Allozyme analysis of the four species of Hypostomus ( Teleostei : Loricariidae ) from the Ivaí river , upper Paraná river basin , Brazil

Allozyme electrophoresis analysis were performed in four species of Hypostomus (Loricariidae), H. albopunctatus, H. hermanni, H. regani, and Hypostomus sp. 1/NUP 5612 from the Ivaí river, a tributary of the upper Paraná river. The study of 14 loci revealed diagnostic characters and exclusive alleles in a low frequency. The heterozygosity ranged from 0.000 in H. albopunctatus to 0.199 in H. hermanni, which was higher than the heterozygosity in other samples of Hypostomus in literature, as well as in other fish groups. Hypostomus albopunctatus and H. regani revealed higher similarity (I = 0.804), while H. hermanni and Hypostomus sp. 1/NUP 5612 showed the least genetic identity (I = 0.569). All samples were genetically distinguished, despite there were several shared alleles. The FST value was 0.671, showing a high genetic differentiation among the samples. Hypostomus sp. 1/NUP 5612 was genetically distinguished from the three congeners by the loci Adh-A and G3pdh-B and by present rare exclusive alleles in other six enzymatic systems.


Introduction
The loricariid representatives of the genus Hypostomus are armored catfishes with a moderate to small and stout body, with nondepressed caudal peduncle, which usually supports an adipose fin (ARMBRUSTER, 2004).Hypostomus, with 128 valid species (ZAWADZKI et al., 2012), is distributed throughout Central and South America (FERRARIS JR, 2007).Several species of Hypostomus are morphologically very similar in shape and most of them present a high intraspecific variability in morphology and color pattern (OYAKAWA et al., 2005;ZAWADZKI et al., 2008a).Therefore, there are still many species with ill-defined specific status.Additionally, according to Weber (2003), based on molecular data, one-third of the current species of Hypostomus are still to be described.
These fishes are important for the aquarium trade where they are known as "plecos".They occur in a wide range of habitats, from still waters to fast running mountain streams.In greater water bodies they are usually found along lateral bank rivers with slower water current speed (BURGESS, 1989), but several species are also found in rapids.According to Suzuki et al. (2000) these fishes exhibit several reproductive strategies, mainly parental care as egg and larvae brooders or nest guarders, and its sucker-like mouth allow them to feed on algae and detritus.
Studies with allozymes have been carried out to assess the reproductive isolation and the genetic variability among populations of fish (LIMEIRA et al., 2009;PAIVA et al., 2005;PHILIPPSEN et al., 2009;RENESTO et al., 2007;ZAWADZKI et al., 2004a), as well as to reveal new species in high species-rich fish genus as it is the case of the Neotropical genera Hypostomus (ZAWADZKI et al., 2004b(ZAWADZKI et al., , 2008b) ) and Neoplecostomus (LUCENA et al., 2012;REUSING et al., 2011).
In samplings in the Ivaí river, a tributary on the left margin of upper Paraná river, southern Brazil, four samples of the genus Hypostomus, H. albopunctatus (Regan), H. hermanni (Ihering), H. regani (Ihering) and also specimens of a putative not described species assigned to as Hypostomus sp.1/NUP 5612 were caught.In the rivers and streams of the Neotropical region it is common to find two or more species of Hypostomus, which due to the aforementioned morphological variation are often difficult to be identified by ichthyologist.The main objective of this work is to verify if the allozyme data can be useful to genetically differentiate the four morphotype/species, as well as to estimate the level of genetic variability of the analyzed samples.

Material and methods
Specimens of Hypostomus sp.1/NUP 5612, H. albopunctatus, H. hermanni and H. regani (Figure 1) were collected in the Ivaí river (23º40'28"S, 52º07'09"W) (Figure 2 Samples of white skeletal muscle and liver were withdrawn from the fish, homogenized, centrifuged and then applied in horizontal starch gel (15%) according to detailed procedures in Zawadzki et al. (1999) and in polyacrylamide gel vertically (11%) (LAPENTA et al., 1995) in order to perform the electrophoresis.Thereafter, the gels were stained for nine enzymatic systems (Table 1), according to Zawadzki et al. (1999).The data were analyzed using POPGENE 1.31 software (YEH et al., 1999).The genetic interpretation of the enzymatic patterns was based on the quaternary structure of the enzymes described by Ward et al. (1992).

Results
The study of the four samples of Hypostomus from the Ivaí river, through nine enzymatic systems allowed the analysis of 14 loci, which presented a total of 40 alleles.Some diagnostic loci were detected among the species: Adh-A and G3pdh-B in Hypostomus sp.1/NUP 5612, sAat-B and Gcdh-A for H. hermanni and Est-2 for H. albopunctatus.An allele of locus sAat-A which expressed a band with a slower cathodic migration for H. albopunctatus also differentiated this species from H. regani (Table 2).
Variations in allelic frequencies also occurred among the samples of Hypostomus analyzed.The allele Est-2-d was diagnostic for H. albopunctatus, while the more frequent allele of Hypostomus sp.1/NUP 5612 (Est-2-c) and that of H. hermanni (Est-2-e) allowed the differentiation in most of their individuals from H. regani (allele Est-2-f).The same case was also verified for locus sMdh-B of Hypostomus sp.1/NUP 5612 in relation to the other species, as the most common allele (sMdh-B-c) was exclusive for this species.Furthermore, in Hypostomus sp.1/NUP 5612 some exclusive alleles were detected in low frequency (Table 2 All the loci in this study presented more than one allele in at least one species, except locus G3pdh-A, which presented the same fixed allele in all samples.The Esterase presented the highest number of alleles per locus.Locus Est-2 presented five different alleles in H. hermanni and three alleles in Hypostomus sp.1/NUP 5612.The genetic variability of the species analyzed presented values from 0.000 for H. albopunctatus to 0.199 for H. hermanni (Table 2).From the allele frequencies, the values of genetic identity (I) and genetic distance (D) were calculated (Table 3) and it was observed that Hypostomus sp.1/NUP 5612 and H. hermanni (D = 0.563) were the most genetically divergent.A dendrogram based on the genetic distance between the four samples is represented in Figure 3. Hypostomus albopunctatus and H. regani were the most similar samples analyzed (D = 0.218).The F ST value among the samples were 0.671.

Discussion
Hypostomus sp.1/NUP 5612 morphotype is a small to medium-lengthened species bearing pale spots over body and fins, large mandibles and large eyes.It is distributed throughout Ivaí river and several tributaries of this river and, apparently, is restricted to this system.The similar pale-spotted species inhabiting Ivaí river basin are H. albopunctatus, H. regani and H. strigaticeps.Hypostomus regani (Figure 1D) is the most likely to be morphologically confounded to Hypostomus sp.1/NUP 5612 (Figure 1A), mainly the young specimens.Hypostomus is traditionally known as having a complex taxonomy, which is mainly due to the incomplete descriptions and broad morphological variation of many of its species (OYAKAWA et al., 2005;WEBER, 2003).
The study of enzymatic loci has been used to discover or to confirm the existence of sibling species or species with dubious taxonomic status in sympatric populations of several organisms (THORPE; SOLÉ-CAVA, 1994) and in syntopic morphotypes of Hypostomus (ITO et al., 2009;ZAWADZKI et al., 2008b).In another study with allozyme data of the genus Hypostomus, Paiva et al. ( 2005) found the loci Acp-A, Gcdh-A and sMdhp-A as diagnostic for three species, H. strigaticeps, Hypostomus sp. 1 and Hypostomus sp. 2 from Maringá stream, a tributary to Pirapó river, Paranapanema river basin.Furthermore, 12 alleles also served as diagnostic for Hypostomus sp. 2 in that work.The allele Adh-A-a was diagnostic for Hypostomus sp. 1 as it clearly distinguished the two morphotypes from each other and from H. strigaticeps (PAIVA et al., 2005).
In the present work, besides the diagnostic loci Adh-A and G3pdh-B in Hypostomus sp.1/NUP 5612, changes in allelic frequencies also revealed a possible differentiation between the species analyzed.The alleles Est-2-c were the most frequent at Est-2 locus and exclusive for Hypostomus sp.1/NUP 5612 and Est-2-f was the most frequent and exclusive for H. hermanni.Est-2-d was fixed and exclusive for H. albopunctatus.The sMdh-B-c also allowed the differentiation of most specimens of Hypostomus sp.1/NUP 5612 from the other three species.Besides, we verified exclusive alleles in low frequencies in H. hermanni: sAat-A-c, Est-2-g, mMdh-A-a and Sod-A-a (Table 2).
In Hypostomus sp.1/NUP5612, the exclusive alleles sAat-B-a, Acp-A-a, Adh-A-b, sMdh-A-b and sMdh-B-a were recorded with low frequencies (Table 2).The low frequency of some exclusive alleles, together with the fact that part of these loci are not in Hardy-Weinberg equilibrium in these species, could characterize a population which is in the process of stabilization after some undetermined disturbing event (eg.drastic drought, pollution or local overfishing).
According to Mayr (1963), new genes are incorporated in populations in heterozygous condition, and the homozygotes usually only become frequent after a long time.Three loci with alleles in similar frequencies were found in H. derbyi from Iguaçu river, GDH-1*-A (0.036), GPI-2*-C (0.071) and IDHP-1*-C (0.036) (ZAWADZKI et al., 1999) and the alleles Gpi-B-a (0.031), mMdh-A-b (0.063) and sMdh-B-a (0.031) in Hypostomus sp. 3 from Keller stream (ZAWADZKI et al., 2004b).According to Thorpe and Solé-Cava (1994), in sympatric populations, significant differences in any locus between two morphs represent a barrier to gene flow and, at least, partial reproductive isolation.In organisms with sexual reproduction by cross fertilization, this variation indicates that the two populations should be considered as different species.
In H. albopunctatus, all the analyzed loci were monomorphic (Table 2) and consequentially, presented a genetic variability with zero value (He = 0.000).Probably, inbreeding due to its sedentary habits, could explain the absence of heterozygotes for this sample.Zawadzki et al. (1999) found the value of He = 0.011 for H. derbyi and 0.017 for H. myersi of Iguaçu river and suggested that inbreeding could have contributed for the maintenance of low levels of heterozygosity.In Hypostomus sp. 2 from Maringá stream (PAIVA et al., 2005) and Neoplecostomus paranensis of Hortelã stream (ZAWADZKI et al., 2004a), both species belonging to the family Loricariidae, all loci were also monomorphic, that is, He equal to zero.According to the authors, inbreeding would be acting in these species, due to a probable isolation by some geographical barriers in these small tributaries.
However, genetic drift events as bottleneck and founder effect can not be rejected.Increasing the analyzed specimens number of each of these aforementioned homozygote populations could probably reveal some rare heterozygote alleles, although it would not be expected to considerably change the allelic frequency levels.
A comparison among the values of average heterozygosity for 84 species of tropical fish, in relation to reproductive strategies (LASSALA; RENESTO, 2007), showed that the group sedentary or short distance migrants with parental care (as is also the case of the species of Hypostomus) had the smallest average heterozygosity (0.046), followed by migratory species of long distance and without parental care (0.064).The highest mean He (0.081) was for the sedentary or short distance migrants species and without parental care group.These results indicate that the parental care could be associated with low values of heterozygosity.According to these authors, species exhibiting parental care tend to be less variable.In species without parental care, few offspring are likely to survive, and a greater genetic variability is important to face environmental challenges.
The genetic variability in H. albopunctatus from the Itaipu reservoir (ZAWADZKI et al., 2005), was slightly higher (He = 0.031) than those found in this work for H. albopunctatus (He = 0.000).Also, the total mean values obtained herein for H. albopunctatus are still low when compared to the value of average heterozygosity for 195 species of fish from several regions of the world (WARD et al., 1992) that was He = 0.051, or with the average value of heterozygosity (He = 0.046) for 49 species of freshwater fish by Ward et al. (1994).
The values of He for Hypostomus sp.1/NUP 5612 (0.126), H. hermanni (0.199) and H. regani (0.085) in this work, different from H. albopunctatus, were considerably higher than the average found by Ward et al. (1992Ward et al. ( , 1994)).The value of He in H. regani from the Corumbá reservoir (0.056) (ZAWADZKI et al., 2008c) were slightly above the average of Ward et al. (1992Ward et al. ( , 1994)), while the population of H. regani from the Itaipu reservoir presented He value higher than the average, 0.078 (ZAWADZKI et al., 2008b).These H. regani populations were previously isolated by several waterfalls along the Paraná river and, currently by several reservoirs.Thus, they would have suffered different evolutionary rates which could have occasioned the differences in the heterozygosity values.
The expected heterozygosity of Hypostomus sp.1/NUP 5612 (0.126) may be similar to those obtained by Zawadzki et al. (2005) for H. margaritifer (0.106) and Hypostomus sp. 1 (0.107) from the Itaipu reservoir, or even, with Hypostomus sp. 1 (0.143) from the Keller river (ZAWADZKI et al., 2004b).However, none of these results are close to the genetic variability reported H. hermanni in this work (0.199), which seems to be one of the greatest values of He ever found for a species of this genus.According to Ward et al. (1994), the estimated values of heterozygosity expected from an average of 107 species of fish, varies from zero to 0.05 for 54% of the species, from 0.05 to 0.10 for 30%, from 0.10 to 0.15 for 12% and exceeding 0.15 there was only 4% of the species analyzed by these authors.In the Corumbá reservoir, Zawadzki et al. (2008b) also verified a significant variation in heterozygosity among ten species of Hypostomus analyzed, ranging from 0.009 in H. iheringii to 0.099 in Hypostomus sp. 4.These authors reported that the different He values revealed a lack of uniformity in the patterns of genetic variability among species of this genus.Although Hypostomus sp.1/NUP 5612 presented exclusive alleles and the same percentage of polymorphic loci of that H. hermanni, the highest genetic variability was found in this last species due to a greater number of alleles per locus (1.928) (Table 2) as, for example, in loci Est-2 and Sod-A.
From the allele frequencies the values of identity (I) and genetic distance (D) were calculated (Table 3) and it was observed that Hypostomus sp.1/NUP 5612 and H. hermanni (D = 0.563) were the most genetically divergent.The genetic distance (D) of Nei (1978) estimates the average number of nucleotide substitutions per locus, detectable by electrophoresis and accumulated in populations since they diverged from a common ancestral, that is, the distance should be proportional to the evolutionary time (DOBZHANSKY et al., 1977;THORPE;SOLÉ-CAVA, 1994).Contrary to the genetic distance, the genetic identity (I) represents the proportion of the products of the genes that are not differentiated by electrophoretic procedure (DOBZHANSKY et al., 1977) and its value varies from 0 to 1.According to Thorpe and Solé-Cava (1994), 85% of the values of I between species of the same genus exceed 0.35 and 97% of the values are below 0.85.Among species of different genus 77% of the values are less than 0.35, while for 98% of populations of the same species exceeds 0.85.These values when compared to the similarity between Hypostomus sp.1/NUP 5612 and H. hermanni (I = 0.569), between Hypostomus sp.1/NUP 5612 and H. albopunctatus (I = 0.711) and between Hypostomus sp.1/NUP 5612 and H. regani (I = 0.768) corroborate the distinction among this morphotype and the three nominal species.Additionally, the differentiation is supported by the high F ST value (0.6707) among the analyzed samples.According to allozyme data, F ST values higher than 0.15 are considered indicative of significant differentiation among populations (FRANKHAM et al., 2008).
The genetic distance among the species studied in the present work resulted in a dendrogram (Figure 3) which showed H. albopunctatus and H. regani as the most similar (D = 0.218).An analysis of the same both species, but from the Itaipu reservoir by Zawadzki et al. (2005), presented practically the same distance between them (D = 0.219), despite a greater number of enzymatic systems ( 14) and loci number (25) that have been analyzed in that study.Zawadzki et al. (2005) also found that H. albopunctatus and H. regani from Itaipu reservoir were genetically similar (I = 0.803), as well as H. albopunctatus and H. regani from the Ivaí river in the present work (I = 0.804).This data corroborates the statement of Thorpe and Solé-Cava (1994), that there is a clear relationship between taxonomic divergence and genetic distance.

Conclusion
These results highlight the usefulness of allozyme data to obtain genetic markers to Neotropical fish, as well as to inferences on the heterozygosity and evolutionary genetics among species of Hypostomus.The allozyme data, therefore, indicate herein that the four species of Hypostomus from the Ivaí river present biochemical diagnostic markers which allow to genetically differentiate them.
), a tributary of the left margin of the upper Paraná river, State of Paraná, Brazil.Voucher specimens were deposited in the collection of fish of the Nupélia (Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura) from the Universidade Estadual de Maringá: NUP 5614 for H. albopunctatus; NUP 4842 for H. hermanni; NUP 4360 for H. regani; and NUP 5612 for Hypostomus sp.1/NUP 5612.

Figure 2 .
Figure 2. Partial map of Paraná State showing the collecting point (black dot) in the Ivaí river.

Figure 3 .
Figure 3. UPGMA dendrogram from the similarity indices of Nei (1978) for the four Hypostomus population from the Ivaí river.