Vertebral monstrosities: phenotypically shortened fish with deformed vertebrae in endemic fish genus Hypselobarbus (Bleeker, 1860), (Teleostei: Cyprinidae) from Western Ghats, India

Abstract

Background

Fish with vertebral monstrosities are very rare in the wild, as those individuals in the natural populations tend to perform poorly to survive in any ecosystem. Species of the fish genus Hypselobarbus as reported (Bleeker in De visschen van den Indischen Archipel, Lange, 1860) are freshwater endemic barbs of Western Ghats and peninsular India. Four species of the genus, namely Hypselobarbus dobsoni (Krishna carp), H. jerdoni (Jerdon’s Carp), H. lithopidos (Canara barb) and H. thomassi (Red Canarese barb), were collected from three different river systems of the Western Ghats biodiversity hotspot of India. Some individuals were found to be different from normal specimens, with extremely large body depth compared to normal specimens. The study was initiated with the aim of bringing an understanding on monstrosities of these four species along with identifying the normal and abnormal individuals in an integrated approach; employing traditional morphometry, X-ray imaging and barcoding mtDNA COI X-ray imaging could elucidate the vertebral monstrosities, which are discussed in detail. The mtDNA COI gene sequences generated were used to draw conclusions on identity of both normal and deformed individuals.

Results

The phenotypic deformities have led to deepening of the body with a more robust and reduced length which is evident from the morphometric comparison of normal specimens with deformed ones. The radiographic images revealed reduced intra-vertebral space in comparison with the normal vertebrae, deformed vertebrae were between 25 and 32, showing significantly altered intra-vertebral space. Slight genetic divergence of 1.1% between normal and deformed specimens in mitochondrial DNA COI gene of H. lithopidos and H. thomassi and no divergence in H. dobsoni and H. jerdoni were also observed.

Conclusion

The specimens were collected from areas with high anthropogenic stresses, abate water quality, and habitat, which could be possible reasons of appearance of individuals with deformed vertebrae. Several environmental and genetic factors might have influenced the development of these robust short-bodied phenotypes in these rivers and possess slight genetic divergence from normal specimens. However, these deformities may also be the result of the stress during embryonic and early life stages in the wild.

Background

Monstrosities in fishes have been of great curiosity for the ichthyologists since the early days, as indicated in the description of fishes with monstrosities even in the sixteenth century (Aldrovandi, 1613). The records of monstrosities have been published as bibliography of anomalies in fishes (Dawson, 1964, 1966, 1971; Dawson & Heal, 1976). Darwin emphasized on distinction between morphological anomalies and typical variations, but evolutionary biologists stressed more on morphological defects that are genetically based and may be influenced by natural selection as potential novelties (Darwin, 1875). Many abnormalities appear to be addressed as monstrosities, specifically when the vertebral column is shortened without visible curvature, resulting in an odd deepened short body (Golubtsov et al., 2021). Morphological abnormalities develop during the embryonic and post-embryonic phases due to unclear causes or poorly understood mechanisms (Gray et al., 2021). The incidence of malformed specimens is minimal in natural aquatic environments, unlike cultured systems where deformities are quite common (Dahlberg, 1970; Daoulas et al., 1991). With the reports of declining water quality, numerous research has identified deformity as the biological variable indicating contamination in natural waterbodies (Bengtsson et al., 1997).

Fishes have dispersed all over the planet and shown an unmatched diversity in their appearance, habitat, physiology, and behaviour since their origin (Nelson et al., 2016). In India, freshwater fishes from the Western Ghats biodiversity hotspot comprise 320 species representing 11 orders, 35 families, and 115 genera (Bijukumar & Raghavan, 2015; Dahanukar & Raghavan, 2013). The barbs of genus Hypselobarbus (Bleeker, 1860) (Cyprinidae, Teleostei; Cypriniformes) (Tan & Armbruster, 2018) are native to rivers of the Western Ghats and peninsular India (Ali et al., 2013; Arunachalam et al., 2012; Knight et al., 2013). The genus includes 13 valid species, (Knight et al., 2016). These fishes vary between 25 and 100 cm in total length. The fishes of the genus Hypselobarbus are caught in riverine fishing gears by the local fishers for local consumption and are economically important inland fish catches in markets (Arunachalam et al., 2012). Some species are candidate species in aquaculture industry, and smaller sizes are in ornamental fish trade (Ali et al., 2013; Raghavan et al., 2018). Preliminary investigation of body shape through visual examination revealed extreme body deformity in individuals of four species of Hypselobarbus collected from three different rivers Tungabhadra, Netravathi, and Periyar of Western Ghats. There are no reports of monstrosities in any fish from Western Ghats of India. Several deformed fishes of the genus Hypselobarbus were collected from different rivers of Western Ghats. With this, the present study aimed at a comparison of body depth and other morphometric measurements with normal and deformed specimens of four species collected. The study also aimed at the radiographic investigation of the vertebral column in normal and deformed specimens, along with genetic comparison of mitochondrial DNA Cytochrome oxidase subunit 1 for species confirmation as well as intraspecific variation between normal and deformed specimens.

Methods

Sampling site

Fishes of genus Hypselobarbus were collected during February to June and October to December, of 2021 to 2023, from River Tungabhadra, originating at Gangamoola Peak in the Khudremukh range of Karnataka in Central Western Ghats, River Netravathi, which is also originating at Gangamoola Peak and thirdly River Periyar, the largest river system in southern Western Ghats, originating in Sivagiri hills in Kerala. Fish samples were collected from the commercial catches of fishers along the rivers who used gill nets of varying mesh sizes ranging from 40 to 100 mm, the normal and deformed specimens of H. dobsoni (n = 10) from Bhadravathi and Shivamogga in Tungabhadra river, H. jerdoni (n = 15), H. lithopidos (n = 3), and H. thomassi (n = 36) near to B.C. Road, Mangaluru, Karnataka and near Aluva in Kerala. The coordinates are available in table (Supp_Table. 1) and has been plotted as a GIS map (Fig. 1).

Fig. 1

Map of sampling sites from Tungabhadra river, Netravathi river, and Periyar river of Western Ghats

Morphological identification

Species were identified using the original descriptions and available literature (Day, 1874, 1876, 1888; Menon & Remadevi, 1995; Sykes, 1839). A total of 23 morphometric characters were measured using a digital Vernier calliper with 0.1 mm accuracy (Hubbs et al., 2004). The X-ray images of both normal and deformed specimens were captured, followed by counting vertebrae (Golubtsov et al., 2021; Naseka, 1996). The fish muscle tissue samples of the specimens were taken and stored in 100% ethanol. After a day, the alcohol was discarded and filled with fresh ethanol. The fish specimens are stored in 8% formalin solution and deposited at the Aquatic Biodiversity Museum and Repository, ICAR-CIFE, Mumbai.

DNA isolation and PCR amplification

Genomic DNA was isolated from 20 mg of muscle tissue stored in absolute ethanol from all four deformed samples and normal representative samples using the organic extraction method (Phenol–chloroform method) (Taggart et al., 1992). The mitochondrial partial cytochrome c oxidase subunit 1 (650 bp) gene was amplified as described in Ward et al. (2005), using the primer set, FishF1-5’TCAACCAACCACAAAGACATTGGCAC3’ FishR1-5’TAGACTTCTGGGTGGCCAAAGAATCA3’.

PCR was performed following Jeevan et al. (2024). The PCR products were visualized on 2% agarose gel, and the amplicons were purified using a PCR purification kit (MinElute PCR Purification Kit) following the manufacturer’s protocol. The purified products were sequenced in both directions using the PCR primers from Eurofins Pvt Ltd., Bangalore, India.

Sequence analysis

The sequence quality was assessed by estimating the Phred score of each base using Finch TV software (version 1.4.0) (Geospiza, 2009). The sequences were aligned using MEGA (version 11.0): Molecular Evolutionary Genetics Analysis version 11 (Tamura et al., 2021) and were subjected to similarity analysis with the NCBI database using the BLAST tool. Sequence divergence values within the normal and deformed specimens were calculated using Kimura two Parameter (K2P) distance model (Kimura, 1980) implemented in MEGA (version 11.0), with 1000 bootstrap replications. Maximum likelihood phylogenetic inference was performed using IQ-TREE (Trifinopoulos et al., 2016), where node support was evaluated through ultrafast bootstrap support over 1000 iterations (Minh et al., 2013). The most suitable partitioning models for the datasets were identified using ModelFinder (Kalyaanamoorthy et al., 2017) within IQ-TREE, following the selection based on the minimum BIC (Bayesian Inference Criterion) score.

Statistical analysis

All analyses were performed using R version 4.1.1 (R Core Team, 2022). The descriptive statistics for the normal specimens of all four species were derived using the package summarytool (Comtois, 2018). All 23 morphometric measurements (in millimetres) were allometrically size-corrected by taking Standard Length as base using PAleontological STatistics PAST 4.13 (Hammer et al., 2001), for three species, except for H. lithopidos as the number of total specimen were not sufficient (n = 3) and hence were not included in further analysis. Univariate analysis of variance (ANOVA) was performed for all the size-corrected measurements (Supp_Table. 2). Significant variables (p < 0.05) after ANOVA were considered for further analysis. Principal component analysis (PCA) was performed as a data reduction technique to determine the variables responsible for the variation. The normal and deformed vertebrae along with the total number of vertebrae (Table 3) are represented in bar graph (Fig. 2). All visualizations were done using ggplot2 (Wickham & Wickham, 2016) and RColorBrewer packages (Neuwirth & Brewer, 2014).

Fig. 2

Bar plot showing number of normal and deformed vertebrae in all four species of Hypselobarbus (HD—H. dobsoni normal specimen, HD*, HD**, HD***—H. dobsoni deformed specimens; HJ—H. jerdoni normal specimen, HJ*—H. jerdoni deformed specimen; HL—H. lithopidos normal specimen, HL*—H. lithopidos deformed specimen; HT—H. thomassi normal specimen, HT*, HT**—H. thomassi deformed specimens) (TNV—total number of vertebrae; NVNS—number of vertebrae in normal specimens; NVDS—number of vertebrae in deformed specimens)

Results

Short definition

The deformed specimens were found to have relatively higher body depth, head depth, and shorter body profile than other normal specimens (Fig. 3) (Table 1 and Table 2). The radiographic image (Fig. 4) showed an inverse proportionate number of deformed vertebrae with shortness of the specimen. The normal specimens H. dobsoni have 35 vertebrae, H. jerdoni and H. thomassi have 34 vertebrae and H. lithopidos have 36 number of vertebrae (Table 3).

Fig. 3

External appearance of normal (left) and deformed (right) specimens. A Normal specimen of H. dobsoni, B short, deformed specimen of H. dobsoni, C normal specimen of H. jerdoni, D short, deformed specimen of H. jerdoni, E normal specimen of H. thomassi, F short, deformed specimen of H. thomassi, G normal specimen of H. lithopidos, H short, deformed specimen of H. lithopidos

Table 1 Body measurements in % standard length and % head length

Table 2 Descriptive statistics of normal specimens of all four species of Hypselobarbus

Fig. 4

Radiographic images of normal (left) and deformed (right) specimens with compressed intra-vertebral space. A Normal specimen of H. dobsoni, B short, deformed specimen of H. dobsoni, C normal specimen of H. jerdoni, D short, deformed specimen of H. jerdoni, E normal specimen of H. thomassi, F short, deformed specimen of H. thomassi, G normal specimen of H. lithopidos, H short, deformed specimen of H. lithopidos

Table 3 Normal and deformed vertebrae in all four species of Hypselobarbus

Morphometric differences in normal and deformed specimens

Hypselobarbus dobsoni (Day, 1876): Normal specimens (Fig. 3A) had an average percentage of 36.5%, Body Depth (BD) to Standard Length (SL), while the deformed specimens (Fig. 3B) had 50.5%; Head Depth (HD) to SL proportion was 17% in normal specimens and 22.5% in the deformed specimens. Caudal Peduncle Depth (CPD) to SL proportion was 13.5% in normal specimens and 17% in deformed specimens; intra-narial width (INW) to proportionate head length (HL) in normal specimens was 30% and 36% in the deformed specimens.

Hypselobarbus jerdoni (Day, 1876): Normal specimens (Fig. 3C) had an average BD to SL proportion of 34.5%, while the deformed specimen (Fig. 3D) had 53%; HD to SL proportion was 18% in normal specimens and 26% in the deformed specimen; CPD to SL proportion was 14% in normal specimens and 18% in the deformed specimen. INW to proportionate HL in normal specimens was 29% and 34% in a deformed specimen.

Hypselobarbus lithopidos (Day, 1874): Normal specimens (Fig. 3G) had an average BD to SL proportion of 31%, while the deformed specimen (Fig. 3H) had 42%; HD to SL proportion was 15.5% in normal specimens and 21% in the deformed specimen; CPD to SL proportion was 12% in normal specimens and 15% in the deformed specimen. INW to proportionate HL in normal specimens was 26% and 28% in a deformed specimen.

Hypselobarbus thomassi (Day, 1874): Normal specimens (Fig. 3E) had an average BD to SL proportion of 31%, while the deformed specimens (Fig. 3F) had 43.5%; HD to SL proportion was 16.5% in normal specimens and 22% in deformed specimens; CPD to SL proportion was 12.5% in normal specimens and 18% in deformed specimens. INW to proportionate HL in normal specimens was 25% and 42% in deformed specimens.

The descriptive statistics of 23 morphometric variables given for normal specimens of all four species with mean, standard deviation, and range are presented (Table 2). PCA is one of the best multivariate analytical techniques which can be used to reduce morphometric data and extract the independent or explanatory variables significant for variation. The independent variables, showing significance (p < 0.05) after univariate ANOVA (Supp_Tables 2), were subjected to PCA. The PCA factor loadings for each species (H. dobsoni, H. jerdoni, and H. thomassi) including proportion of variance and cumulative proportion of each PCA for three species shows each of the variables contributing to the PC loadings (Supp_Table 3). The PCA biplot (Fig. 5), shows PC1 contribution of 77.17%, and PC2 18.84% in the case of H. dobsoni. In the case of PCA biplot (Fig. 6) of H. jerdoni, PC1 shows 84.23% of cumulative proportions and 9.27% in PC2. PCA biplot of H. thomassi (Fig. 7), shows that 71.83% of variation in PC1 and 9.74% in PC2. In all the three PCA biplots generated for three species, the deformed individuals are clearly separated from the normal specimens which are clustered in ellipse (Figs. 5, 6, 7).

Fig. 5

PCA plot distinguishing normal and deformed specimens of H. dobsoni, (HD—normal specimen, HD*, HD**, HD***—deformed specimens)

Fig. 6

PCA plot distinguishing normal and deformed specimens of H. jerdoni, (HJ—normal specimen, HJ**—deformed specimen)

Fig. 7

PCA plot distinguishing normal and deformed specimens of H. thomassi, (HT—normal specimen, HT*, HT**—deformed specimens)

Genetic divergence and neighbour-joining tree

The K2P distance estimated for deformed individuals and the normal specimens (Table 4) shows the conspecific divergence with a maximum in H. thomassi (1.17%), as fewer haplotypes were found in the sequences. The average congeneric divergence was 5.75%, 15 folds more than the average conspecific divergence value. The results of maximum likelihood tree (Fig. 8) also showed four different nodes formations for all four species with high bootstrap coverage and the individuals of the species restricted to the particular node, The alignment had 12 sequences with 684 bases, 68 distinct patterns, 68 parsimony-informative, 556 constant sites and the best-fit model chosen according to BIC was TIM2 + F + I with optimal log-likelihood value − 1695.621. With the molecular sequencing data, we confirm all the deformed specimens were identified in the correct taxon (accession numbers: Supp_Table. 5).

Table 4 Genetic divergence table converted to percentage variation between normal and deformed specimens

Fig. 8

Maximum likelihood tree for normal and deformed specimens of genus Hypselobarbus

Discussion

The present study could clearly distinguish deep-bodied phenotypic specimens with shortened inter-vertebral space, the only report of such deformities in the genus Hypselobarbus and the second in cyprinids. Short and deep-bodied morphotypes are not reported so far in any of the cyprinids group from the rivers of Western Ghats, India. The species Hypselobarbus curmuca showing several anomalies is the only report in the genus Hypselobarbus (Jeevan et al., 2024). A report on barbs of family Cyprinidae is of a species of the genus Labeobarbus from Africa (Golubtsov et al., 2021). The other reports include spinal compression in Gadus morhua (Atlantic cod) from the German Wadden sea, where the seasonal prevalence and the rate of occurrence were studied (Hilger, 1992). The most probable causes for the common skeletal malformations is linked to physiological, environmental, xenobiotic, dietary, and genetic factors (Gjerde et al., 2005; Huggett, 2018; Madsen et al., 2001), while there is no confirmed proof of any of the single parameter or factor affecting fish to be malformed. In our study, we also found a slight genetic divergence of 1.13% in H. lithopidos and 1.17% in H. thomassi and no genetic divergence in H. dobsoni and H. jerdoni between normal and deformed specimens, which is insufficient to prove them as different species.

Possible factors for shortened phenotypic deformities

To the best of our knowledge, no records are available on the factors leading to such phenotypic short-bodied forms from any wild fish population. However, several studies have been reported on farmed salmonids (Gjerde et al., 2005; Kvellestad et al., 2000; McKay & Gjerde, 1986; Vagsholm & Djupvik, 1998; Witten et al., 2009). These studies also could not conclude with the exact cause and explained as multiple factors, viz. parasitic and bacterial infections, deficiency of micro and macronutrients, vaccination, higher temperatures during the early embryonic stage, fluctuation in photoperiod, fluctuation in water quality, water current and environmental pollutions (Witten et al., 2009).

Possible effect of predator-driven phenotypic plasticity

Extreme climatic events like floods during the monsoon season are very common in Western Ghats on account of heavy rainfall (Vijaykumar et al., 2021). These events caused the escape of exotic fish species under the culture system into the wild and are currently found in almost all the freshwater ecosystems of the Western Ghats (Raj, et al., 2021a, b). Non-native species are being reported from the lakes, rivers, and reservoirs of the Western Ghats (Bijukumar, 2019). Reservoirs of Western Ghats have greatly been focused on stocking non-native species for capture-based culture fisheries, which became the source for exotic species to spread all over the river's catchment (Sugunan, 1995). The alien fish species have outcompeted and established successfully (Raj et al., 2021b). The accidental or deliberate introduction of exotics increases the predation stress along with the native predators, particularly in the early life stages (De Leaniz et al., 2010). It is already known that smaller sized individuals have more pressure due to predation during their fast-growing phase (Sogard, 1997). The predator-driven environment tends to have greater body depth (Eklöv & Svanbäck, 2006) and also tends to diverge genetically from the normal ecosystems (Ingley et al., 2014), which is true in some species of Brachyrhaphis fishes. Similarly, the presence of invasive predatory fish species, such as African catfish, in huge numbers from rivers of the Western Ghats (Pillai et al., 2016; Raghavan et al., 2016; Raj, et al., 2021a, b; Ranjan, 2018; Roshni et al., 2020; Sreenivasan et al., 2021), maybe the reason for the occurrence of short-bodied phenotypic specimens.

Environment-driven phenotypic plasticity

Humans pose the world’s greatest evolutionary force, altering global ecology and evolutionary trajectories and dramatically accelerating mutation in species associated with particular ecosystems (Palumbi, 2001). There have been ample reports on phenotypic plasticity from anthropogenically altered riverine habitats and significant divergence in the body shape of fishes residing in reservoirs and streams (Brinsmead & Fox, 2002; Franssen, et al., 2013a, b). Few fishes which occupy the reservoir have shown relatively deeper bodies and smaller heads compared to their riverine counterparts, and few others showed deeper bodies in the riverine habitat, which concludes that phenotypical evolutionary traits are purely species-specific (Franssen, et al., 2013a, b; Haas et al., 2010). Environmental factors affect phenotypes as a single parameter or an interactive environmental variable with two or more combined parameters, such as dissolved oxygen and water flow (Langerhans et al., 2007). Alteration in riverine habitat possesses a novel selection pressure on the native fish fauna by altering body shape, evident through genetic-based morphometric plasticity in reservoir and stream fishes (Franssen, 2011). The stress due to pollution or any anomalies in abiotic environmental parameters could also influence the eggs and larvae to get deformed body shapes while they grow out (Sfakianakis et al., 2015), During peak summer, Netravathi River dries up completely. The dammed segments show the least fish species richness due to low or no flow, as a result of small hydropower projects along the river at various places (Jumani et al., 2017). In addition to that, waters have been characterized by elevated temperature and reduced dissolved oxygen (Jumani et al., 2018). The surface sediment and water are found to have a high load of heavy metal pollutants like lead, which may severely affect metabolic activities in riverine biota and community structures (Gayathri et al., 2021). All three species, except H. dobsoni, collected from Netravathi river, are in the catchment of Tumbe small hydropower project may have led to such phenotypic deformities. The location from where the specimens of H. dobsoni were collected is reported to have comparatively low dissolved oxygen and high biological oxygen demand as the river stretch is packed with several large-scale industries with a high inflow of effluents at Bhadravathi in Bhadra river (Shahnawaz et al., 2010). Similarly, Periyar river has about 15 impoundments that have affected the hydro flow regime, and the water quality index is reported to be poor, along with high loads of heavy metal pollutants (Abdu Rahiman et al., 2009; Mohan et al., 2019; Rajappan & Joseph, 2017). All these natural and anthropogenic stresses might lead to loss of ichthyofaunal diversity of river ecosystem. Hence, it is very important to know the exact causes, as most of the species of the genus are under threatened category of IUCN. Thus, for conservation of these species, critical riverine habitats of Western Ghats of India must be protected from anthropogenic activities.

Conclusion

The appearance of phenotypes with vertebral deformity as a result of reduced inter-vertebral space led to the excessive deepening of the body. This kind of deformity is rare in wild fish populations. The presence of robust deep-bodied phenotypes in four species from three different river systems of Western Ghats in a single genus is inquisitive. The slight genetic divergence in the least mutating gene in two species has made us think it is as an evolutionary trait induced anthropogenically or by natural selection. Also, the occurrence of these deformed phenotypes within the normal phenotypic fish population is confirmed. From these observations of monstrosities, it maybe concluded that anthropogenic induced changes in fish vertebrae and body exists rarely in fishes of the genus Hypselobarbus. We consider these occurrences as deformities caused due to several unknown anthropogenic or natural stresses.