Introduction

The Mahseer, an iconic group of fish in the Cyprinidae family, is characterized by large scales and a large body was once considered a single species (Day 1876), is now distributed under three genera Tor, Neolissochilus, and Naziritor (Ogale 2002; Sati et al. 2013; Eschmeyer et al. 2017; Froese and Pauly 2018). However, members only under the genus Tor are considered true mahseers (Desai 2003; Nguyen et al. 2008) and there are 16 valid species, all of which exhibit highly potamodromous behaviors, with upstream spawning migration (Nautiyal et al. 2001, 2008; Shrestha 1997). Eight out of 16 species are reported in India (Pinder et al. 2019). The “Deccan Mahseer,” Tor khudree (Sykes 1839) (Order: Cypriniformes, Family: Cyprinidae) is one of the charismatic mahseer species distributed in Peninsular India (Menon 1992; Jayaram 1994, 2005; Basavaraja 2011; Raghavan et al. 2011). This mahseer is an esthetic food and game fish endemic to the Southwestern part of India (Day 1889). They are commonly called “King of Indian Freshwater systems” (Sati et al. 2013; Sarkar et al. 2015). The slow-growing fish with a long life span and high nutritional value (Day 1876; Froese and Pauly 2010; Sarkar et al. 2015) is overexploited for food by the forest-dwelling communities in central and western peninsular regions of India (Raghavan et al. 2008; Basavaraja 2011). Due to its close resemblance with Tor malabaricus from Peninsular India, they were considered synonymous (Menon 1992, 1999; Pinder et al. 2019). Now malabaricus is considered a valid species with the present name (Silas et al. 2005; Raghavan and Ali 2013). The similarity in the morphology of both species has also led the breeders to misidentify the species. Moreover, overfishing due to food and sports are responsible for the decrease in khudree population. Habitat destruction due to the building of dams has also been the cause of concern for the declining population of the species. Some of the minor and major dams and reservoirs of the important rivers of Deccan India include Tunga Anicut Dam, Tunga, Tungabhadra reservoir, Tungabhadra dam, Harangi dam/reservoir, Cauvery and Bhoothathankettu dam, and Periyar. The highest number (~ 100) of dams present is known for River Cauvery. Due to its declining status, it has been recorded as an endangered species on the Red List of the International Union for Conservation of Nature (Pinder and Raghavan 2013). Their presence indicates the health of the rivers as a little change in the environment would lead to a decline in their number (Basavaraja 2011) and decline in number of natural populations might lead to eradication of unique genotypes that are not found anywhere else (Parmaksiz 2019). Knowledge of the genetics of the species is lacking and research is limited to taxonomy and phylogeny of the species (Nguyen et al. 2008). Information on the stock structure of the species is also very essential before designing an effective management program, conservation strategy, standardized breeding, or selective breeding program. Economic fish species provide both a source of food and an opportunity of employment for people (Parmaksiz 2023). Therefore, the purpose of this study was to determine the genetic variation and population structure of the endangered mahseer across the Deccan riverine systems of India. The effective stock characterization is possible now due to the usage of Molecular markers as they have been utilized in various studies to determine effective population size, population demography, and gene flow (Habib et al. 2012) in fish populations. Maternal inheritance and absence of recombination are a few of the important properties of mitochondrial DNA that make it useful in tracing evolution history, founder events, introductions, and population bottlenecks (Harrison 1989).

The mitochondrial gene has demonstrated its importance as a maker in species identification through DNA barcoding (Mohanty et al. 2015; Bhattacharya et al. 2016), thus helping in overcoming the uncertainty in the morphological identification of species. The mitochondrial gene COI has been declared a basic tool for molecular taxonomy and identification (Ratnasingham and Hebert 2007) and we have used it to resolve the morphological ambiguities that we faced during the sample collection. We have also used concatenated mitochondrial genes: ATPase 6/8 and Cyt b as the marker of choice for the population studies in khudree (Sahoo et al. 2019), since it is effective in showing considerable among and within-population variation, which provides information on the structuring of populations (Avise et al. 1987). Hence, the present study is done to comprehend the genetic diversity and population structure among four different populations of khudree as a step toward understanding the genetics of this fish that might help further in the management and conservation of the species.

Material and Methods

Sample Collection and DNA Extraction

Fin clipping of the species was performed between April 2015 and March 2018 from River Tungabhadra, River Tunga, River Periyar, and River Cauvery (Table 1, Fig. 1). Live fish were caught with prior permission from the concerned Department of Forests and wild life at the protected area randomly and after careful collection of fin tissue, they were released back into the water. Fin clippings from 174 collected specimens were preserved in 95% ethanol at −20 °C until further use. The congeneric species Neolissochilus hexagonalepis from GenBank was chosen as the outgroup (KU380329) in Phylogenetic analysis. High molecular weight genomic DNA was extracted using the standard phenol–chloroform method (Sambrook et al. 1989), resuspended in Tris–EDTA buffer, and stored at −20 °C until use. The integrity of the DNA was checked using agarose gel electrophoresis and Nanodrop Biospectrometer (Eppendorf India Pvt. Ltd.). The DNA was then diluted to 25 ng/µl working solutions.

Table 1 Sampling details and Molecular Diversity indices in Tor khudree
Full size table
Fig. 1
figure 1

Schematic map of Indian Rivers showing the place of collection of samples

Full size image

PCR Amplification and Sequencing

To avoid any ambiguities in the identification of the species during collection, all 174 samples were verified using DNA barcoding which involved PCR amplification of a 600 bp fragment of the COI gene using universal primers, FishF1-5′ TCA ACC AAC CAC AAA GAC ATT GGC AC 3′ and FishR2-5′ ACT TCA GGG TGA CCG AAG AAT CAG AA- 3′ (Ward et al. 2005). A 25 µl PCR mixture contained 1.0–2.0 µl of DNA template (25 ng/ml), 2.5 µl 10X PCR buffer, 1 µl of each primer (10 pmol/µl), 1.5 µl MgCl2 (20 mM), 2 µl dNTP mix (2.5 mM each), and 0.25 U of Taq DNA polymerase. The reactions were performed on a GeneAmp PCR System (Applied Biosystems Inc., Foster City, CA) for 35 cycles under the conditions: one cycle of initial denaturation at 94 °C for 5 min followed by 30 s at 94 °C, 45 s at 58 °C, and 30 s at 72 °C for COI, with a final extension at 72 °C for 7 min. The PCR products were purified using Qiagen PCR purification kit followed by bidirectional cycle sequencing on ABI 3100 PE automated capillary sequencer.

Partial ATPase 6/8 and Cyt b genes from the verified samples were amplified using universal primers, ATPase 6/8—L8331 5′AAA GCR TTR GCC TTT TAA GC 3′ and H9236 5′ GTT AGT GGT CAK GGG CTT GGR TC 3′ (Thai et al. 2005) and a combination of Universal forward primer (Kocher et al. 1989) and carp-specific reverse primer (Das et al. 2014) in case of Cyt b—L14841 5′ AAA AAG CTT CCA TCC AAC ATC TCA GCA TGA TGA AA 3′ and CIFA-Cyt b-R2 5′ CTC ATT CTA GTG CTT TAT TTT C 3′. A 25 µl PCR mixture contained 1.0–2.0 µl of DNA template (25 ng/ml), 2.5 µl PCR buffer, 1 µl of each primer (10 pmol/µl), 1.5 µl MgCl2 (20 mM), 2 µl dNTP mix (2.5 mM each), and 0.25 U of Taq DNA polymerase. The reactions were performed on a GeneAmp PCR System (Applied Biosystems Inc., Foster City, CA) for 35 cycles under the conditions: one cycle of initial denaturation at 94 °C for 5 min followed by 30 s at 94 °C, 1 m at 55 °C, and 2 m at 72 °C for ATPase 6/8 and 30 s at 94 °C, 40 s at 50 °C, and 45 s at 72 °C for Cyt b, with final extension at 72 °C for 7 min. The PCR products were purified using Qiagen PCR purification kit followed by bidirectional cycle sequencing on ABI 3100 PE automated capillary sequencer.

Species Identification, Generation of Consensus Sequences

The COI sequences obtained were edited in BioEdit version 7.0.5.3 (Hall 1999). The edited sequences were fed in FASTA format into BOLD Systems v3 (https://v3.boldsystems.org/) for species identification. The sequences obtained after the sequencing of ATPase 6/8 and Cyt b genes of the confirmed individuals were edited and aligned using CLUSTALW in BioEdit version 7.0.5.3 (Hall 1999) to get the final set of consensus sequences.

Genetic Diversity and Population Differentiation Study

The consensus sequences of mitochondrial genes were concatenated manually using BioEdit version 7.0.5.3 (Hall 1999) based on the concept of supergene topology (Havird and Santos 2014). The sequential genetic information was estimated, which included the number of polymorphic sites, parsimony informative sites, Haplotype diversity, nucleotide diversity, private haplotypes using the DnaSP version 6 (Rozas et al. 2017). Genetic differentiation among the Tor khudree populations was calculated using the F-statistics (FST) which was estimated by using conventional method using haplotypic frequencies with 1023 nonparametric permutations. The significance of FST was determined by the p value generated in Arlequin v3.1.1 (Excoffier and Lischer 2006) and hierarchical partitioning of genetic diversity within and between populations was determined by Analysis of Molecular Variance as implemented in Arlequin v3.1.1.A median-joining network was also constructed using PopART v. 1.7 (Bandelt et al. 1999) to better illustrate the association between the haplotypes.

Phylogenetic Analysis, Population Demography

The phylogenetic analysis was conducted using MEGA 10 (Kumar et al. 2018). A maximum-likelihood tree was constructed with HKY + GAMMA + I model as the best-predicted model from jModelTest 2.1.10. The congeneric species Neolissochilus hexagonalepis was taken as the outgroup (KU380329). The robustness of the test was found out using the bootstrap method of 1000 replications. Only the haplotypes of the concatenated sequences were included for the generation of a phylogenetic tree.

Demographic history was calculated for Tor khudree by distribution of the number of differences in site between the pair of sequences referred as mismatch distribution. For a model of constant population size, calculated expected values were plotted against the observed values using DnaSP6 program. To check whether the observed distribution deviated significantly from the expected values under the population expansion model, Harpending’s raggedness index (Hri) Harpending (1994) and the sum of squared deviation (SSD) were calculated using Arlequin. The Tajima’s D test (Tajima 1989) and Fu’s Fs test (Fu 1997) were used to test the hypothesis of neutrality of the concatenated mitochondrial genes in DnaSP v.6.

Results

Sequence Analysis, Genetic Diversity, and Population Differentiation Study in T. khudree

Out of 174 samples collected, 152 samples, i.e., 39 out of 39 in Tungabhadra, 47 out of 47 in Tunga, 27 out of 43 in Periyar, and 39 out of 45 in Cauvery got verified as Tor khudree. Consensus sequences of 790 bp of ATPase 6/8 gene (MZ057291–MZ057442) and 1000 bp of Cyt b gene (MZ057443–MZ057594) were obtained from the sequencing of the PCR products. Analysis of the ATPase 6/8 and Cyt b genes revealed 10 and 21 different haplotypes, respectively. The genetic diversity analysis of both the genes is described in the supplementary data 1 (S1). Analysis of Molecular Variance (AMOVA) using ATPase 6/8 gene revealed 50.83% among population variation and 49.17% within population variation with pairwise FST ranging from −0.01624 to 0.70977 (p < 0.05). Significant genetic differentiation was observed among all population pairs except Periyar-Cauvery (−0.01624). Similarly, AMOVA using the Cyt b gene showed 14.87% among population variation and 85.13% within population variation with pairwise FST ranging from 0.05 to 0.28 (p < 0.05) indicating moderate to high genetic differentiation among all populations.

Analysis of concatenated sequences of both genes in four populations revealed 44 haplotypes. Tungabhadra has the highest with 24 haplotypes followed by Cauvery with 14, Tunga with 11, and Periyar with 07 haplotypes. Out of 1790 sites in the concatenated genes, 27 sites were polymorphic, 14 were parsimony informative sites, and 13 were singleton variable sites. The average nucleotide composition was 27.2% T, 29.2% C 30.0% A, and 13.6% G and it was biased toward the A + C content. The haplotype diversity (Hd) and nucleotide diversity (π) ranged from 0.792 to 0.968 and 0.00068 to 0.00157, respectively. Periyar population showed the lowest Hd and π and Tungabhadra showed the highest Hd and π (Table 1). The frequency of each haplotype in a population is provided in Supplementary data 2 (S2).

AMOVA illustrated the population structure using the concatenated sequences of both genes. The results revealed 28.98% among population variation and 71.02% within population variation (Table 2). Pairwise FST ranged from 0.121 to 0.372 (p < 0.05) indicating significant genetic differentiation among all population pairs. The highest genetic differentiation was observed between Tunga and Periyar populations and it was lowest between Tunga and Tungabhadra populations (Table 3).

Table 2 AMOVA for concatenated gene sequences of four riverine populations of khudree
Full size table
Table 3 Pairwise FST values determined from concatenated mitochondrial gene analysis (p < 0.05)
Full size table

Median-joining tree revealed Haplotype 4 (TKACH04) to be found in most of the individuals of the three populations: Tungabhadra, Tunga, and Periyar as it was present in the center of the network but absent in the Cauvery population (Fig. 2A). Haplotype 9 was also shared by the above three populations. Haplotype10 was shared by three populations (Tungabhadra, Tunga, and Cauvery). Haplotype 17 was shared by three populations (Tungabhadra, Periyar, and Cauvery). Haplotypes 27 and 29 were shared by two populations (Tunga and Cauvery). Haplotypes 34 and 35 were shared by two populations (Periyar and Cauvery). Many unique haplotypes were found not shared by any populations; the highest number of unique haplotypes was seen in the Tungabhadra population (20) and the lowest was seen in the Periyar population (1). Haplotypic distribution has been shown in the sampling locations (Fig. 2B).

Fig. 2
figure 2

A Median-Joining Network of 44 haplotypes observed in T. Khudree populations B Mitochondrial haplotype distribution at sampling locations

Full size image

Phylogenetic Analysis, Population Demography Analysis

Phylogenetic analysis revealed the monophyly of the Tor khudree haplotypes that were separated a hundred percent from the outgroup species Neolissochilus hexagonolepis (Fig. 3). The results of the mismatch analysis and neutrality tests along with the associated p-values are represented in Table 4. The Tajima’s D was negative for all populations, which indicated the presence of an excess of rare nucleotide variants compared to what was to be expected under the neutral model of evolution. However, the results (−0.27, p > 0.05) showed no significant deviations from the null hypothesis of selective neutrality for all the populations. The individual population also showed similar results. In contrast, Fu’s Fs tests for the overall population and individual stocks, which are based on the distribution of haplotypes, though showed negative values had a significant deviation from neutrality (−8.718, p < 0.05) suggesting populations are showing population size expansion and/or purifying selection.

Fig. 3
figure 3

Phylogenetic relationship among Tor khudree populations using concatenated sequences of ATPase 6/8 and Cyt b genes. A maximum Likelihood tree with 1000 replicates, and HKY + G + I model with Neolissochilus hexagonolepis as outgroup

Full size image
Table 4 Mismatch distribution and neutrality tests for Tor khudree
Full size table

Mismatch distribution Supplementary data 3 (S3) for the overall population as well as individual population showed a unimodal plot which indicated that the population had undergone a recent population expansion after a bottleneck. The estimated values of Harpending’s raggedness index and SSD were non significant which suggested that the data is a good fit between the observed and expected values of the pairwise distributions. The data supported the recent population expansion.

Discussion

Our study aimed at revealing the population genetic structure of Deccan mahseer across the Deccan region of peninsular India. The utility of maternally inherited Mt DNA has been used in the present study for species identification as well as to assess the genetic diversity and population structure of the endangered species, endemic to India. The highly conserved mitochondrial gene, COI, was used in species identification through barcoding (Ward et al. 2005; Lakra et al. 2011). Out of collected 174 samples, 152 got confirmed as Tor khudree, and the rest were found to be mostly Tor malabaricus, followed by Tor putitora and Tor tor. The least number of samples that got confirmed was from the Kerala population also indicated the coexistence of two types of mahseer in Periyar waters, Tor khudree and Tor malabaricus (Raghavan et al. 2011). Since mitochondrial genes have a high rate of evolution, it allows for distinguishing between haplotypes in small sample sizes and threatened species (Habib et al. 2012; Sharma et al. 2019). They have been used in the study to examine the genetic variability and structure of the species. Analysis using the two mitochondrial genes revealed both the genes are AT rich as reported in other studies (Johns and Avise 1998). Genetic variability (haplotypes diversity and nucleotide diversity) of ATPase 6/8 gene was lower than the Cyt b gene due to high conservness of ATPase 6/8 gene during evolution (Sati et al. 2013). We have also concatenated both the genes based on the super gene concept to get a better resolution on the population genetics of Tor khudree since mtDNA markers show different levels of rate of evolution (Reeb and Avise 1990; Havird and Santos 2014; Kim et al. 2014; Sahoo et al. 2019). The combined mitochondrial sequences have proved to be very useful in delineating genetic diversity and differentiation in many aquatic species including mahseers (Kim et al. 2014; Sahoo et al. 2019; Sharma et al. 2019; Sah et al. 2020). The nucleotide composition in the concatenated sequences was biased toward A + T as seen in other fishes (Johns and Avise 1998). The order of base composition was A > C > T > G where nucleotide G was least represented similar to other carps (Bej et al. 2012). The presence of low Guanine is a special feature of the mitochondrial genome (Cantatore et al. 1994; Zhu et al. 2017; Balasubramaniam et al. 2022). Genetic diversity study using the concatenated mitochondrial genes revealed the presence of 44 haplotypes with the Tungabhadra population having the highest number of haplotypes (24). The number of haplotypes ranged from 7 to 24. Due to a large number of haplotypes and polymorphic sites, the Tungabhadra population could be considered a highly divergent population. Sharing of haplotypes was seen to be limited to three populations as evident from the median-joining network (Fig. 2A). This indicated the common evolutionary history among the individuals of those populations as well as gene flow between populations (Das et al. 2014). A close genetic relationship is suggestive among the populations, though the oldest haplotype is absent in the Cauvery population. Many haplotypes were also having limited distribution to a single population; out of 44 haplotypes, 36 were private haplotypes (81.8%) which suggested notable population differentiation, as well as the absence of population admixture (Sharma et al. 2019), or these haplotypes, may have arisen due to independent mutation events (Das et al. 2018). These unique haplotypes may have arisen due to environmental changes at their respective locations (Balasubramaniam et al. 2022). These private haplotypes could be used as a marker for stock identification (Sahoo et al. 2019). The distribution of mitochondrial haplotypes is reported to be proportional to their age (Neigel and Avise 1993; Kim et al. 2020). Haplotype 4 (TKACH04) is present in most of the individuals and perhaps the ancestral haplotype which is evident from the median-joining network as it is present in the center of the network. Haplotype diversity ranged from 0.792 to 0.968 and π ranged between 0.00068 and 0.00157 which was comparable with the other mahseers (Sati et al. 2015; Sharma et al. 2019) in terms of high haplotypic diversity and low nucleotide diversity. This may be due to sudden population growth and accumulation of mutation after diverging from a population facing bottleneck events (Avise et al. 1984; Grant and Bowen 1998). The range of Hd and π seen in khudree is similar to the range observed in other freshwater fishes (Habib et al. 2012).

Population Differentiation and Stock Structure

FST is one of the most widely used estimates for detecting genetic differentiation between and among populations. This measure provides information on the evolutionary events that may cause genetic differentiation among populations (Holsinger and Weir 2009). In general, FST values of 0–0.05 represent little differentiation; values of 0.05–0.25 indicate moderate differentiation and values higher than 0.25 indicate great differentiation among populations (Wright 1978). In this study, the FST between the four populations ranged from 0.121 to 0.372, indicating a moderate to high level of genetic differentiation (Sati et al. 2013; Sahoo et al. 2019; Sharma et al. 2019). The overall FST was found to be 0.289, p < 0.05 suggesting high and significant genetic differentiation among the khudree populations and freshwater fishes tends to exhibit higher levels of genetic differentiation as compared to marine fishes (Ward et al. 1994). AMOVA using ATPase 6/8 gene revealed high among population variation as compared to high within population variation using Cyt b. The results based on the concatenated sequences also showed high intrapopulation variation, concurrent with the information on the Cyt b from our study and the other study on Tor mahseer (Sah et al. 2020), suggested possible gene flow among the populations and this result is comparable to other mahseers where high among population variation is observed (Esa and Abdul Rahim 2013; Sati et al. 2013; Sharma et al. 2019). This observation is also in congruence with Vrijenhoek 1998, who described that low among population variation and high within population variation is expected in migratory fishes like mahseer.

Phylogenetic Study and Population Demography

Phylogenetic relationships using the HYK + I + G model did not show the populations to separate into clusters instead mixed clusters were seen indicating the sharing of haplotypes among the population. The neutrality test of Tajima’s D revealed nonsignificant negative values for all populations of T. khudree which supported the changes in concatenated Cyt b and ATPase 6/8 for the species in the collection sites are selectively neutral. However, the neutrality test of Fu’s F revealed significant negative values and the mismatch distribution of all the population was seen to be unimodal thus suggesting possibilities of recent population expansion or a purifying selection (Hudson et al. 1992; Rogers and Harpending 1992). The unimodal pattern could also be due to range expansion with high levels of migration in between the demes (Ray et al. 2003; Excoffier 2004).

Conclusion

Our work on genetic diversity of Tor khudree using the mitochondrial markers with samples from four different rivers of Deccan India has been done for the first time. The result suggested the importance of mitochondrial markers for studying the variations within and among population of Tor khudree. High genetic differentiation was seen in all khudree populations. The highly divergent Tungabhadra population can serve as conservation unit. The within population variation in khudree would provide wide genetic base for any genetic improvement program. The results from this study can be compared with the results obtained on using highly advanced markers such as SSR and SNP for better understanding the genetic differentiation and genetic structure of the esthetic Deccan mahseer.