Local adaptation towards divergent ecological conditions often results in genetic differentiation and adaptive phenotypic divergence. To illuminate the ecological distinctiveness of the schizothoracine fish, we studied a Gymnocypris species complex consisting of three morphs distributed across four bodies of water (the Yellow River, Lake Qinghai, the Ganzi River and Lake Keluke) in the Northeast Tibetan Plateau. We used a combination of mitochondrial (16S rRNA and Cyt b) and nuclear (RAG-2) genetic sequences to investigate the phylogeography of these morphs based on a sample of 277 specimens. Analysis of gill rakers allowed for mapping of phenotypic trajectories along the phylogeny. The phylogenetic and morphological analyses showed that the three sparsely rakered morphs were present at two extremes of the phylogenetic tree: the Yellow River morphs were located at the basal phylogenetic split, and the Lake Keluke and Ganzi River morphs at the peak, with the densely rakered Lake Qinghai morphs located between these two extremes. Age estimation further indicated that the sparsely rakered morphs constituted the oldest and youngest lineages, whereas the densely rakered morph was assigned to an intermediate-age lineage. These results are most compatible with the process of evolutionary convergence or reversal. Disruptive natural selection due to divergent habitats and dietary preferences is likely the driving force behind the formation of new morphs, and the similarities between their phenotypes may be attributable to the similarities between their forms of niche tracking associated with food acquisition. This study provides the first genetic evidence for the occurrence of convergence or reversal in the schizothoracine fish of the Tibetan Plateau at small temporal scales.
Citation: Zhang R, Peng Z, Li G, Zhang C, Tang Y, Gan X, et al. (2013) Ongoing Speciation in the Tibetan Plateau Gymnocypris Species Complex. PLoS ONE 8(8): e71331. https://doi.org/10.1371/journal.pone.0071331
Editor: Thomas Mailund, Aarhus University, Denmark
Received: April 9, 2013; Accepted: June 28, 2013; Published: August 15, 2013
Copyright: © 2013 Zhang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was partially supported by grants from the Chinese Academy of Sciences (Key Innovation Plan: KSCX2-YW-N-101 and KSCX2-1W-N-004) and from the National Natural Science Foundation of China (30970341, 31172070, 31071903). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The discovery of the ecological and evolutionary forces responsible for population divergence and adaptation has long been a major objective of evolutionary biology –. Local adaptation driven by differing ecological conditions often results in the adaptive phenotypic and genetic divergence of geographically isolated populations , – and may drive the formation of new taxa –. Such a process requires a source of divergent selection, for which environmental differences and niche adaptation are driving mechanisms , , , . Cases of local adaptation that demonstrate adaptive phenotypic divergence have been well documented in a number of taxa , , , –. However, when two independent evolutionary lineages occupy comparable trophic niches, similar phenotypes can be generated through analogous evolutionary responses –. Evolutionary convergence or reversal in a selected morphological character may then occur –. Schizothoracine fish are specialized for high-elevation rivers and exhibit a number of unique adaptations to the Tibetan Plateau , . This large taxon has become an important model for research on speciation in nature and the ecological factors underlying divergence in general –. In this study, we present genetic and morphological evidence of such a genetic divergence in a Gymnocypris species complex of schizothoracine fish. And we demonstrate an evidence of ongoing speciation with a tendency to retain the ancestral phenotype on the complex by convergence or reversal.
This small species complex is endemic to the Yellow River water system and to the Lake Qinghai water system of the Northeast Tibetan Plateau. The complex includes three members , , , : Gymnocypris eckloni eckloni, distributed in the upper reaches of the Yellow River; Gymnocypris przewalskii przewalskii, from Lake Qinghai; and G. p. ganzihonensis, endemic to the Ganzi River (Fig. 1) , . The current speciation hypothesis for these species is that they evolved allopatrically through geographic isolation from an ancestral taxon , . This geographical isolation is thought to have occurred at least twice: once through an early series of well-known geological events ca. 15 Ma that caused the separation of Lake Qinghai from the upper reaches of the Yellow River, producing the species G. przewalskii (Fig. 1) –, and a second, more recent, event caused by the drying of the climate, which separated the Ganzi River from Lake Qinghai and produced the endemic G. p. ganzihonensis (Fig. 1) .
This map shows the locations of the samples in the Northeastern Tibetan Plateau and their corresponding codes, as listed in Table 1.
Local adaptation can result in genetic differentiation and phenotypic divergence, as evidenced by the divergence of gill raker numbers. Gill raker number is associated with food acquisition, and it has been suggested that variations in gill raker number may be at least partially influenced by natural selection , –. In the studied Gymnocypris species complex, different morphs are often separated by differences in gill raker number, and this trait is assumed to be a reliable marker for assessing the systematic relationships between the species , . Interestingly, although G. e. eckloni and G. p. ganzihonensis evolved in different habitats, they exhibit similar gill raker numbers and morphologies: each bears relatively few, short, sparsely spaced rakers. In contrast, G. p. przewalskii is distinctly different, having many long, densely spaced gill rakers , .
This unique model system has proven to be a good candidate for the study of adaptive convergence or reversal, enabling an evaluation of the relative roles of historical identity, phenotypic distinctness and ecological parameters in the adaptive diversification of the sparsely and densely rakered morphs , –. To date, few studies of schizothoracine fish have fully integrated ecological, morphological and life historical information into a genetic framework to address ecological speciation , . In this study, we used both mitochondrial and nuclear gene sequences to investigate the evolutionary history of the Gymnocypris fish species complex and the possible driving forces underlying its speciation. We examined in detail the phenotypic trajectories of these species along a phylogeny, specifically evaluating whether the speciation of the Gymnocypris fish species complex underwent an adaptive convergence or a reversal in gill raker characteristics at small temporal scales.
According to two previous genetic studies of mitochondrial markers, the Yellow River G. e. eckloni, Lake Qinghai G. p. przewalskii and Ganzi River G. p. ganzihonensis form a species complex , . The first study (based on Cyt b) primarily focused on the phylogeny and population structures of G. p. przewalskii and G. e. eckloni, only examining a limited number (14) of G. p. ganzihonensis individuals. The latter study (based on a control region and Cyt b) primarily focused on G. eckloni and contained no G. p. ganzihonensis individuals. In the present study, we amplified three DNA regions (mitochondrial 16S rRNA, mitochondrial Cyt b and nuclear RAG-2) to compare sequence variations among geographic morphs because assessing variations in multiple genes, particularly in both nuclear and mitochondrial genes, generates a more robust inference of genetic structure than assessing a single DNA region , .
Samples were collected from across the entire range of the complex. In addition, given the high degree of diversification found in this species complex, we considered it likely that collecting populations from a previously unstudied adjacent lake would result in the discovery of new DNA lineages. We therefore sampled individuals from Lake Keluke, which is adjacent to Lake Qinghai. A Gymnocypris population has been reported in Lake Keluke  and is currently classified as G. p. przewalskii , although this taxonomic status is not widely accepted. No genetic surveys of this population have been previously conducted due to sampling difficulties; here, we provide the first molecular data for this population.
Materials and Methods
All necessary permits for collection and experimentation were acquired for the described field study from the Agriculture Department of Qinghai Province, China. All samples of fish used in this study followed the guidelines of the regulations of experiments on animals, and was approved by China Zoological Society. Samples were collected using gill nets or cast nets between January 2011 and December 2012. All the specimens were preserved in 95% ethanol for laboratory analyses.
Samples and Laboratory Analyses
A total of 277 individuals were obtained from the Yellow River, Lake Qinghai, the Ganzi River and Lake Keluke (Fig. 1; Table 1). G. e. chilianensis, G. waddelli and G. potanini, three species closely related to G. e. eckloni, were chosen as outgroups. Voucher specimens were deposited at the Northwest Plateau Institute of Biology, in the Chinese Academy of Science in Xining.
Total genomic DNA was extracted from muscle tissue using a phenol/chloroform extraction procedure . A complete sequence of the Cyt b gene and partial sequences of the 16S rRNA and RAG-2 genes were obtained for all the sampled individuals. The primer information is listed in Table 2. The polymerase chain reaction (PCR) mixture contained approximately 100 ng of template DNA, 1 µL of each primer, 5 µL of 10× reaction buffer, 2 µL of dNTPs (each 2.5 mM) and 2.0 U of Taq DNA polymerase, in a total volume of 50 µL. The PCR conditions included an initial denaturation at 94°C for 3 min; 30 cycles of denaturation at 94°C for 1 min, annealing at 58–64°C for 1 min and extension at 72°C for 1 min; and a final extension at 72°C for 5 min. The amplified DNA was fractionated by electrophoresis through 0.8% low-melting agarose gels, recovered from the gels and purified with Gel Extraction Mini kit (Watson Biotechnologies, Shanghai, China). The purified DNA was sequenced on an ABI 3730 capillary sequencer with the Perkin-Elmer BigDye DNA Sequencing Kit according to the manufacturer’s protocol, using the primers previously employed in the PCR (Beijing Tianyi Huiyuan Bioscience and Technology Incorporation, Beijing, China). Newly reported sequences in this paper have been deposited in GenBank under accession numbers KC733894-KC734002, KC76749-KC767659 and KC757127 (see Appendix S1).
Mismatch distributions and Fu’s Fs test  were conducted using Arlequin version 3.5 . The moment estimator of the time to expansion (τ) was computed, and the time of the main expansion in generations (t) was estimated by the equation τ = 2ut –, with u representing the mutation rate per sequence and per generation. The value of u was calculated using u = 2 µk, with µ representing the mutation rate per nucleotide and k representing the number of nucleotides in the analyzed fragment. Finally, the approximate time of expansion in years was calculated by multiplying t by the generation time (4 years; ) of the schizothoracine fish. We used the previously established average substitution rate of 1.69% per million years , which was calibrated for the mitochondrial DNA of the schizothoracine fish.
Analyses of Genetic Structure and Gill Raker Numbers
The nucleotide diversity (π) indices  and pairwise genetic differentiation (FST) values ,  were calculated using Arlequin version 3.5 . The overall mean divergence among haplotypes and the net divergence between lineages were calculated using MEGA3 software , and standard errors were estimated by bootstrapping with 10,000 replicates. To determine the associations between gill raker number and genetic structure, we counted the outer rakers and inner rakers of the first gill arch for all 277 sampled individuals, a decisions based on the morphological characteristics of the different morphs , .
We sequenced the Cyt b, 16S rRNA and RAG-2 genes of 280 individuals, including the three outgroup specimens. The alignment performed for the ingroup individuals revealed 43 variable sites in the Cyt b sequences (1140 base pairs [bp]), 25 of which were parsimony informative. The 16S rRNA gene was 1,118 bp in length and included 22 variable sites, 14 of which were parsimony informative. The RAG-2 sequence was 894 bp in length, with 28 variable sites, 14 of which were parsimony informative. No stop codons, insertions or deletions were found in any of the sequences.
Only 43, 22 and 34 unique haplotypes were defined for the Cyt b, 16S rRNA and RAG-2 genes, respectively; therefore, we combined the sequences from the three DNA regions to assess variation. Combined mitochondrial DNA regions have proven useful for elucidating both inter- and intraspecific relationships among the schizothoracine fish , , . A partition homogeneity test indicated the absence of significant incongruence among the Cyt b, 16S rRNA and RAG-2 genes (P>0.05). This apparent phylogenetic congruence justified the combination of the three partial sequences into a single, 3152-bp fragment for the phylogenetic analysis. A total of 113 haplotypes were defined, including three from the outgroup individuals (see Appendix S1).
Haplotype Distribution and Genetic Diversity
No haplotypes were shared between the three morphs (G. p. przewalskii, G. p. ganzihonensis and the Yellow River G. e. eckloni). Each haplotype was present in only one of the river systems (Fig. 2; Appendix S1).
A) The neighbor-joining tree (GTR+I+G model) obtained by combining the Cyt b, 16S rRNA and RAG-2 genes (3,152 bp) of all haplotypes defined in the present study. The geographic origins of the haplotypes are illustrated by the color codes used in Fig. 1. The numbers on the branches correspond to a bootstrap support of ≥60% obtained in the BI, NJ and MP analyses. B) The median-joining network based on the combined sequence data for all the haplotypes assessed in the present study. The haplotype numbers correspond to those in the Appendix S1. The circle sizes represent the approximate numbers of individuals, and the scale is provided in the lower right corner. The black dots indicate the nucleotide substitutions inferred for that branch. The numbers of mutational steps joining the clades are indicated along the connecting branches. Lineage B was linked to lineage C via haplotype 74 by the smallest number of mutations (five mutations; data not shown). The geographical origins of the haplotypes are illustrated by the same colors used in Fig. 1.
The nucleotide diversity was highest in the Yellow River G. e. eckloni (π = 0.0022±0.0011), lowest in the Ganzi River G. p. ganzihonensis and the Lake Keluke G. p. przewalskii (π = 0.0003±0.0001 and 0.0002±0.0001, respectively) and intermediate in the Lake Qinghai G. p. przewalskii (π = 0.0012±0.0007). The estimated maximum divergences between the haplotypes were 0.26% for the Yellow River G. e. eckloni, 0.19% for the Lake Qinghai G. p. przewalskii, 0.07% for the Lake Keluke lineage and 0.06% for the Ganzi River lineage.
All three phylogenetic methods (NJ, MP and BI) used for the 113 haplotypes obtained from the combined data resulted in trees with similar topologies. We found evidence of three distinct lineages (lineages A, B and C; Fig. 2A) rather than two taxonomic species (G. przewalskii and G. eckloni) or three subspecies (G. p. przewalskii, G. p. ganzihonensis and G. e. eckloni). A clear branching order for the Yellow River G. e. eckloni individuals was found in each of the three lineages, whereas the branching of G. p. przewalskii and G. p. ganzihonensis occurred only in lineage C (Figs. 2A, 2B). The divisions of these lineages were well supported (81–100%) and showed strong geographical associations. Lineages A and B both corresponded to the Yellow River G. e. eckloni, whereas the haplotypes of lineage C were widely distributed in both G. p. przewalskii from Lake Qinghai and Lake Keluke and G. p. ganzihonensis from the Ganzi River, as well as in G. e. eckloni from the Yellow River. The mean sequence divergence between lineage A and the other lineages was 0.66% ±0.09, and the net divergence was 0.37% ±0.08; the mean sequence divergence between lineages B and C was 0.38% ±0.07, with a net divergence of 0.19% ±0.07.
The haplotype network for lineage C exhibited seven clusters of haplogroups (haplogroups 28, 29, 39, 47, 58, 66 and 74; Fig. 2B). Although each pair of haplogroups was separated by no more than three mutations, each haplogroup was clearly arrayed in a star-like structure around a central common haplotype (Fig. 2B). These seven haplogroups corresponded to four geographically defined bodies of water. Haplogroups 28, 29, 39 and 47 contained only samples from the Lake Qinghai G. p. przewalskii. Haplogroup 66 was composed of only the Lake Keluke G. p. przewalskii, whereas haplogroup 58 was restricted to the G. p. ganzihonensis endemic to the Ganzi River. The remaining haplogroup, 74, comprised only samples of G. e. eckloni from the Yellow River. Lineage B was linked to lineage C via haplotype 74 by the smallest number of mutations (five mutations; data not shown).
Unimodal mismatch distributions and significant negative values for Fu’s Fs were observed for Gymnocypris individuals from the four bodies of water (Fig. 3). These results suggest the occurrence of expansion events in the demographic histories of these populations. The earliest recent expansions were estimated to have occurred approximately ca. 0.017–0.21 Ma, as shown in Fig. 3.
Associations between Gill Raker Numbers and Genetic Structure
The Gymnocypris fish of all four populations were characterized by a bimodal frequency distribution of gill raker number (Fig. 4), which was composed of three sparsely rakered morphs and one densely rakered morph (Fig. 4). The gill raker numbers were similar among the sparsely rakered Gymnocypris morphs from the Yellow River, Lake Keluke and the Ganzi River (with average outer/inner raker numbers of 15.31/21.47, 17.91/30.71 and 17.59/26.3, respectively) and among the densely rakered morphs from Lake Qinghai (with average outer/inner raker numbers of 28.09/44.57) (Table 3). The sparsely and densely rakered morphs showed highly significant differences in these measures (t-test: P<0.01).
The densely rakered morph in Lake Qinghai were compared to the sparsely rakered Yellow River G. e. eckloni, Ganzi River G. p. ganzihonensis and Lake Keluke G. p. przewalskii, respectively. Sample size (N) is given.
Pairwise comparisons of genetic differentiation (FST) based on the combined data revealed strong differences among the Gymnocypris fish from each of the four environments (Table 4). A high level of differentiation was detected between the three sparsely rakered morphs (the Yellow River, Lake Keluke and Ganzi River morphs), whereas the densely rakered Lake Qinghai morph exhibited only an intermediate level of differentiation from any of the other morphs (Table 4). The FST test results were in agreement with the results of the phylogenetic analyses.
Phylogeographic Processes and Age Estimation
Lake Qinghai and the Yellow River water system together represent the clearest geologically based barrier to fish movement, and the key aspects of this area’s geological history are well documented. Lake Qinghai is the largest brackish water lake in the Tibetan Plateau. This lake originated due to tectonic activity, and its complete closure has been dated to approximately 0.15 Ma , , . Originally, the lake was connected to the ancient Yellow River , , , . The “Gonghe Movement” event of the Tibetan Plateau in the Late Pleistocene led to the uplift of the present barrier, and the single body of water was separated into its current configuration of two large water systems (Fig. 1) , –. Lake Qinghai was completely closed, and the Yellow River began to cut through the Gonghenanshan Mountains and captured further ancient limnetic basins, reaching upward to its present headwaters. This event offers sufficient support for a Late Pleistocene origin of G. p. przewalskii from the Yellow River G. eckloni . More recently, the Ganzi River was separated from Lake Qinghai by a water level drop, which led to the transformation of the lower reaches of the Ganzi River into a subterranean river that now flows into Lake Qinghai (Fig. 1). As a result, G. p. przewalskii evolved into the endemic G. p. ganzihonensis within an isolated environment . There are no detailed data on the origin of the fish in Lake Keluke, but the stratigraphic evidence indicates that the lake's two largest tributaries and those of Lake Qinghai were adjacent and that both derived from the northern Qilian Mountains, which exhibit a recent stratigraphic uplift that may have promoted the isolation of the two lakes .
We discovered two lineages (A and B) of the Yellow River G. e. eckloni at the basal phylogenetic split of this species complex and one lineage (C) that was shared among all of the Gymnocypris individuals from Lake Qinghai, Lake Keluke and the Ganzi River (Fig. 2a, b). This branching order implies that the genetic differentiation of the Yellow River G. e. eckloni may have existed long before Lake Qinghai separated from the Yellow River. These observations suggest that three lineages of the Yellow River G. e. eckloni may have originated from an ancestral population that fragmented into isolated populations and that the ancient lake basins in the limits of the upper Yellow River might have acted as evolutionary reservoirs for the fish.
Our haplotype network approach enabled a fine-grained reconstruction of the evolutionary history of the young lineage C, showing through a network tree approach that the Yellow River G. e. eckloni should be considered ancestral to lineage C. First, the central predominant haplotype (haplotype 74; Fig. 2B) was composed entirely of Yellow River G. e. eckloni. Second, haplotype 74 was linked to the older lineage B through the smallest number of mutations (five mutations, data not shown). Third, at least four lineages derived from haplotype 74 have seeded Lake Qinghai to form the G. p. przewalskii population. It is apparent that the Yellow River G. e. eckloni was crucial to the evolution of the Gymnocypris fish in this complex.
The discovery of two well-defined groups endemic to the Ganzi River and Lake Keluke, respectively, was surprising. This result demonstrates the usefulness of sampling a crucial population from Lake Keluke and of utilizing combined sequence data, particularly those from both nuclear and mitochondrial DNA. The Gymnocypris fish from the Ganzi River and Lake Keluke formed two monophyletic groups (haplogroups 58 and 66, respectively) that were sisters to two G. p. przewalskii lineages (haplogroups 47 and 28, respectively). It is also apparent that the Lake Qinghai G. p. przewalskii played a critical role in the evolution of the Gymnocypris fish of the Ganzi River and Lake Keluke. A clear pattern of decreasing genetic diversity was detected among the four bodies of water, and the age estimation suggested that the most recent demographic expansion occurred in the two newer morphs (Fig. 3). Together with the geological evidence, these data indicate that a sequential process of adaptive divergence occurred in this Gymnocypris species complex.
Ecological Speciation with a Tendency toward Convergence or Reversal in Gill Raker Traits
The three phylogenetic lineages (A, B and C) were not associated with operational taxa. This lack of congruence between the genetic and phenotypic patterns calls into question the validity of gill raker number as a marker of taxonomic separation. However, gill raker number may still be valuable for identifying ecological distinctiveness and evolutionary forces , .
Gymnocypris fish (three sparsely rakered morphs and one densely rakered morph) in the four environments were characterized by a bimodal frequency distribution of gill raker number with very few overlapping individuals. The densely rakered Lake Qinghai G. p. przewalskii fish were characterized mainly by their many long gill rakers, in contrast to the sparsely rakered Yellow River G. e. eckloni, Ganzi River G. p. ganzihonensis and Lake Keluke G. p. przewalskii, which had fewer, shorter rakers (Table 3). However, this study demonstrated that the three sparsely rakered lineages were located at two extremes of the phylogenetic tree: the Yellow River G. e. eckloni was located at the basal phylogenetic split, and the Lake Keluke G. p. przewalskii and Ganzi River G. p. ganzihonensis were at the peak, with the densely rakered Lake Qinghai G. p. przewalskii clearly located between the two extremes (Fig. 2AB). These results suggest that the similar morphologies of the three sparsely rakered morphs may have originated through a complex mechanism, such as introgressive hybridization, retention of ancestral polymorphism or parallel or convergent evolution triggered by directional selection toward similar morphology . The recently derived morphs with sparse rakers formed two monophyletic groups (haplogroups 58 and 66), and each was fully geographically separated from the other, making an introgressive hybridization scenario unlikely. The hypothesis of ancestral polymorphism is more difficult to exclude, but morphological data have demonstrated that G. p. ganzihonensis has a distinct lower jaw horn, which is absent in G. e. eckloni , . This observation may indicate that the sparsely rakered trait in these two species originated through independent evolution along different trajectories. This leaves convergence or reversal in the gill raker number trait as the most likely mechanism. Support for this hypothesis includes the detection of a high level of differentiation (P<0.01) among the three sparsely rakered morphs (Table 4). The phylogeographic analyses and age estimation further indicated that the sparsely rakered morphs constituted the oldest and youngest lineages, with the densely rakered morph assigned to an intermediate type.
The evolutionary trends in gill raker number have also been associated with potentially causal ecological factors . Typically, fish with densely rakers prefer a plankton diet, whereas sparse rakered morphs primarily consume zoobenthos , –, . The character of trophically relevant structures is tightly linked to the ability of schizothoracine fish to process alternative food types . In this Gymnocypris species complex, gill raker number is also associated with food acquisition, and the observed feeding preferences are consistent with ecological specialization (Table 3) , . The Lake Keluke G. p. przewalskii and the Ganzi River G. p. ganzihonensis have been found to feed primarily on benthic organisms, such as mud-dwelling gammarid and chironomid larvae, and occasionally on algae attached to stones . The Lake Qinghai G. p. przewalskii have been reported to prefer the consumption of plankton, such as copepoda and bacillariophyta , . In contrast, the Yellow River G. e. eckloni have a wider niche corresponding to a variety of food items, with a preference for zoobenthos and a tolerance for plankton and algae as alternative food sources (Table 3) , , which is consistent with the usual location of the Yellow River G. e. eckloni in deeper waters. Geographically localized natural selection can drive local adaptation and often results in the adaptive phenotypic and genetic divergence of populations , –. Therefore, for the Gymnocypris fish in this study, disruptive natural selection due to divergent habitats and dietary preferences is the most likely driving force underlying the observed divergence.
Although these variable environments may have promoted rapid divergence, changes in these environments may lead to rapid convergence or the reversal of divergence . Our data suggest that evolution has sometimes favored the same trait combinations and food type changes may have played an important role in the process. Lake Qinghai contains relatively high plankton densities, which may have promoted the evolution of the densely rakered G. p. przewalskii instead of the sparsely rakered Yellow River G. e. eckloni. In contrast, the Ganzi River and Lake Keluke both provide abundant zoobenthos and algae. These trophic specializations resulted in morphological innovation through their effects on trophically-related traits. The trophic niches of the Ganzi River and Lake Keluke were comparable to that of the Yellow River G. e. eckloni, which may have encouraged the rapid adaptation of gill raker traits toward the ancestral state. Interestingly, the highest genetic differentiation was observed between the two youngest lineages, the Lake Keluke and Ganzi River lineages (Fst = 0.6754; Table 4), indicating an earlier divergence from different ancestral lineages. However, the similarities between their phenotypes might also suggest analogous evolutionary responses for local adaptation associated with ecological specialization.
Unexpectedly, average gill raker numbers among the Lake Keluke Gymnocypris fish were substantially lower (17.91 for the outer rakers and 30.71 for the inner rakers of the first gill) than those previously described (31.29 and 48.71) by Wu and Wu . This earlier study, which was based on seven individuals, did not demonstrate clear differences in external characters and gill rakers between the Lake Keluke population and that of Lake Qinghai; therefore, the Lake Keluke fish were included in the G. p. przewalskii subspecies . In the present study, we observed gill raker counts completely different from those of G. p. przewalskii, and the molecular data supported the separation of the Lake Keluke Gymnocypris fish into a monophyletic sister group to the Lake Qinghai G. p. przewalskii. A comprehensive morphological and ecological analysis based on a more extensive sample of Lake Keluke Gymnocypris fish will be necessary to determine the appropriate taxonomic status of this population.
Lineage C contained all four morphs, suggesting that the greatest morphological diversity was associated with recent evolutionary events. Indeed, the rate of divergence was rapid among the morphs as early as 0.017 Ma. This observation implies that these species have a recent origin of divergence and an extreme sensitivity to environmental change. They also exhibit a great ability to adapt to different habitats.
The high genetic divergence of the three Yellow River G. e. eckloni lineages reflects their ancient history, but no consistent morphological differentiation within this population has been recorded thus far , . This result suggests that the similarity of traits among the G. e. eckloni lineages is not due to insufficient time for divergence. We hypothesize that stable selection pressures and homogeneous conditions within the water system explain this observed pattern; further research is necessary to confirm this hypothesis.
The results of this study suggested a rapid evolutionary radiation with a tendency toward the ancestral phenotype in these Gymnocypris fish. Our study provides the first genetic evidence for the occurrence of convergence or reversal in the schizothoracine fish of the Tibetan Plateau at small temporal scales and highlights the need for a detailed examination of additional species’ phenotypic trajectories along a phylogeny to test the generality of the observed patterns.
Conceived and designed the experiments: RYZ KZ. Performed the experiments: RYZ GGL CFZ YTT. Analyzed the data: RYZ ZGP XNG SPH KZ. Contributed reagents/materials/analysis tools: RYZ GGL ZGP KZ. Wrote the paper: RYZ KZ.
- 1. Schluter D (2000) The ecology of adaptive radiation. Oxford: Oxford University Press.
- 2. Dieckmann U, Doebeli M, Metz JAJ, Tautz D (2004) Adaptive speciation. Cambridge: Cambridge University Press.
- 3. Østbye K, Amundsen PA, Bernatchez L, Klemetsen A, Knudsen R, et al. (2006) Parallel evolution of ecomorphological traits in the European whitefish Coregonus lavaretus (L.) species complex during postglacial times. Molecular Ecology 15: 3983–4001.
- 4. Rowe KC, Aplin KP, Baverstock PR, Moritz C (2011) Recent and rapid speciation with limited morphological disparity in the genus Rattus. Systematic Biology 60: 188–203.
- 5. Mayr E (1947) Ecological factors in speciation. Evolution 1: 263–288.
- 6. Endler JA (1986) Natural selection in the wild. Princeton, New Jersey: Princeton University Press.
- 7. Collin H, Fumagalli L (2011) Evidence for morphological and adaptive genetic divergence between lake and stream habitats in European minnows (Phoxinus phoxinus, Cyprinidae). Molecular Ecology 20: 4490–4502.
- 8. Rice WR (1987) Speciation via habitat specialization: the evolution of reproductive isolation as a correlated character. Evolutionary Ecology 1: 301–314.
- 9. Kawecki TJ, Ebert D (2004) Conceptual issues in local adaptation. Ecology Letters 7: 1225–1241.
- 10. Kirkpatrick M, Barton N (2006) Chromosome inversions, local adaptation and speciation. Genetics 173: 419–434.
- 11. Schluter D (1994) Experimental evidence that competition promotes divergence in adaptive radiation. Science 266: 798–801.
- 12. Bolnick DI (2004) Can intraspecific competition drive disruptive selection? An experimental test in natural populations of sticklebacks. Evolution 58: 608–618.
- 13. Saether SA, Fiske P, Kalas JA, Kuresoo A, Luigujoe L, et al. (2007) Inferring local adaptation from QST-FST comparisons: neutral genetic and quantitative trait variation in European populations of great snipe. Journal of Evolutionary Biology 20: 1563–1576.
- 14. Leinonen PH, Sandring S, Quilot B, Clauss MJ, Mitchell-Olds T, et al. (2009) Local adaptation in European populations of Arabidopsis Lyrata (Brassicaceae). American Journal of Botany 96: 1129–1137.
- 15. Antoniazza S, Burri R, Fumagalli L, Goudet J, Roulin A (2010) Local adaptation maintains clinal variation in melanin-based coloration of European barn owls (Tyto alba). Evolution 64: 1944–1954.
- 16. Losos JB (1992) The evolution of convergent structure in Caribbean Anolis communities. Systematic Biology 41: 403–420.
- 17. Losos JB, Jackman TR, Larson A, de Queiroz K, Rodriguez-Schettino L (1998) Contingency and determinism in replicated adaptive radiations of island lizards. Science 279: 2115–2118.
- 18. Rüber L, Verheyen E, Meyer A (1999) Replicated evolution of trophic specializations in an endemic cichlid fish lineage from Lake Tanganyika. Proc Natl Acad Sci U S A 96: 10230–10235.
- 19. Rüber L, Adams DC (2001) Evolutionary convergence of body shape and trophic morphology in cichlids from Lake Tanganyika. Journal of Evolutionary Biology 14: 325–332.
- 20. Grant T, Humphrey EC, Myers CW (1997) The median lingual process of frogs: A bizarre character of Old World ranoids discovered in South American dendrobatids. American Museum Novitates 3212: 1–40.
- 21. Vences M, Kosuch J, Boistel R, Haddad CFB, La Marca E, et al. (2003) Convergent evolution of aposematic coloration in Neotropical poison frogs: a molecular phylogenetic perspective. Organisms Diversity & Evolution 3: 215–226.
- 22. Taylor EB, Boughman JW, Groenenboom M, Sniatynski M, Schluter D, et al. (2006) Speciation in reverse: morphological and genetic evidence of the collapse of a three-spined stickleback (Gasterosteus aculeatus) species pair. Molecular Ecology 15: 343–355.
- 23. Syme AE, Oakley TH (2012) Dispersal between shallow and abyssal seas and evolutionary loss and regain of compound eyes in cylindroleberidid ostracods: conflicting conclusions from different comparative methods. Systematic Biology 61: 314–336.
- 24. Cao WX, Chen YF, Wu YF, Zhu SQ (1981) Origin and evolution of schizothoracine fishes in relation to the upheaval of the Xizang Plateau. In: Science TEotCAo, editor. Studies on the period, amplitude and type of the uplift of the Qinghai-Xizang Plateau. Beijing: Science Press. 118–130.
- 25. He DK, Chen YF (2006) Biogeography and molecular phylogeny of the genus Schizothorax (Teleostei : Cyprinidae) in China inferred from cytochrome b sequences. Journal of Biogeography 33: 1448–1460.
- 26. Zhao K, Li JB, Yang GS, Duan ZY, He SP, et al. (2005) Molecular phylogenetics of Gymnocypris (Teleostei : Cyprinidae) in Lake Qinghai and adjacent drainages. Chinese Science Bulletin 50: 1325–1333.
- 27. Zhao K, Duan ZY, Peng ZG, Guo SC, Li JB, et al. (2009) The youngest split in sympatric schizothoracine fish (Cyprinidae) is shaped by ecological adaptations in a Tibetan Plateau glacier lake. Molecular Ecology 18: 3616–3628.
- 28. Zhao K, Duan ZY, Peng ZG, Gan XN, Zhang RY, et al. (2011) Phylogeography of the endemic Gymnocypris chilianensis (Cyprinidae): Sequential westward colonization followed by allopatric evolution in response to cyclical Pleistocene glaciations on the Tibetan Plateau. Molecular Phylogenetics and Evolution 59: 303–310.
- 29. Qi DL, Chao Y, Guo SC, Zhao LY, Li TP, et al. (2012) Convergent, parallel and correlated evolution of trophic morphologies in the subfamily schizothoracinae from the Qinghai-Tibetan Plateau. PLoS One 7: e34070.
- 30. Kai Z, Ziyuan D, Gongshe Y, Zuogang P, Shunping H, et al. (2007) Origin of Gymnocypris przewalskii and phylogenetic history of Gymnocypris eckloni (Teleostei: Cyprinidae). Progress in Natural Science 17: 520–528.
- 31. Duan ZY, Zhao K, Peng ZG, Li JB, Diogo R, et al. (2009) Comparative phylogeography of the Yellow River schizothoracine fishes (Cyprinidae): Vicariance, expansion, and recent coalescence in response to the Quaternary environmental upheaval in the Tibetan Plateau. Molecular Phylogenetics and Evolution 53: 1025–1031.
- 32. Wu Y, Wu C (1992) The fishes of the Qinghai-Xizang plateau. Chendu, China: Sichuan Publishing House of Science & Technology.
- 33. Chen YF, Cao WX (2000) Schizothoracinae. In: Yue PQ, editor. Fauna Sinica, Osteichthyes, Cypriniformes II Beijing: Science Press. 273–335.
- 34. Zhu SQ, Wu YF (1975) Studies on the ichthyofauna of Lake Qinghai. In: Qinghai Institute of Biology, editors. Ichthyofauna of Lake Qinghai and biology of Gymnoypris przewalskii (Cyprinidae). Beijing: Science Press. 9–26.
- 35. Chen ZM. A preliminary analysis of paleo-limnology and its environment at the upper reaches of the Huanghe River; 1988; Beijing, China. Science Press. 244–253.
- 36. Zhang Z, Yu Q, Zhang K, Gu Y, Xing S (2003) Geomorphological evolution of Quaternary River from upper Yellow River and geomorphological evolution investigation for 1: 250000 scale geological mapping in Qinghai–Tibet plateau. Earth Science-Journal of China University of Geosciences 28: 621–633.
- 37. Bergstrand E (1982) The diet of four sympatric whitefish species in lake Parkijaure. Report-Institute of Freshwater Research, Drottningholm 60: 5–14.
- 38. Amundsen P-A, Knudsen R, Klemetsen A, Kristoffersen R (2004) Resource competition and interactive segregation between sympatric whitefish morphs. Annales Zoologici Fennici 41: 301–307.
- 39. Amundsen P-A, Bohn T, Vaga GH (2004) Gill raker morphology and feeding ecology of two sympatric morphs of European whitefish (Coregonus lavaretus). Annales Zoologici Fennici 41: 291–300.
- 40. Østbye K, Bernatchez L, Naesje TF, Himberg KJM, Hindar K (2005) Evolutionary history of the European whitefish Coregonus lavaretus (L.) species complex as inferred from mtDNA phylogeography and gill-raker numbers. Molecular Ecology 14: 4371–4387.
- 41. Edwards SV, Beerli P (2000) Perspective: Gene divergence, population divergence, and the variance in coalescence time in phylogeographic studies. Evolution 54: 1839–1854.
- 42. Baird HP, Miller KJ, Stark JS (2011) Evidence of hidden biodiversity, ongoing speciation and diverse patterns of genetic structure in giant Antarctic amphipods. Molecular Ecology 20: 3439–3454.
- 43. Sanbrook J, Fritsch E, Maniatis T (1989) Molecular cloning: a laboratory manual. New York: Cold Spring Harbor Laboratory Press.
- 44. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 25: 4876–4882.
- 45. Swofford DL (2003) PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods), Version 4. Sunderland, Massachusetts Sinauer Associates.
- 46. Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574.
- 47. Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817–818.
- 48. Bandelt HJ, Forster P, Rohl A (1999) Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16: 37–48.
- 49. Fu YX (1997) Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147: 915–925.
- 50. Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10: 564–567.
- 51. Nei M, Tajima F (1981) DNA polymorphism detectable by restriction endonucleases. Genetics 97: 145–163.
- 52. Rogers AR, Harpending H (1992) Population growth makes waves in the distribution of pairwise genetic differences. Molecular Biology and Evolution 9: 552–569.
- 53. Huang S, He SP, Peng ZG, Zhao K, Zhao EM (2007) Molecular phylogeography of endangered sharp-snouted pitviper (Deinagkistrodon acutus; Reptilia, Viperidae) in Mainland China. Molecular Phylogenetics and Evolution 44: 942–952.
- 54. Jin YT, Brown RP, Liu NF (2008) Cladogenesis and phylogeography of the lizard Phrynocephalus vlangalii (Agamidae) on the Tibetan plateau. Molecular Ecology 17: 3033–3034.
- 55. Tsao W, Wu X (1962) An investigation of the fish biology and fishery problems in Ganze-Apa region of western Szechwan Province. Acta Hydrobiologica Sinica 2: 79–111.
- 56. Nei M (1987) Molecular evolutionary genetics. New York: Columbia University Press.
- 57. Reynolds J, Weir BS, Cockerham CC (1983) Estimation of the coancestry coefficient: basis for a short-term genetic distance. Genetics 105: 767–779.
- 58. Slatkin M (1995) A measure of population subdivision based on microsatellite allele frequencies. Genetics 139: 457–462.
- 59. Kumar S, Tamura K, Nei M (2004) MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5: 150–163.
- 60. Li J, Fang X, Ma H, Zhu J, Pan B, et al. (1996) Geomorphological and environmental evolution in the upper reaches of the Yellow River during the late Cenozoic. Science in China Series D(Earth Sciences) 39: 380–390.
- 61. Li JJ, Fang XM, Pan BT, Zhao ZJ, Song YG (2001) Late Cenozoci intensive uplift of Qinghai-Xizang Plateau and its impacts on environments in surrounding area. Quaternary Sciences 21: 381–391.
- 62. Pan BT (1994) A study on the geomorphic evolution and development of the upper reaches of Yellow River in Guide Basin. Arid Land Geography 17: 43–50.
- 63. Pan BT, Wu GJ, Wang YX, Liu ZG, Guan QY (2001) Age and genesis of the Shagou River terraces in eastern Qilian Mountains. Chinese Science Bulletin 46: 509–513.
- 64. Kaukoranta M KM-L, Koskiniemi J, Pennanen J, Tammi J (2000) Atlas of finnish fishes, English summary. Lamprey, brook lamprey, Atlantic salmon, brown trout, Arctic charr, whitefish, vendace, grayling, asp, vimba, spined loach and bullhead - the distribution and status of stocks. Helsinki: Finnish Game and Fisheries Research Institute. 40 p.
- 65. Chiari Y, Vences M, Vieites DR, Rabemananjara F, Bora P, et al. (2004) New evidence for parallel evolution of colour patterns in Malagasy poison frogs (Mantella). Molecular Ecology 13: 3763–3774.
- 66. Kahilainen K, Østbye K (2006) Morphological differentiation and resource polymorphism in three sympatric whitefish Coregonus lavaretus (L.) forms in a subarctic lake. Journal of Fish Biology 68: 63–79.
- 67. Wang JL. Studies on the diet of Gymnocypris przewalakii przewalakii. In: Qinghai Institute of Biology, editors. Ichthyofauna of Lake Qinghai and biology of Gymnoypris przewalskii (Cyprinidae); 1975. Science Press. 27–36.
- 68. Zhao L, Wang J (1990) Environment of Keluke Lake and introduction of valuable fishes. Journal of Fisheries of China 14: 286–296.
- 69. Xiao WH, Zhang YP, Liu HZ (2001) Molecular systematics of Xenocyprinae (Teleostei : Cyprinidae): taxonomy, biogeography, and coevolution of a special group restricted in east Asia. Molecular Phylogenetics and Evolution 18: 163–173.
- 70. Li JB, Wang XZ, Kong XH, Zhao K, He SP, et al. (2008) Variation patterns of the mitochondrial 16S rRNA gene with secondary structure constraints and their application to phylogeny of cyprinine fishes (Teleostei : Cypriniformes). Molecular Phylogenetics and Evolution 47: 472–487.
- 71. Lovejoy NR, Collette BB (2001) Phylogenetic relationships of new world needlefishes (Teleostei : Belonidae) and the biogeography of transitions between marine and freshwater habitats. Copeia 2001: 324–338.
- 72. Qin GX, He GC, Li JX, Xu SC (2001) Study on the biology of Gymnocypris przewalskii przewalskii Kessler. Heilongjiang Journal of Animal Science and Veterinary Medicine 12: 7–8.