Hybrids between Pseudoplatystoma corruscans and P . reticulatum ( Siluriformes : Pimelodidae ) previously reported in the Upper Paraná River are likely escapes from aquaculture farms : evidence from microsatellite markers

The production of hybrids of the ‘pintado’, Pseudoplatystoma corruscans (Spix & Agassiz, 1829) and ‘cachara’, Pseudoplatystoma reticulatum (Eigenmann & Eigenmann, 1889) in captivity has generated many concerns about the possibility of introduction of farmed hybrids into natural environments. In the last decade, hybrids between these species, known as ‘pintachara’ or ‘cachapinta’, were reported from different regions of the Upper Paraná River basin. Prospection of these hybrids is important in order to orient conservation programs for the species involved. Knowledge of the presence of these hybrids will direct conservation strategies towards prevention and/or mitigation of the effects of cross breeding in natural populations of P. corruscans (the native species of the genus) and farmed hybrids. In this study, surveyed the larval population using molecular tools to detect the presence and assess the origin (natural hybridization or escapes from fish farms) of hybrids in natural water bodies. Nine microsatellite markers were used to detect signals of hybridization and introgression of P. reticulatum in larvae and adults of P. corruscans in Upper Paraná River basin, between Itaipu Dam and Porto Primavera Dam. The specimens were sampled in the Upper Paraná channel and in tributaries where hybrids had been detected in the past, during two reproductive seasons. Despite of that, no sign of hybridization and introgression was found in the 171 larvae and 75 adults sampled, suggesting that the specimens detected in previous studies had originated from escapes of

Hybridization and introgression between different species is one of the greatest challenges faced by global biodiversity (RhymeR & SimbeRloff 1996, AllendoRf et al. 2001, olden et al. 2004).Hybrids are becoming more common in natural environments as a result of environmental degradation, introduction of foreign species, breeding of artificial interspecific hybrids for aquaculture, and shifts in the distribution of species associated with climate changes (RhymeR & SymbeRloff 1996, AllendoRf et al. 2001, ScRibneR et al. 2001, WAltheR et al. 2009).Hybridization can homogenize distinct populations and species, and cause a reduction in local adaptations and genetic diversity (olden et al. 2004).
The 'pintado', Pseudoplatystoma corruscans (Spix & Agassiz, 1829), is the single native species of Pseudoplatystoma from the Upper Paraná River basin (buitRAgo-SuáRez & buRR 2007).This species is the largest catfish and one of the most widely commercialized fish in this region (PetReRe-JR et al. 2002).'Pintado' populations have suffered negative effects from dams, fishery activities, and habitat loss (Welcomme 1985).In order to meet market demands in the face of decreasing populations, farming of the 'pintado' and its hybrid with the 'cachara', Pseudoplatystoma reticulatum (Eigenmann & Eigenmann, 1889), has increased in the region, (cARvAlho et al. 2013).With that, concerns about the possibility of escapes from aquaculture farms have also increased, since farmed hybrids 'pintachara' (crossbreed between female 'pintado' and male 'cachara') and 'cachapinta' (crossbreed between female 'cachara' and male 'pintado') can potentially hybridize with natural populations of P. corruscans (feRnAndeS et al. 2003, PoRto-foReSti et al. 2008, PRAdo et al. 2012a, hAShimoto et al. 2012)  Thus, in this study, we apply microsatellite markers to detect evidence of hybridization in larvae and adults of Pseudoplatystoma spp. in the Upper Paraná River basin, between Itaipu Dam and Porto Primavera Dam.We consider that the inclusion of larvae in prospection of hybrids in the natural environment is more informative than studies exclusively with adults, especially in migratory species.Larvae allows us to assess both the presence and the origin of hybrids.Because larvae of migratory fishes are passively carried in the water flow (nAkAtAni et al. 2001), using larvae to detect hybrid individuals may help determining specific tributaries and areas were escapes from fish farms and/ or reproductive areas of hybridization are located within a watershed.This information can contribute significantly with the development of conservation programs for specific rivers, maximizing their efficiency and minimizing costs.2009) to test for hybridization between the two species of Pseudoplatystoma.The 10 µl reactions contained 10x buffer, 1.5 mM of MgCl 2 , 0.2 mM of dNTP, 0.05 pmol of each primer, 0.3 U of Taq and 5 ng of DNA.The PCR conditions included an initial denaturation at 95°C for 3 min, 35 cycles at 95°C for 30 s for denaturation, 55°C (50°C for Pcor10) for 1 min to annealing and 72°C for 1 min for extension, and a cycle at 72°C for 1 h for final extension.Genotyping was carried out in an ABi 3130 sequencer (Applied Biosystems).
We analyzed the presence of null alleles and scoring errors using the software Micro-Checker 2.2.3 (vAn ooSteRhout et al. 2004).Hardy-Weinberg disequilibrium, linkage disequilibrium, diversity, and genetic differentiation analysis were done using the software Arlequin 3.5 (excofieR & liScheR 2010).Whenever necessary (multiple analyses), critical p-values were corrected using the B-Y correction (nARum 2006).Assignment tests were used to assign larvae to their respective pure species or hybrid cluster.Five runs of 5 million of generations (500 thousand of burn-in) and 1 < K < 5 were performed in the software Structure 2.3.1 (PRitchARd et al. 2000).The ad hoc method of evAnno et al. ( 2005), implemented on the online tool Structure Harvester (eARl & vonholdt 2012), was used to assess the most likelihood value of K. Individuals with 0.1 < q < 0.9 were considered hybrids (vähä & PRimmeR 2006).A run of 5 million of generations (500 thousand of burn-in) in the software NewHybrids 1.1 (AndeRSon & thomPSon 2002) was used to estimate the posterior probability of individuals belonging to the categories pure P. corruscans, pure P. reticulatum, hybrids F1 and F2, and both backcrosses (F1 with each pure population).
A total of 171 larvae of Pseudoplatystoma species, identified P. corruscans by morphological and DNA barcode methods (GenBank accession numbers KU220028-KU220190), were collected and genotyped for the 9 loci of microsatellites (Table 1).All larvae were in the pre-flexion or flexion stage.
The presence of null alleles, scoring errors, linkage, and Hardy-Weinberg disequilibrium were recurrent between populations (larvae and of P. corruscans and P. reticulatum) and among loci.Adults of P. corruscans and P. reticulatum presented different values of genetic diversity (0.61 and 0.49, respectively -Table 2) and numbers of alleles (mean of 9.8 and 6.6, respectively).Adults of P. corruscans presented 88 alleles of which 69 are private.Pseudoplatystoma reticulatum presented 54 alleles, of which 35 are private.The 104 private alleles of these species compose 84.6% of the 123 alleles found.Although all loci presented private alleles, only the loci Pcor07, Pcor10, Pcor21 and Pcor23 did not present overlap in allele range (Table 2).
Tests of genetic differentiation support disjunction between the adults of species according to AMOVA, Fst, and Rst analyses.A total of 74% and 38% of the total genetic variation is due to differences between the species according with Rst (Rst = 0.74, p = 0.00) and Fst methods (Fst = 0.38, p = 0.00), respectively.The locus-by-locus AMOVA supports this distinction: significant Fst values varied between 0.11 and 0.74, and Rst values between 0.48 and 0.99, except for the Rst analysis with the Pcor02 locus (Rst = 0.00, p = 1.00).
The assignment analysis supports disjunction of these species (Fig. 1).The analysis in Structure using the most probable number of groups is 2 (K = 2), according to the method of evAnno et al. ( 2005), indicates that 99.7% of the genotype of adults of P. corruscans and 99.7% of the genotype of the P. reticulatum   individuals belonged to their own independent clusters.All individuals presented more than 95% of its genotype relative to its species cluster.Likewise, all adults of each species belong to its pure species with probabilities higher than 99.7%, as suggested by the NewHybrid analysis.
The number of alleles of larvae was 12.0 and genetic diversity was 0.61, with no evidence of inbreeding (Table 1).The assignment test supported that 99.8% of the genotype of the sampled larvae belongs to P. corruscans; 98.2% of the total of larvae presented more than 99% of its genotype assigned as P. corruscans; and no larvae had more than 4% of its genotype associated to P. reticulatum (Fig. 1).The assignment test with NewHybrids supports these results and all larvae presented more than 99.7% of probability of belonging to P. corruscans.
Among the 75 adults of P. corruscans and the 171 larvae sampled in the Upper Paraná River basin, no individual presented mitochondrial DNA compatible with P. reticulatum nor evidence of hybridization and introgression in the nuclear DNA with this species.
Natural populations of P. corruscans and P. reticulatum occur in sympatry in some river basins (e.g., Paraguay River, Lower Paraná River, Uruguay River) but they present low level of natural hybridization (cARvAlho et al. 2013).This may be a consequence of differences in growth and body size at sexual maturity (ReSende et al. 1996(ReSende et al. , godinho 2007)) To assess the possibility of future natural hybridization between P. corruscans and P. reticulatum in the Upper Paraná River basin, we recommend systematic surveys on larvae of P. corruscans, using molecular markers as part of a monitoring program.Continuous or sporadic escapes from aquaculture farms represent a risk of introduction of hybrid specimens into natural waters by increasing propagule pressure (see SimbeRloff 2009 for details).Thus, monitoring is fundamental to preserve the natural populations of the 'pintado'.Monitoring should primarily focus on larvae to control for the origin of hybrids detected in the Upper Paraná River (e.g., tributaries contributing to escapes or areas of hybridization).Specific control campaigns may be directed to fish farms located in those tributaries.This approach is promising and can be used to prospect hybrids of other species along Neotropical basins.
ABSTRACT.The production of hybrids of the 'pintado', Pseudoplatystoma corruscans(Spix & Agassiz, 1829)  and 'cachara', Pseudoplatystoma reticulatum(Eigenmann & Eigenmann, 1889)  in captivity has generated many concerns about the possibility of introduction of farmed hybrids into natural environments.In the last decade, hybrids between these species, known as 'pintachara' or 'cachapinta', were reported from different regions of the Upper Paraná River basin.Prospection of these hybrids is important in order to orient conservation programs for the species involved.Knowledge of the presence of these hybrids will direct conservation strategies towards prevention and/or mitigation of the effects of cross breeding in natural populations of P. corruscans (the native species of the genus) and farmed hybrids.In this study, surveyed the larval population using molecular tools to detect the presence and assess the origin (natural hybridization or escapes from fish farms) of hybrids in natural water bodies.Nine microsatellite markers were used to detect signals of hybridization and introgression of P. reticulatum in larvae and adults of P. corruscans in Upper Paraná River basin, between Itaipu Dam and Porto Primavera Dam.The specimens were sampled in the Upper Paraná channel and in tributaries where hybrids had been detected in the past, during two reproductive seasons.Despite of that, no sign of hybridization and introgression was found in the 171 larvae and 75 adults sampled, suggesting that the specimens detected in previous studies had originated from escapes of aquaculture farms.KEY WORDS.Conservation genetics, fish larvae, interspecific hybrids, introgression, molecular markers.the 'cachara' is native to the Paraguay River and Lower Paraná River basin (buitRAgo-SuáRez & buRR 2007), adult specimens have been reported in the Upper Paraná River basin, introduced by aquaculture (vAini et al. 2014).Natural hybridization between the 'cachara' and P. corruscans is possible, since hybrids are fertile (PRAdo et al. 2012b), but highly undesirable for the maintenance of the genetic integrity of the natural local populations of P. corruscans (RhymeR & SimbeRloff 1996, AllendoRf et al. 2001).Hybrids between P. corruscans and P. reticulatum are often reported from the Aquidauana River (Paraguay River basin), the Mogi-Guaçu River and the Ivinheima River sub-basin (Upper Paraná River basin) (PRAdo et al. 2012a, vAini et al. 2014).Presently, they have been reported from the Upper Paraná River basin only sporadically between the Itaipu Dam and Porto Primavera Dam (E.A. RoSA, pers.comm., A.A. SilvA, pers.comm.).Genetic markers are important tools to detect the presence of hybrids (SAnz et al. 2009), and microsatellite markers have been used to detect hybrids between P. corruscans and P. reticulatum in the recent past (e.g., PRAdo et al. 2012a, cARvAlho et al. 2013).
Ichthyoplankton were collected from February 2012 to October 2015 (the reproductive season of Pseudoplatystoma spp.occurs between October and March, godinho et al. 2007) in the Upper Paraná River basin (between Itaipu and Porto Primavera Dam) and its main tributaries (Ivaí, Ivinheima, Ivinheiminha, Iguatemi, Piquiri and Amambai rivers) (Appendix 1) using a conical-cylindrical plankton net (500 µm mesh) in the water surface for a period of 10 minutes, at night.Samples were fixed in 70% ethanol.The larvae were identified preliminarily by morphological methods using nAkAtAni et al. (2001) and molecular methods using the DNA barcode method (hebeRt et al. 2003, see also cARvAlho et al. 2012).A database of pure species was created with 75 adults of P. corruscans collected at the same sites than the larvae (Ivaí, Ivinheima, Iguatemi, Piquiri and Amambai rivers) and 20 individuals of P. reticulatum were obtained from the Aquidauana River (Paraguay River basin) (Appendix 1).Specimens of P. corruscans and P. reticulatum from the Coleção Ictiológica do Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura (Nupélia) at the Universidade Estadual de Maringá, Maringá, Brazil (catalog number NUP 12798 and NUP 3587) were used as reference adult specimens.The larvae were photographed and their heads were deposited in the Coleção Ictiológica do Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura (Nupélia) (accession numbers NUP17754 to NUP17757); the remaining larvae were used in genetic analysis.Total DNA of the larvae were isolated with the DNeasy kit ® (Qiagen), and the DNA of adult specimens were extracted with the EZ DNA ® kit (Biological Industries).The amplification reaction of the mitochondrial gene cytochrome oxidase, subunit 1 (COI), was carried out in 25 µl PCR with 25x buffer, 3 mM of MgCl 2 , 0.4 uM of dNTP, 1 pmol of each primer (FF2d and FR1d, ivAnovA et al. 2007) and 20-40 ng of DNA.The PCR program included an initial denaturation at 94°C for 4 min, 35 cycles at 94°C for 4 min for denaturation, annealing at 58°C for 45 s, extension at 72°C for 1 min, and a cycle at 72°C for 5 min for final extension.Sequencing with the BigDye ® kit (Applied Biosystems) followed the manufacturer's protocol.The sequencing products were purified with Sephadex G-50 (GE) and sequenced with an ABi 3130 automatic sequencer (Applied Biosystems).Barcode identification of each larva was performed in BoldSystems v3 (RAtnASinghAm & hebeRt 2007).Larvae and the adults were genotyped for nine microsatellite loci (Pcor01, Pcor02, Pcor05, Pcor07, Pcor08, Pcor10, Pcor21, Pcor23, Pcor28) described by RevAldAveS et al. (2005) and PeReiRA et al. (

Figure 1 .
Figure 1.Genetic assigment of adults of Pseudoplatystoma corruscans and P. reticulatum and larvae sampled in the Upper Paraná River basin.Each column represent an individual and each color represent a species.
, as well as fidelity to the reproductive area, as proposed for P. corruscans by PeReiRA et al. (2009) (PRAdo et al. 2012a).Alternatively, elevated proportions of hybrids were reported in regions with high density of fish farms that produce these hybrids, such as the Mogi Guaçu River (50% -Paraná River basin), the Ivinheima River (61% -Paraná River basin), and the Aquidauana River (30.75% -Paraguay River basin) (PRAdo et al. 2012a, vAini et al. 2014).Most hybrids sampled by vAini et al. (2014) in the Upper Paraná River basin and Paraguay basin correspond to "cachapinta", which is also the most traded hybrid in fish farms (PoRto-foReSti et al. 2011, PRAdo et al. 2012a, vAini et al. 2014).Furthermore, hybrids have already been collected in low frequency (3.6%) in an upper stretch of the Upper Paraná River, close to Ilha Solteira Dam, between 2003 and 2008 (PRAdo et al. 2012a).During the period of these studies, A.B. Silva (pers.comm.) reported that hybrids were frequently caught in professional fisheries in the stretch of Upper Paraná River between Itaipu reservoir and Porto Primavera dam.Alternatively, our results indicate an absence of any sign of hybridization and introgression in the larvae and adults of the native population of P. corruscans from the Upper Parana River basin, including in the Ivinheima River population, the sub basin that presented hybrids in the study of vAini et al. (2014).Supported also by the currently sporadic catch of hybrids in this region (E.A. Rosa, pers.comm., A.B. Silva, pers.comm., vAini et al. 2014), we suggest that the hybrids captured there are most likely escapes from local fish farms, supporting the hypothesis of bignotto et al. (2009) and PRAdo et al. (2012a).

Table 2 .
Number of adults of P. corruscans and P. reticulatum sampled (N), gene diversity (H), inbreeding coefficient (Fis), number of alleles (N a ), range of alleles (R a ) and number of private alleles (P a ) for each of nine microsatellites loci.Bold number of Fis represent significant values and bold number of N a represent deviation of Hardy-Weinberg equilibrium.