Introduction

Habitat alterations caused by water flow control induced by dams, for example, disrupt the structure of the aquatic biota and facilitate the establishment of invasive species (Havel, Lee, & Zanden, 2005). The installation of fish passages in dams, with the aim to preserve migratory species, may pose an additional threat to the biota from upstream stretches by increasing the dispersion of non-native species. This may occur when the dam separates distinct native faunas along the watershed, as observed in the Paraná River (Agostinho et al., 2015). In the case of Paraná River, the Itaipu Reservoir, filled in 1982, covered the Sete Quedas Falls, which separated distinct ichthyofauna provinces. As a consequence, 33 species of native fishes in the lower Paraná basin successfully colonized the upper Paraná River (Julio Junior, Dei Tós, Agostinho, & Pavanelli 2009). In 2002, a channel for fish migration named “Canal de Piracema” was opened, allowing 28 of the 68 fish species registered downstream of the dam to rise towards the Itaipu Reservoir (Agostinho, Gomes, Fernandez, & Suzuki, 2002).

Hemiodus orthonops is a small fish of the Hemiodontidae family, order Characiformes, with a maximum of 25 cm standard length (Froese & Pauly, 2015), described by Eigenmann & Kennedy (1903) from specimens collected in the Paraguay River, near the town of Asuncion.

Until recently, H. orthonops was an endemic species from the Paraná-Paraguay basin and it was absent from the upper Paraná River basin (Agostinho et al., 2015). Since 2008, it has started to be collected in the upper Paraná River, representing up to 10% of catches (Agostinho et al., 2015).

In the Paraná River population studied by Agostinho et al. (2015), the maximum standard length observed for H. orthonops was 26 cm for males and 29 cm for females; individuals become sexually mature at one year of age when they reach 15.2 cm of length for males and 19.4 cm for females, releasing a maximum amount of 79,653 oocytes with 0.79 mm diameter each and feeding primarily on detritus and algae.

According to Agostinho et al. (2015), the success of H. orthonops at colonizing the upper part of Paraná River was due to dispersal ability, favorable environmental conditions, the ability to exploit highly available food resources, early maturation and high body growth rate.

Invasive species are predicted to suffer from reductions in genetic diversity when only a few individuals from the source population colonize new habitats, thus undergoing the founder effect and reducing adaptive potential (Mayr, 1963; Allendorf & Luikart, 2007).

The founder effect on genetic variability of experimental populations of mosquito fish were measured by the allozymes electrophoretic technique by Leberg (1992), who estimated the values of proportion of polymorphic loci, number of alleles per locus and heterozygosity. Compared to source populations, Leberg found that 23 populations had a smaller He and 15 had an equal or greater He. On average, all three indexes of genetic variability decreased as the effective size of the founding population decreased.

Roman and Darling (2007) reviewed data of 43 introduced aquatic species and found that only 16 (37%) showed clear evidence of significant loss of genetic diversity relative to native populations, while in 27 species there was no alteration.

Dlugosch and Parker (2008) analyzed data from 80 species in the literature covering 18 plants, 2 fungi, and 60 animals (including 7 birds, 6 reptiles, 8 fish, 3 amphibians, 8 mammals, 13 insects, 4 crustaceans, 6 mollusks, 3 annelids, 1 cnidarian, and 1 tunicate) and showed that, in introduced populations, there was an average reduction of 15.5% in allelic richness and a decrease of 18.7% in average heterozygosity compared to the original population.

However, other studies have shown that many invasive species did not lose the variability and in some of them it was even higher. Bossdorf et al., (2005) revealed that 69% of invasive plants had the same or greater genetic diversity than native populations. A similar study of 29 terrestrial and aquatic animal species showed that introduced species have preserved 80% of the genetic variability existent in native populations (Wares, Hughes, & Grosberg, 2005).

Thus, in this study we aimed to estimate the genetic variability of Hemiodus orthonops using the allozymes electrophoresis technique and compare it to other species from Paraná River basin, which was also studied with allozymes electrophoretic technique. Furthermore, we verified the presence of a founder effect in the invasive population of this migratory species in the Paraná River.

Material and methods

Hemiodus orthonops individuals were collected in the upper Parana River basin, area of Porto Camargo (23°21’27.87”S, 53°44'54.14"W), municipality of Icaraíma and in Porto Figueira (23°23'43.93"S, 53°48'47.79"W), municipality of Alto Paraíso (Figure 1) since March, 2012 to April, 2013.

Fifteen specimens were captured in the first mentioned municipality, and 13 in the second one. Voucher specimens were deposited in the collection of Nupelia (Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura) of the Universidade Estadual de Maringá NUP 10610. All specimens were frozen and stored at low temperature (-20° C) until the extraction of enzymes. The enzymes were extracted from samples of muscle, liver and heart tissue, using buffer Tris-HCl, 0.02 M, pH 7.5. The samples were centrifuged in a refrigerated centrifuge at 4° C and 16,434.6 g. The protein extract was applied on corn starch gel (Val, Schwantes, Schwantes, & De Luca, 1981) Penetrose 50 at a concentration of 17%, with small filter paper strips (4 mm x 8 mm) Whatman 3 MM^® soaked with the samples. Subsequently, the gel was submitted to a continuous horizontal electrophoresis under refrigeration. Three buffer solutions were used: 0.135 M Tris/Citric Acid 0.043 M, pH 5.9 (TC), diluted 15 times in the preparation of the gel; 0.1 M Tris/Maleic Acid 0.1 M/EDTA 0.004M/Magnesium chloride 0.01M, pH 7.4 (TEM) in tank and diluted 10 times in the gel, and Tris 0.18M/Boric acid 0.1M/EDTA 0.004M, pH 8.6 (TBE), diluted four times in the gel preparation. The polyacrylamide gel, applied for the revelation of Esterase enzyme, was prepared in two glass plates, consisting of two systems of gel with different porosities and pH. The concentration of the separating gel was of 10% of acrylamide-Bisacrylamide and the stacking gel was of 5%.

Muscle samples were prepared as described in the starch gel protocol, however, the buffer employed to macerate was 1.5 M Tris-HCl, pH 8.8 with 10% glycerol, and the samples were centrifuged for 20 minutes at 4°C to 16,434.6 g. Ten enzyme systems in starch gel were analyzed: Aspartate amino Transferase (AAT), Alcohol Dehydrogenase (ADH), Glucose-3-phosphate dehydrogenase (G3PD), Glucose-6-phosphate dehydrogenase (G6PD), Glucose phosphate Isomerase (GPI), Isocitrate Dehydrogenase (IDH), Lactate dehydrogenase (LDH), Malate dehydrogenase (MDH) Malic Enzyme (ME), Phosphoglucomutase (PGM), and Esterase (EST) in the polyacrylamide gel. The nomenclature of enzymes used was previously proposed by Murphy, Sites, Buth, and Haufler (1996). The genetic interpretation of zymograms was based on the quaternary structure of enzymes according to Ward, Skibinski, and Woodwark (1992).

Data were analyzed in the Popgene software 3.1 (Yeh, Yang, Boyle, Ye, & Mao, 1997). The genetic variability was estimated by the calculation of the observed and expected heterozygosity (He and Ho) according to Nei (1978). Considering the values of allele frequencies, the identity (I) and the genetic distance (D) from Nei (1978) were evaluated. To compare the genetic variability of H. orthonops with other species of the Paraná River basin, the authors calculated the mean of 74 species distributed in 12 families (49 Loricariidae, three Callichthyidae, five Pimelodidae, one Cynodontidae, one Erythrinidae, six Anostomidae, one Prochilodontidae, one Auchenipeteridae, three Gymnotidae, one Serrasalmidae, three Characidae and one Cichlidae).

Results

In the Porto Camargo sample (PC), 27 alleles distributed in 21 loci were detected in H. orthonops, six loci of which were polymorphic (Adh, G6pdh, Ldh-1, Est-1, Est-2, Est-3); in Porto Figueira sample (PF), 25 alleles were detected in 21 loci, of which three were polymorphic (Adh, Est-1, Est-3).

Table 1 illustrates the allele frequencies of each locus of each sample. The Aspartate enzyme Amino Transferase (AAT) expressed two loci: Aat-1, which codifies cathodic enzymes and Aat-2 that codifies anodyne enzymes. The loci Aat-1 and Aat-2 were monomorphic for the two population samples.

The ADH enzyme expressed one locus with two alleles for the two samples in Hardy-Weinberg equilibrium. The G3PDH enzymes were encoded by two loci, being both loci G3pdh-1 and G3pdh-2 monomorphic. The G6PDH enzyme presented one polymorphic locus for H. orthonops in the PC region, being in Hardy-Weinberg equilibrium and monomorphic for PF region. Two loci for the GPI enzyme were detected, both Gpi-1 and Gpi-2 monomorphic for both regions. The IDH enzyme showed two loci Idh-1 and Idh-2, which were both monomorphic. Two loci were seen for LDH: Ldh-1 and Ldh-2. Ldh-1 was monomorphic for the PF sample and polymorphic with two alleles in Hardy-Weinberg equilibrium for the PC sample. The MDH enzyme expressed three loci, all monomorphic in both samples. Two loci for the ME enzyme were detected, named Me-1 and Me-2, which showed no variation. For the PGM enzyme, a monomorphic locus Pgm-1 was found for both samples.

Table 2 exhibits a summary of the genetic variability results for both population samples. Through analysis of this data, it was possible to verify that the H. orthonops sample from the PC region had a polymorphic loci proportion of 26.09%, of which six loci were polymorphic and presented a ratio of 13.04% for the PF sample with three polymorphic loci. The mean number of allele per locus was 1.3 and 1.2 for PC and PF, respectively.

In the present study, the average Ho and He were of 0.0668 and 0.1017 for samples from PC, and 0.033 and 0.058 for samples from PF. The fixation index (F) was 0.34 for PC and 0.41 for PF, that is, there was an excess of 34% of homozygotes in the PC sample and 41% in the PF sample.

In order to verify the inbreeding level and the population structure, Wright’s F statistics (1965), FIS, FIT and FST were calculated for the two H. orthonops population samples (Table 3). FIS measures the mean excess or deficiency of heterozygotes in the two populations and showed that are an excess of 35.8% of homozygotes. The FIT (0.4917) coefficient showed that if the two populations were united in one there would be an excess of 49.17% of homozygotes. The FST value (0.2081) represents a differentiation measure of the two populations. In other words, 20.81% of the expected heterozygosity was due to the separation between the populations. All F ratios were statistically significant (Table 3). Bringing the two samples together in a single population, we obtained the values of HO = 0.0497, HT = 0.0979 and HS = 0.0791. The identity and the genetic distance between them were estimated according to Nei (1978) in I = 0.9558 and D = 0.0452.

Discussion

The data obtained in this work showed that H. orthonops has little mean number of alleles by locus (1.29 for PC sample and 1.19 for PF sample) and revealed that there is an excess of homozygotes in the two samples. Based on data of literature, the authors calculated the genetic variability for 74 species from Paraná River that were also studied with the allozymes electrophoresis technique. The result was a mean of 21.5 loci, 24.4% of which were polymorphic, 1.3 allele per locus and values of Ho and He of 0.037 and 0.076, respectively. In this work, we found mean values of Ho and He of 0.05 and 0.10 respectively (Table 2), 26.09% of polymorphic loci and 1.3 of alleles per locus. Therefore H. orthonops has greater genetic variability than the mean of 74 species from Paraná River basin and greater than the mean of 195 fish species estimated (He = 0.051) by Ward et al. (1992).

FST values show that, despite the short distance separating the two sampled populations (7 Km), they are genetically structured. According to Wright (1978), FST values ranging from 0.0 to 0.05 may indicate slight genetic differentiation; from 0.05 to 0.15 indicates moderate differentiation; from 0.15 to 0.25 indicates large differentiation and over 0.25 indicates a very large differentiation. Therefore, it can be assumed that the value obtained (0.2081; χ2 = 10.82; P<0.01)) demonstrates a large differentiation among the two populations.

The FIS coefficient measures the heterozygosity reduction due to non-random mating within a subpopulation. Observing Table 3, a significant FIS value (0.3581; χ2 = 9.31; P<0.01) can be noticed, indicating an excess of homozygotes. Moreover, the average FIT value (0.4917; χ2 = 25.57; P<0.01) denotes an excess of homozygotes if the two populations were grouped in only one. Such high excess of homozygotes may be owing to a considerable level of consanguinity or to a founder effect due to the recent colonization of the sampled environment. The results obtained by Agostinho et al., (2015) points that the number of captured individuals in the samplings increased from 8 in 2008 to 753 in 2013, which means that this species multiplied 94 times in five years. Therefore, the population dwelling in the Paraná River should be numerous, avoiding consanguinity. In accordance with such data, it is likely that the excess of homozygotes is due to founder effect of the population that had multiplied quickly.

We believe that the high variability is an intrinsic feature of the species, more related to high fertility and population size. To support such hypothesis, there are the results obtained by Ribeiro, Morán, and Caballero (2008) in Atlantic salmon (Salmo salar). The mentioned authors found that reducing the salmon population size from the Eo River, Spain, did not reduce the level of genetic diversity.

Allelic richness is generally predicted to be more sensitive to founder effects than heterozygosity is (Nei, Maruyama, & Chakraborty, 1975; Allendorf 1986; Leberg 1992). For studies that reported both metrics, a paired comparison showed that proportional losses of allelic richness were on average 5.1% more severe (more negative) than losses of heterozygosity in Hypericum canariense (Dlugosch & Parker, 2008). This difference is expected because allelic richness will reflect the loss of rare alleles that contributed little to heterozygosity (Dlugosch & Parker, 2008). In this study, we had not analyzed samples of H. orthonops from localities downstream Itaipu Reservoir to verify if the values of He, proportion of polymorphic loci and mean number of alleles were reduced or not. The fact of these values in the sample of PF is lower than that of PC is suggestive that the colonization of the upper Parana River basin by H. orthonops had reduced those values.

Founder effect was detected in natural populations of various species (Dlugosch & Parker, 2008) and in experimental populations of the mosquito fish Gambusia Holbrooki (Leberg, 1992). It was not yet evaluated if the species that migrated through Canal de Piracema passed through a bottleneck due to founder effect. This artificial corridor may contribute to founder effects in other fish species that are capable to transpose these barriers and colonize new environments upstream as did H. orthonops.

To support or refute these findings, it would be interesting to conduct genetic analysis with DNA markers in both populations here analyzed and in the Paraguay River populations, where the species is native.

Conclusion

We conclude that the population of Hemiodus orthonops dwelling the upper Paraná River basin is genetically structured and they show homozygote excess probably due to founder effect.

Acknowledgements

The Authors would like to thank Henrique Bueno Ruiz and Rafael Vieira for helping with field work; to Rafaela Granzotti for helping with the language; to Corn Products Brazil for providing Penetrose 50.

References

Agostinho, A. A., Suzuki, H. I., Fugi, R., Alves, D. C., Tonella, L. H., &. Espindola, L. A. (2015). Ecological and life history traits of Hemiodus orthonops in the invasion process: looking for clues at home. Hydrobiologia, 746(1), 415-430.

Agostinho, A. A., Gomes, L. C., Fernandez, D. R., & Suzuki, H. I. (2002). Efficiency of fish ladders for neotropical ichthyofauna. River Research and Applications, 18(3), 299-306.

Allendorf F. W. (1986) Genetic drift and the loss of alleles versus heterozygosity. Zoo Biology, 5(2), 181-190.

Allendorf, F. W., & Luikart, G. (2007). Conservation and the genetics of populations. Malden, MA, Blackwell publishing.

Bossdorf, O., Auge, H., Lafuma, L., Rogers, W. E., Siemann, E., & Prati, D. (2005). Phenotypic and genetic differentiation between native and introduced plant populations. Oecologia, 144(1), 1-11.

Dlugosch, K. M., & Parker, I. M. (2008). Founding events in species invasions: genetic variation, adaptive evolution, and the role of multiple introductions. Molecular Ecology, 17(1), 431-449.

Eigenmann, C., H., & Kennedy, C. H. (1903). On a collection of fishes from Paraguay, with a synopsis of the american genera of Cichlids. Proceedings of the Academy of Natural Sciences of Philadelphia, 55, 497-537.

Froese, R., & Pauly, D. (2015). FishBase. World Wide Web electronic publication. Retrieved from http://fishbase. org/summary/55704.

Havel, J. E., Lee, C. E., & Zanden, J. M. V. (2005). Do reservoirs facilitate invasions into landscapes? BioScience, 55(6), 518-525.

Julio Junior, H. F., Dei Tós, C., Agostinho, A. A., & Pavanelli, C. S. (2009). A massive invasion of fish species after eliminating a natural barrier in the upper Parana basin. Neotropical Ichthyology, 7(9), 709-718.

Leberg, P. L. (1992). Effects of population bottlenecks on genetic diversity as measured by allozyme electrophoresis. Evolution, 46(2), 477-494.

Mayr, E. (1963). Animal species and evolution. Cambridge, MA: The Belknap Press of Harvard University Press.

Murphy, R. W., Sites Jr., J. W., Buth, D. G., & Haufler, C. H. (1996). Proteins: Isozyme electrophoresis. In D. M. Hillis, C. Moritz, & B. K. Mable, (Eds.), Molecular Systematics (2nd ed., p. 51-120). Sunderland, MA: Sinauer Associates.

Nei, M. (1978). Estimation of average of heterozygosity and genetic distance from small number of individuals. Genetics, 89(3), 583-590.

Nei M., Maruyama T., & Chakraborty, R. (1975). The bottleneck effect and genetic variability in populations. Evolution, 29(1), 1-10.

Ribeiro, Â., Morán, P., & Caballero, A. (2008). Genetic diversity and effective size of the Atlantic salmon Salmo salar L. inhabiting the River Eo (Spain) following a stock collapse. Journal of Fish Biology, 72(8), 1933-1944.

Roman, J., & Darling, J. A. (2007). Paradox lost: genetic diversity and the success of aquatic invasions. Trends in Ecology and Evolution, 22(9), 454-464.

Val, A. L., Schwantes, A. R., Schwantes, M. L. B., & De Luca, P. H. (1981). Amido hidrolisado de milho como suporte eletroforético. Ciência e Cultura, 33, 737-741.

Ward, D. R., Skibinski, D. O. F., & Woodwark, M. (1992). Protein heterozygosity, protein structure, and taxonomic differentiation. Evolutionary Biology, 26, 73-159.

Wares, J. P., Hughes, A. R., & Grosberg, R. K. (2005) Mechanisms that drive evolutionary change: insights from species introductions and invasions. In D. F. Sax, J. J. Stachowicz, & S. D. Gaines, (Eds.), Species invasions: Insights into ecology, evolution, and biogeography (p. 229–257). Sunderland, MA: Sinauer Associates.

Wright, S. (1965) The interpretation of population structure by F-statistics with special regards to systems of mating. Evolution, 19(3), 395-420.

Wright, S. (1978). Evolution and the genetics of populations. Variability within and among natural populations (Vol. IV). Chicago, IL: University of Chicago Press.

Yeh, F. C., Yang, R. C., Boyle, T. J. B., Ye, Z. H., & Mao, J. X. (1997) POPGENE, the user - friendly shareware for population genetic analysis. Edmonton, CA: Molecular Biology and Biotechnology Centre, University of Alberta Canada. Retrieved from https://www. ualberta.ca/~fyeh/popgene_download.html