied population of each was assigned, retrospectively, to Cytoform ' A' .
We inferred a phylogeny based on uniquely shared, derived chromosomal rearrangements, primarily inversions, from the polytene complement. We used a two-step procedure [24]. Briefly, we first resolved all rearrangements in our material, relative to the Simulium subgeneric standard for the IS, IL, IIL, and IIIS arms [26] and the IIS and IIIL arms [19,27]. To provide directionality, we then rooted the phylogeny by resolving the subgeneric standard where possible, particularly for the entire IIIL arm, relative to the common sequences [28,29] in two outgroups, Simulium (Boophthora) erythrocephalum and Simulium (Psilozia) vittatum.

Results
The banding patterns of 280 larvae (including 8 from Malaysia) were analyzed completely; the chromosomes of 18 additional larvae (6.0%) were not of sufficient quality for full resolution, and were not included in any tabulations or analyses. A total of 88 chromosomal rearrangements, primarily (86.4%) paracentric inversions, but also differential band expressions (13.6%), were discovered in Vietnamese material (plus 1 additional novel inversion in our Malaysian samples), relative to the standard sequence for the S. tani complex and the Southeast Asian S. tuberosum species subgroup. Rearrangements were concentrated (69.3%) in the IIIL arm. Chromocenters, ectopic pairing of centromeres, and supernumerary (B) chromosomes were absent.
The cytoforms fell into 2 previously defined [19] lineages: the Simulium tani complex and the Southeast Asian Simulium tuberosum subgroup. We describe each cytoform under its chromosomally assigned name. Table 1 links the initial morphological identification of each cytoform with its chromosomal designation, and Table 2 summarizes the diagnostic information for each cytoform in Vietnam (plus one in Malaysia). the S. tani complex of Southeast Asia, which included 3 inversions in IL, 6 inversions in IIS, 3 inversions in IIL, and 3 inversions in IIIL (IIIL-1, IIIL-2, and IIIL-3) [19]. These inversions, therefore, are not repeated in descriptions of the taxa presented here, nor in Table 3. Figs 1-7 show all rearrangements discovered in our material of the S. tani complex.
Simulium tani Cytoform 'B2' . We analyzed the banding patterns of all 48 chromosome preparations of larvae from Site 5. Fixed inversions were absent. IL-2 was absent, in contrast to its preponderance in 'B' in Thailand [19]; on this basis, we recognized 2 subunits of Cytoform 'B': 'B1' in Thailand and 'B2' in Vietnam. The sex chromosomes were undifferentiated. Polymorphisms were scarce; 1 male and 1 female were heterozygous for IS-25 (Fig 1) and IL-15 (Fig 3), respectively (Table 3).

Southeast Asian Simulium tuberosum subgroup
We recognized 11 cytoforms among 6 morphoforms in the Southeast Asian S. tuberosum subgroup in Vietnam (Table 1). Relative to the Simulium subgeneric standard sequence, all Vietnamese members of this subgroup shared IL-1, IL-tuberosum, 4 inversions in IIS, 3 inversions in IIL, and IIIL-1 [19]; these inversions are not repeated in the descriptions below or in Table 4. Figs 1, 2, 4, and 8-12 depict all rearrangements discovered in the subgroup.

Simulium brevipar Cytoform 'A' .
To establish the chromosomal characteristics of S. brevipar, we examined 8 larvae from the Cameron Highlands of Malaysia (Sites 17, 18) about 90 km from the type locality. Accepting these larvae as chromosomally representative of the type of S. brevipar, and assigning them to Cytoform ' A' , we found that our 2 samples, albeit small, were cohesive. Larvae were fixed for IIIL-13. The only polymorphism (autosomal) was IS-26 (Fig 2A), homozygous in 4 males, heterozygous in 2 females and 1 male, and standard in 1 female ( Table 4). The small sample revealed no evidence of differentiated sex chromosomes.
Simulium doipuiense Cytoform 'C' . We analyzed 1 larva (female, Site 3) of this species, mixed with a sample identified morphologically as S. congi. IIIL-11 and IIIL-13 were present. IS-32, IL-16, and IIIL-79 also were homozygous (Table 4; Figs 1A, 2B, and 8B), but whether they actually were fixed, as provisionally represented, could not be determined without larger samples. Although only a single larva was found, we tentatively recognized it as a separate cytoform on the strength of homozygous inversions IS-32, IL-16, and IIIL-79.
Simulium doipuiense Cytoform 'F' . The single larva (female, Site 15) of this cytoform had a novel IIIL banding pattern with 3 homozygous inversions-IIIL-87, IIIL-88, and IIIL-89-on  top of IIIL-11,13 (Table 4; Fig 12B). We provisionally show IIIL-87,88,89 as fixed (italicized), pending more material. Our interpretation of bands in sections 93/93/94, and consequently the included breakpoints, is tentative. The banding sequence of the remainder of the polytene complement conformed to that of the Southeast Asian S. tuberosum subgroup. The larva was coinfected with 2 unknown species of microsporidia. Simulium rufibasis Cytoform 'B' . All 14 larvae of this species from Sites 11-13 were analyzed completely. IIIL-8, IIIL-11, and IIIL-13 were fixed (Table 4; Fig 12A). IIIL-12 (Fig 9A), previously known only as a fixed inversion in S. yuphae and S. 'unknown sp. 2' of Tangkawanit et al. [19], was an autosomal polymorphism at Site 12. IIIL-60 was a common autosomal polymorphism, and IIIL-61 was heterozygous in 1 male larva (Fig 12A). Two heterobands in IIIL (87hb and 96hb) were found as single heterozygotes (Figs 11A and 12A). IIS-4, an inversion originally interpreted as present in S. rufibasis [19] was absent in our larvae; however, the symmetry of bands within this small inversion in some preparations suggests that the original [19] interpretation of the presence of this inversion in S. rufibasis might have been erroneous. IIIL-64 (Fig 9B), which appeared heterozygously in the only larva (female) of the S. rufibasis complex at Site 11, was shared with S. doipuiense ' A' at the same site. The presence of IS-29,31 +13hb (= Y 1 ) in 7 of 9 males but in none of the 3 females at Site 12 suggests that IS is the sex arm (Table 6; Fig 1A). IS-27 (X 1 ) and IS-28 (X 2 ) might represent alternative X sequences to the undifferentiated X 0 ; males also were polymorphic for an undifferentiated Y 0 chromosome ( Table 6; Fig 1A). We tentatively recognize our Vietnamese material as a new cytoform, 'B' , on the basis of probable differentiated sex chromosomes, in contrast to material of S. rufibasis with cytologically undifferentiated sex chromosomes, previously analyzed from Thailand [19], and recognized here, retrospectively, as Cytoform ' A' . Simulium rufibasis 'B' and S. doipuiense ' A' occurred in the same streams and were reproductively isolated from one another. Simulium yuphae Cytoform 'A' . We analyzed 21 larvae from Sites 6-9. The larvae, identified morphologically as S. cavum Takaoka & Ya'cob, were chromosomally classic for S. yuphae, having IIIL-12 and IIIL-13 (Fig 9A), with undifferentiated sex chromosomes and few polymorphisms. One male larva from Site 6 and 1 female from Site 7 had the typical sequence for S. yuphae but were heterozygous and homozygous, respectively, for IIIL-59 (Table 4; Fig 9A). We tentatively consider IIIL-59 an autosomal polymorphism of S. yuphae ' A' , although the possibility that it is X linked (and possibly associated with a separate breeding population) cannot be excluded. Simulium yuphae Cytoform 'B' . Two male larvae (Site 10), initially segregated as morphologically distinct, were chromosomally identical to S. yuphae ' A' , with IIIL-12 and IIIL-13, except both were heterozygous for IIIL-58 (Table 4; Fig 11A) and 1 also was heterozygous for IIIL-57 (Fig 9A), suggesting possible sex linkage of the 2 inversions. We, therefore, provisionally regard them as a separate cytoform-'B' . No other rearrangements were present. If 'B' is consistently defined by a differentiated Y chromosome, then the 2 females from Site 9, about 215 km away, also could belong to 'B' .

Phylogenetic Relationships
Rearrangements previously identified as synapomorphies for the S. tuberosum group, S. tani complex, and Southeast Asian subgroup [24] were included in our phylogenies. Among the 88 chromosomal rearrangements discovered in the S. tuberosum group in Vietnam, 8 had phylogenetic potential (IL-2, IL-19, IIIL-11, IIIL-12, IIIL-13, IIIL-64, IIIL-79, and 100hb1); that is, they were shared between at least 2 taxa. Seven of these rearrangements were uniquely derived (synapomorphic), based on outgroup comparisons; the breakpoints of IL-19 (shared by S. doipuiense ' A' and 'E') could not be determined in the outgroups and, therefore, was not used for phylogenetic inference. The probability that a shared heteroband (e.g., 100hb1) represents common ancestry versus independent origins is not known. However, the likelihood of independently enhancing DNA content of a band is probably greater than independently sharing an inversion with two microscopically identical breakpoints; thus, we consider the phylogenetic value of 100hb1 weak.
Within the S. tani lineage, S. xuandei was the sister species of the northern clade of S. suzukii, based on IIIL-34, whereas S. tani 'B2' and 'M' , lacking both IL-2 and IIIL-5, were in an unresolved trichotomy with all other members of the S. tani complex (Fig 13). Simulium tani 'N' was in an unresolved trichotomy with cytoforms 'E' and 'K' .
The Southeast Asian S. tuberosum subgroup was uniquely defined by fixed inversion ILtuberosum (Fig 14). Within the IIIL-13 clade, one lineage included the IIIL-11 clade consisting of the cytoforms of S. rufibasis, defined by IIIL-8, and the cytoforms of S. doipuiense, whereas another lineage, defined by fixation of IIIL-12, included the S. yuphae cytoforms. IIIL-12 is shown as a polymorphism in the ancestor of the S. doipuiense-rufibasis-yuphae lineage to accommodate its presence as a polymorphism in S. rufibasis 'B' and as a fixed inversion in the  [19] can be obtained by alphabetically ordering the fragments indicated by the letters a-p; cs = cup and saucer marker.

Taxonomic status of cytoforms
We discovered 15 cytoforms among 9 morphoforms of 6 nominal species in the S. tuberosum group in Vietnam. The cytoforms fall into two categories based on the evidence that can be mustered for reproductive isolation: (1) valid (i.e., reproductively isolated) species and (2)  Simulium tani 'B' is precariously defined by a lack of diagnostic chromosomal rearrangements relative to all known members of the S. tani complex [19,24,30]. Two points merit discussion: (1) Our material of Cytoform 'B2' in Vietnam conforms chromosomally to 'B1' in Thailand [19], except IL-2 is absent in Vietnam, compared to its typically high frequency in Thailand where 9 of 16 samples had IL-2 frequencies of 1.00. This discrepancy is reconciled if 'B1' in Thailand actually includes two species, one lacking IL-2 and another fixed or polymorphic for IL-2. One large Thai population (Site 41) that was not in Hardy-Weinberg equilibrium had a dearth of IL-2 heterozygotes [19], supporting the idea of two species within 'B' . (2) Molecular and morphological analyses, including material from our Site 5 in Vietnam and from Site 37 in Thailand (ca. 680 km distant), previously recorded [19] as fixed for IL-2, indicate that Vietnamese 'B2' (= S. tani 'a') and Thai 'B1' are distinct species and that Vietnamese 'B2' is distinct from all other analyzed populations of the S. tuberosum group [13]. Complicating this analysis, however, is the chromosomal evidencethree fixed-inversion differences-for the existence of a second, separate breeding population (S. tani 'N') in the same stream with 'B2' . The molecular analyses [13], however, did not discern two species; therefore, we do not know if the single species recognized molecularly was 'B2' or 'N' (or both), although 'B2' had greater representation (80%) in our chromosomal sample.
Simulium tani 'M' (= S. tani 'b') has molecular and morphological support as a distinct species [13], but only moderate chromosomal support-a high frequency (0.75) of the unique inversion IIIL-47. Simulium xuandei, on the other hand, has strong molecular, morphological [13], and chromosomal support for species status. Yet, molecular analyses of material of S. xuandei collected simultaneously with our chromosomal sample revealed two or three cryptic species [13], whereas our chromosomal sample recovered only a single species. With a larger chromosomal sample, the molecular hypothesis of separate breeding populations could be tested.

Gender
Sex-chromosome classes 1 Female  1  1  0  0  0  0  0   Male  0  0  1  4  1  1  2 1 Assuming IS-29,31+13hb is Y-linked, X 0 and Y 0 = standard sequence, X 1 = IS-27, X 2 = IS-28, and Y 1 = IS-29,31+13hb. The only female at Site 11 and the only male at Site 13 were X 0 X 0 and X 0 Y 0 , respectively; they are not included in the table. 2 1 additional larva of undetermined gender (infected with a microsporidium) was heterozygous for IS-28. doi:10.1371/journal.pone.0163881.t006 Our Malaysian sample of S. brevipar sensu stricto (= ' A') from near the type locality had a fixed banding sequence identical to that of larvae collected with S. yuphae in Thailand (Site 59) by Tangkawanit et al. [19] and recognized by those authors as a probable species distinct from   possibility that they are separate species. However, the alternative possibility that 'B' is conspecific with ' A' cannot be excluded; more than 1900 linear km separate our samples. The Nearctic example of S. congareenarum provides a caveat. Two populations of S. congareenarum more than 1200 km apart originally were proposed as sibling species based on two fixed-inversion differences supported by slight morphological differences; eventual analysis of an intervening population revealed heterozygosity for the two inversions, reflecting polymorphism gradients and, thus, a single species with fixation of alternate sequences at the two geographic sampling extremes [31]. Geographically intermediate collections and larger samples are needed to test reproductive isolation of S. brevipar ' A' and 'B' .
Three unique chromosomal inversions and the novel swollen basal fenestra of the pupal gill of S. congi [18] support its species status. Simulium brevipar 'C' and S. doipuiense 'F' , however, are enigmatic victims of inadequate sample sizes, and little can be said about their status. Nonetheless, the three unique inversions in each cytoform have a low probability of being polymorphisms that happen to be expressed homozygously in single larvae. The microsporidia-infected larva of 'F' might have been a remnant of a larger population of chromosomally similar larvae that already had developed, leaving only parasitized individuals. Parasitized larvae typically persist in populations after unparasitized larvae have pupated [32].
Simulium doipuiense sensu stricto (= ' A') has a consistent fixed banding sequence and undifferentiated sex chromosomes over its known range, including sites within 4 km of the type locality [19]. None of its polymorphisms, however, are shared between populations in Thailand and Vietnam. Chromosomal evidence demonstrates that S. doipuiense ' A' is reproductively isolated from the two members (' A' and 'B') of the S. rufibasis complex; no hybrids have been found in sympatry in Thailand [19] or in our study. The presence of IIIL-64 in one larva of S. rufibasis 'B' and in two larvae of S. doipuiense ' A' from the same site suggests either introgression or retention of an ancestral polymorphism. Molecular analyses fail to distinguish the S. doipuiense and S. rufibasis complexes in Thailand [33,34]. Ecologically, the S. rufibasis complex inhabits higher elevations (1100-2300 m) than does S. doipuiense ' A' (400-1800 m) [19].
Although S. doipuiense 'D' in central Vietnam differs from ' A' by fixation of IIIL-79, our samples of the two cytoforms are separated by about 800 linear km. The scenario represents another example of the difficulty of interpreting the extent of reproductive isolation between distant populations. The case for S. doipuiense 'C' (previously referred to as S. rufibasis [18]), collected about 450 km to the south of 'D' , is similar, but is further confounded by a sample of only one larva. Geographically intermediate collections are needed to test reproductive isolation of ' A' , 'C' , and 'D' . Simulium doipuiense 'E' , collected from the same stream with ' A' , is tentatively regarded as a distinct species, based on an absence of hybrids. Greater confidence in claiming reproductive isolation would come from a larger sample or molecular or morphological corroboration.
Simulium rufibasis 'B' differs from its nearest relative, S. rufibasis sensu stricto (= ' A'), only in its putative sex chromosomes and autosomal polymorphisms. 'B' might be merely an example of sex-chromosome polymorphism, which is common in the Simuliidae [11], but we treat it here as a cytoform in recognition of the asserted role that sex chromosomes play in speciation of Diptera [35], including the Simuliidae [36,37]. To provide a taxonomic anchor, we provisionally regard ' A' as conspecific with chromosomally unstudied topotypical material of S. rufibasis from India. In the absence of molecular and morphological data, an assessment of the taxonomic status of 'B' relative to ' A' is precluded.
Simulium yuphae ' A' in our samples was identified morphologically as S. cavum. However, we found no chromosomal differences between Thai populations of ' A' , including topotypical material of S. yuphae [19], and our material of S. yuphae (morphologically S. cavum) collected about 750-900 km to the east. Morphological differences between S. cavum and S. yuphae ' A' are slight: number of columns of upper-eye facets in males and size of the tubercles on the pupal frons [18]. Either S. cavum is conspecific with S. yuphae ' A' , in which case the morphological differences represent intraspecific variation and cavum becomes a synonym of yuphae, or S. cavum and S. yuphae are homosequential species [38][39][40]. Simulium yuphae 'B' was identified morphologically as a new species. Chromosomally, however, the only evidence for species status, separate from bona fide S. yuphae ' A' , which was collected about 800 km to the south, was a potentially differentiated sex chromosome in the two male larvae in our sample. The hypothesis that IIIL-57 and IIIL-58 are sex (Y?) linked in 'B' requires testing.
Agreement among chromosomal, molecular, and morphological taxonomic divisions is encouraging, but the discrepancies argue for closer scrutiny and an integrated approach [41,42]. For instance, not all cytoforms in the S. tani complex in Thailand can be evaluated for species status based on chromosomal evidence alone [19]. Molecular evidence, however, suggests that Thai cytoforms ' A' , 'C' , and 'G' are merely cytotypes-polymorphic members of a single species in an early stage of differentiation [13]. Discrepancies between cytogenetic and molecular analyses have been found in other Oriental members of the S. tuberosum group. Simulium weji, for example, has low cytogenetic diversity, suggesting a single species, but high molecular genetic diversity that partitions into groups of possible cryptic species [43].

Phylogenetic relationships
Phylogenies based on polytene chromosomes can provide excellent topological agreement with those based on nucleotide sequence data, and can even be richer in information [33,44]. One of the most strongly supported phylogenetic relationships is the split of the S. tani lineage from all other members of the Southeast Asian S. tuberosum group [33,34]. A molecularly inferred phylogeny of the Oriental S. tani complex indicates that taxa cluster according to geography; thus, members of the complex are arranged in four monophyletic groups corresponding to Malaysia, Taiwan, Thailand, and Vietnam [13]. The chromosomally inferred phylogeny for the S. tani complex, however, does not show country fidelity of clades.
As with the limited set of morphological characters available in the Southeast Asian S. tuberosum group [45], a dearth of shared chromosomal characters also limits the extent to which phylogenetic relationships can be inferred. Although certain chromosomal synapomorphies (e.g., IIIL-11, IIIL-13) provide a strong phylogenetic signal, the scarcity of shared rearrangements for taxa in Vietnam, coupled with the challenge of determining if they are derived (i.e., by comparison of the often-scrambled sequences against the sequences in outgroups), can limit their utility. An integrated approach that taps the potentially larger set of molecular characters [13] will be needed for a fully resolved phylogeny.

Chromosomal and taxonomic biodiversity
The chromosomal rearrangements discovered in our Vietnamese samples bring to 180 the number now known for the Asian S. tuberosum group. These 180 rearrangements are distributed among 40 cytologically distinct taxa, 38% of which are known from Vietnam. Intra-and interspecific inversions are disproportionately concentrated (>50%) in the IIIL arm, not only for taxa in Vietnam, but also for the entire S. tuberosum group in the Palearctic and Oriental Regions [19,24,30,46]. The highly labile nature of IIIL suggests that the arm is given to increased fragility, that the retention rate of the breakage products (i.e., inversions) is higher, or both. In contrast, not a single rearrangement is known from the IIIS arm in the Oriental Region, other than the displaced nucleolar organizer, which is a synapomorphy for the entire S. tuberosum group [19]. The Nearctic members of the group express the majority of their interspecific chromosomal differences in IIS rather than IIIL, reflecting an independent evolutionary trajectory sometime after divergence from the ancestor of the S. tuberosum group [22,47].
The discovery of 88 different rearrangements, beyond the 25 characteristic of the basic sequences, among 272 Vietnamese larvae contrasts sharply with the 50 rearrangements found among 3347 Thai larvae [19]. The number of cytoforms discovered in Vietnam (15), however, is roughly the same as that known in Thailand (16). Chromosomal comparisons of the four nominal species shared between Vietnam and Thailand (S. tani, S. doipuiense, S. rufibasis, and S. yuphae) reveal the same trend-a greater number of different rearrangements for each of the four nominal taxa in Vietnam, when corrected for sample size, despite fewer to roughly the same number of sampling sites in about the same number of ecoregions (4 or 5) (sensu [48]). These rearrangements represent 2.0-2.5 times more cytoforms of each of these four taxa in Vietnam, except S. tani, which consists of 2.3 times more cytoforms in Thailand where the number of its analyzed larvae was 25 times greater than in Vietnam. The high chromosomal diversity in Vietnam, thus, is not an artifact of sampling, lending credence to the country's status as a biodiversity hotspot.
The diversity of rearrangements and the taxa they represent in Vietnam overlaps minimally with that in Thailand. Only two cytoforms (S. doipuiense ' A' and S. yuphae ' A') and seven rearrangements (IL-2, IIIL-5, IIIL-8, IIIL-11, IIIL-12, IIIL-13, and IIIL-34), beyond the basic sequences, are shared between Thailand and Vietnam. The minimal congruence might reflect distance, local adaptation, and periods of population isolation resulting from glacial cycles. The ecological diversity among Oriental members of the S. tuberosum group has been suggested as an indication that ecological adaptation has played a role in driving evolution in the group [33,43,49]. The S. tuberosum group in Southeast Asia is found in the mountains, which has promoted population divergence in other black flies, with the intervening lowlands restricting gene flow [50]. The isolation of populations at higher elevations would have been particularly acute during glacial periods when tropical areas were drier and streams flowed only in high mountains [49].
More generally, the S. tuberosum group has 10 nominal species in the Nearctic Region, 23 in the Palearctic, and 32 in the Oriental, with 8 of these in Thailand, 10 in Malaysia, and 14 in Vietnam [21,51], plus 1 additional valid, but unnamed, species in Vietnam revealed in our study. Thailand is 1.5 times larger than Malaysia and Vietnam and has been intensively surveyed for simuliids since 1984 [52]. In contrast, focused simuliid exploration began in Vietnam only in 2014 and has been restricted to a limited portion of the country [15][16][17][18]51]. The only comparable analyses of chromosomal diversity in the group beyond the Oriental Region have been conducted in eastern Canada (Nearctic Region) and Hokkaido, Japan, plus two provinces in South Korea (Palearctic Region). Analysis of 1190 larvae (350 larvae for the IS arm) from eastern Canada revealed five cytoforms and 93 different rearrangements [22], whereas 118 larvae from Japan and Korea revealed three cytoforms and 28 different rearrangements [24]. Correcting for sample size, the number of different rearrangements per larva remains greater for Vietnam: 0.32 versus 0.24 for Japan plus Korea, 0.06 for eastern Canada, and 0.01 for Thailand.
Speciation in the Simuliidae has been associated with chromosomal phenomena, particularly coadaptation of sex chromosomes [37], cooption of individual rearrangements for different roles (e.g., fixation, X linkage) in different lineages, and more rarely, larger genomic restructuring events such as translocations [36]. Of the 15 Vietnamese cytoforms, four or fewer (S. brevipar 'B' , S. doipuiense 'E' , S. rufibasis 'B' , and S. yuphae 'B') have differentiated sex chromosomes, compared with seven of 16 in Thailand [19]. At most, only three pairs of Vietnamese taxa provide examples of a differentially expressed rearrangement: (1) Simulium tani 'B1' and 'B2' are defined on the basis of IL-2, which is polymorphic or fixed in 'B1' and absent in 'B2' , (2) the S. yuphae complex and S. rufibasis 'B' carry IIIL-12, which is fixed in the former and polymorphic in the latter, and (3) S. doipuiense ' A' and 'E' have heteroband 100hb1, the former as an autosomal polymorphism and the latter as an X-linked rearrangement. This situation contrasts with the pattern in the S. tani lineage in Thailand [19] and numerous other groups of simuliids [26,53] in which closely related cytoforms are defined in whole or in part by the same rearrangement operating in as many as five different roles. Thus, diversification of simuliids in Vietnam, although corresponding in part to general patterns of chromosomal restructuring associated with speciation, might include additional chromosomal phenomena or altogether different phenomena, such as those operating at the molecular level.
The discovery of hidden diversity in the S. tuberosum group in Vietnam follows an emerging pattern in the Simuliidae-cryptic diversity is not uniformly distributed across taxa. Rather, certain nominal (morphologically based) "species" in each zoogeographic region have a disproportionately greater degree of cryptic diversity. Diversity-rich taxa, formerly considered single species, include Helodon onychodactylus and S. arcticum in the Nearctic Region [54][55][56], Simulium metallicum in the Neotropical Region [57], Simulium colombaschense in the Palearctic Region [25], and the super-rich Simulium damnosum-the largest species complex of bloodfeeding arthropods in the world-in the Afrotropical Region [58]. Simulium tani and S. doipuiense represent this pattern in the Oriental Region. As additional morphospecies are screened for cryptic biodiversity, attention to this trend should provide insights into the process(es) responsible for uneven cryptic diversification.

Conclusions
Comparative analyses indicate that chromosomal and species diversity in the S. tuberosum group is greatest in the Oriental Region, particularly in Vietnam. Our samples of the 15 cytoforms in Vietnam show a typical right-skewed distribution of relative abundance, with one abundant cytoform (S. doipuiense ' A') and five others represented by one or two individuals. The implication of the high proportion (33%) of rare cytoforms is that increased sampling across space and time would reveal additional taxa. The urgency is increasing to discover the extent of biodiversity in Vietnam before it is too late. The montane forests that provide suitable habitat for the S. tuberosum group and other taxa in Vietnam are under threat from the pressures of a rapidly growing human population [4] that now approaches 100 million. Given the restricted geographical distributions suggested by our findings, some taxa, including those not yet discovered, could be imminently threatened.