DNA Barcodes 2015; Volume 3: 118–128

Research Article Open Access

Jefferson Monteiro Henriques, Guilherme José da Costa Silva, Fernando Yuldi Ashikaga, Robert Hanner, Fausto Foresti and Claudio Oliveira Use of DNA barcode in the identification of fish species from Ribeira de Iguape Basin and coastal rivers from São Paulo State (Brazil)

DOI 10.1515/dna-2015-0015 Received February 25, 2015; accepted July 20, 2015 1 Introduction Abstract: Species identification is a difficult task, ranging Biodiversity is threatened by problems arising from from the definition of the species concept itself to the human population growth [1,2], and species identification definition of the threshold for speciation. DNA Barcode is an important step for the preservation of biodiversity technology uses a fragment of the Cytochrome Oxidase [2,3]. However, species identification is a difficult task, I (COI) gene as a molecular tool that many studies have since many controversies have persisted to the present already validated as a tool for species identification. DNA day, ranging from the definition of the species concept barcode sequences for COI were generated and analyzed to the definition of the threshold for speciation [4]. Many from 805 specimens. The General Mixed Yule Coalescent types of methods have been proposed as the best tools for (GMYC) analysis recognized 99 independent evolution units, species identification, with genetic tools considered as an and the Barcode Index Numbers (BIN) approach pointed to alternative approach [5–12]. the existence of 104 BINs (interpreted as distinct species). By The integration of research areas for species cross-tabulating the results of all approaches, we identified identification can bring great benefits in biodiversity 109 Molecular Operational Taxonomic Units (MOTU) by characterization in regions where such information is at least one methodology. In most cases (89 MOTUs), the lacking, avoiding the loss of endemic species that may genetic approaches are in agreement with morphological perhaps occur in these regions, enabling the creation of identification, and the discrepant results of MOTUs are conservation areas. Examples of this characterization in the complex groups, which have many morphological include the checklist of ichthyofauna from the Ribeira de similarities but may represent species complexes. Iguape River Basin [13–15] and the Coastal Rivers of São Paulo [13,16]. These two water bodies are located in two Keywords: Single locus analysis, species identification of the most urbanized and densely populated regions of and delimitation, COI gene Brazil, but data on the number of freshwater fish species in both regions are incomplete. *Corresponding author: Jefferson Monteiro Henriques, Departamento Although molecular tools have provided new insights de Morfologia - Instituto de Biocięncias Universidade Estadual into several areas of study, such as evolutionary biology, Paulista - UNESP Botucatu 18618-970 Săo Paulo Brazil phylogenetic systematics, and population genetics, only Guilherme José Costa Silva, Departamento de Morfologia - Instituto in 2003 it was proposed that a short segment of 648 de Biociências Universidade Estadual Paulista - UNESP Botucatu 18618-970 São Paulo Brazil nucleotides from the mitochondrial gene cytochrome c Fernando Yuldi Ashikaga, Departamento de Morfologia - Instituto de oxidase I (COI) could standardize species identification Biociências Universidade Estadual Paulista - UNESP Botucatu 18618- [6,7]. The methodology called “DNA Barcode” 970 São Paulo Brazil subsequently has been widely used, and many studies Robert Hanner, Biodiversity Institute of Ontario and Department of Inte- have already validated it as a tool for species identification grative Biology University of Guelph Guelph N1G 2W1 Ontario Canada Fausto Foresti, Departamento de Morfologia - Instituto de Biociênci- in selected geographical areas [17–28]. as Universidade Estadual Paulista - UNESP Botucatu 18618-970 São However, many questions have been raised about the Paulo Brazil use of genetic markers for species delimitation. Population Claudio Oliveira, Departamento de Morfologia - Instituto de Biociên- genetic parameters such as gene flow and population size cias Universidade Estadual Paulista - UNESP Botucatu 18618-970 and phylogenetic parameters such as the pattern and São Paulo Brazil

© 2015 Jefferson Monteiro Henriques et al. licensee De Gruyter Open. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License. Unauthenticated Download Date | 12/4/15 10:38 PM DNA barcode fish Ribeira de Iguape coastal rivers Brazil 119 timing of lineage diversification are being very actively 2 Methods studied, because the limit of genetic divergence within and between species is a factor that must be considered [25,29– 2.1 Sampling 32]. “Phylogenetic analysis is traditionally concerned with estimating relationships among species, while For this study, 805 specimens of freshwater fishes (89 population genetics is concerned with adaptation and morphospecies) were used, including 490 samples from understanding the population-level forces that change the Ribeira de Iguape Basin and 315 from Coastal Rivers allele frequencies” Carstens et al.[4]. This brings us to a (Figure 1). All specimens used were identified by specialist question: to what extent do genetic differences correspond PhD Osvaldo Takeshi Oyakawa and identification-keys. to individual variation within a population or a divergence After identification, the specimens were deposited in the between the two species? Thus, for species identification, fish collection of the Laboratory of Biology and Genetic it is necessary to use different methodologies for data of Fish (LBP), Department of Morphology, Biosciences analysis so that the conflicting data can be treated with Institute, UNESP, Botucatu, São Paulo, Brazil. (see Table S1 greater care [4]. More recent studies have used analysis of the Supplementary Material). Specimen data, including of multispecies coalescent models such as the GMYC [33] geospatial coordinates of collection sites, are recorded in for the determination of species. In addition, the Barcode the publicly accessible BOLD projects titled “Fishes from of Life Database provides a tool for analysis that allows Brazilian Coastal Rivers; Fishes from Brazilian Coastal us to identify different molecular MOTUs without prior Rivers part II” (project codes: FBCR; FBCRB), this projects morphological identification, known as the BIN [34]. were merged and analyzed as one. Based on this information, the aim of this study was to use the methodology of DNA Barcode to identify fish 2.2 DNA Extraction, PCR amplification, and species of the Ribeira de Iguape River Basin and Coastal Rivers of São Paulo, and to evaluate the efficiency of Sequencing GMYC and BIN analysis in order to determine the number DNA barcoding was performed at the Laboratório de of MOTUs in these areas. Biologia e Genética de Peixes (LBP), UNESP, Botucatu, Brazil, and at the Canadian Centre for DNA Barcoding

Figure 1 Map of Ribeira de Iguape Basin (green area) and coastal rivers (red area), showing the sample sites (yellow points) where the speci- mens were obtained.

Unauthenticated Download Date | 12/4/15 10:38 PM 120 J. Monteiro Henriques, et al.

(CCDB), Canada. Total genomic DNA was isolated from 2.3 Sequence analysis fins or muscle tissue of each specimen using one of two different methods: with a DNeasy Tissue Kit (Qiagen), Consensus sequences from forward and reverse strands according to the manufacturer’s instructions; or with a were obtained using GENEIOUS PRO 5.4.2 [41]. Alignments vertebrate lysis buffer containing proteinase K digested were generated using MUSCLE [42] under default overnight at 56°C and subsequent extraction using a parameters. After alignment, the matrix was checked membrane-based approach on a Biomek NX (www.pall. visually for any obvious misalignments and potential com) liquid handling station using AcroPrep96 (www. cases of sequencing errors. A quality control step was beckman.com) and 1-mL filter plates with 10 mm PALL included in our workflow to detect contamination, glass fiber media [35] according to the CCDB protocol. A paralogous copies, or pseudogenes. Then, the presence of segment of approximately 648 bp from the 5ʹ end of the stop codons was checked using GENEIOUS PRO 5.4.2. mitochondrial cytochrome c oxidase subunit I (COI) gene Nucleotide variation and genetic distances were was amplified by the polymerase chain reaction (PCR) examined using MEGA 6.0 [43]. The best nucleotide using different combinations of primers: FishF1, FishR1, evolution models for the COI gene were evaluated using FishF2, FishR2 [36], the M13-tailed primer cocktails C_ MODELTEST 3.7 [44]. FishF1t1 – C_FishR1t1 and C_VF1LFt1 – C_VR1LRt1 [37], and the pair L5698-Asn [38] and H7271-COI [39]. A total volume 2.4 GMYC molecular species hypothesis of 12.5 μL of the PCR mixture included 1.25 μL of 10X buffer

(10 mM Tris-HCl + 15 mM MgCl2), 0.5 μL dNTPs (200 nM The lognormal relaxed molecular clock tree was of each), 0.5 μL each 5 mM primer, 0.05 μL Platinum® estimated using BEAST v.1.6.2 [45], since GMYC requires Taq Polymerase (Invitrogen), 1 μL template DNA (12 ng), an ultrametric tree, but the tree was not time calibrated. and 8.7 μL ddH2O. The PCR reactions consisted of 30–40 The nucleotide evolutionary model used to estimate the cycles, 30 s at 95°C, 15–30 s at 48–54°C (according to ultrametric tree was the GTR+I+G model with gamma each combination of primers), and 45 s at 72°C. At LBP, distribution (estimated by the program MODELTEST all PCR products were first visually identified on a 1% 3.7). A random tree was used as a starting tree for the agarose gel and then purified using ExoSap-IT® (USB Markov chain Monte Carlo searches. Eight chains were Corporation) following instructions of the manufacturer. run simultaneously for 10,000,000 generations and a tree The purified PCR products were sequenced using the Big was sampled every 100th generation. The distribution DyeTM Terminator v 3.1 Cycle Sequencing Ready Reaction of log-likelihood scores was examined to determine the Kit (Applied Biosystems) in a final volume of 7.0 μL stationary phase for each search and to decide if extra containing 1.4 μL of template, 0.35 μL of primer (10 μM), runs were required to achieve convergence, using the 1.05 μL of 5x buffer, 0.7 μL of BigDye mix and water. The program Tracer 1.5 [46]. All sampled topologies beneath cycle sequencing conditions included initial denaturation the asymptote (10,000,000 generations) were discarded at 96°C for 2 min followed by 35 cycles of denaturation at as part of a burn-in procedure, and the remaining trees 96°C for 45 s, annealing at 50°C for 60 s, and extension were used to construct a 50% majority-rule consensus tree at 60°C for 4 min. The PCR sequencing products were in TreeAnnotator v1.6.2 [47]. purified with EDTA/sodium acetate/alcohol following For species delimitation and consequent identification the protocol suggested in the BigDye™ Terminator v.3.1 of independent evolutionary units (IUE) we used the GMYC Cycle Sequencing kit’s manual (Applied Biosystems) and model. We performed the single threshold method with loaded on an automatic sequencer 3130-Genetic Analyzer standard parameters (interval = c(0,10)). This analysis capillary sequencer (Applied Biosystems) At the CCDB, was developed in the program R version 3.0.0 [48] with the PCR products were labeled using the BigDye Terminator package “splits” (Species Limits by Threshold Statistics) v.3.1 Cycle Sequencing Ready Reaction kit (Applied (http://r-forge.r-project.org/projects/splits) and default Biosystems) using standard methods [40] and were parameters. bidirectionally sequenced using an ABI3730 capillary sequencer. Sequence data, trace files, primer details, and 2.5 Analytical Tools of Barcode of Life collection localities for specimens are available within the Database (BOLD) projects FBCR and FBCRB on BOLD (http://v3.boldsystems. org/). Sequences have also been deposited in GenBank The Barcode of Life Database (BOLD) offers analytical (Accession numbers at Table S1 of the Supplementary tools for identifying and delimiting species. Thus, in order Material). to test the consistency of the data obtained by Bayesian

Unauthenticated Download Date | 12/4/15 10:38 PM DNA barcode fish Ribeira de Iguape coastal rivers Brazil 121 analysis, two further analyses were performed. the distribution of K2P divergence values showed that To evaluate the existence of a “barcoding gap” [49], we 88.7% of the intra-specific comparisons were less than used the distance matrix generated on the Barcode of Life 2%; however, 14.6% of the divergence values between Database (BOLD) (available on http://www.boldsystems. congeners were also less than 2%. The pairwise distance org), performed under a K2P model and preliminary comparison showed that the range of genetic divergence alignment with MUSCLE algorithm. between distinct morphospecies overlaps the variation Beyond that, the BIN was used to estimate the within some morphospecies (Figure 2), but the average of number of species directly from the barcode records (the genetic divergence found between the nearest neighbor full BIN database is available at bold system), following distance (NND) is almost 14 times the average found the conditions specified in research performed by within each morphospecies (synthesis of divergence Ratnasingham and Herbert [34] genetic comparison shown in Figure 2). Although there were 89 previously recognized morphospecies, the molecular approaches indicate 3 Results that this number may be underestimated. The GMYC analysis identified an intra/inter-specific threshold time We analyzed 805 samples from 89 morphologically of “-0.01112289” and recognized 99 (confidence interval: distinct groups (morphospecies) belonging to 64 genera, 95-102) independent evolution units (interpreted as 24 families, and 8 orders. The number of specimens distinct species or OTUs [33]), while the BIN approach analyzed ranged from 1 to 56 per morphospecies (mean indicated the existence of 104 BIN (interpreted as distinct = 9). The consensus sequences of COI from every sample species or MOTUs) [34] (Figure 3). By cross-tabulating have a mean size of 654 bp (minimum 600 pb) with no the results of all approaches, we identified 109 MOTUs stop codons, insertions, or deletions detected. After by at least one methodology. In most of the cases (89 alignment and editing a matrix was obtained with 600 MOTUs) the genetic approaches are in concordance with characters and a nucleotide composition of 24.12% morphological identification. adenine, 27.28% cytosine, 18.2% guanine, and 30.4% The present analyses show that the different thymine. The MODELTEST analyses select the GTR + I + G Pimelodella morphospecies cannot be identified by our as best nucleotide evolution model for the dataset. genetic data (Figure 3) (genetic divergence ranging from The mean Kimura-2-Parameter (K2P) genetic 0.3 to 1.1% and an average of 0.5% between the two divergence ranged from 0% to 8.7% (mean = 0.8%) for morphospecies). Other groups that show similar patterns intra-specific comparisons and from 0% to 24.5% (mean are Gymnotus (ranging from 0 to 5.4%, average 4.2%) = 10.9%) for congeneric comparisons, establishing and Phalloceros (ranging from 1.7 to 9.7%, average 6.1%), a barcode gap of about almost 14 times between whose morphospecies could not be distinguished by congeneric and intra-specific variation. The analyses of genetic analyses. However, in contrast to the Pimelodella

Figure 2 A) Histogram plotting the distribution of the mean intra-specific distances for each species. B) Histogram plotting the distribution of the nearest neighbor distances for each species. The distribution details are summarized below each histogram.

Unauthenticated Download Date | 12/4/15 10:38 PM 122 J. Monteiro Henriques, et al.

Figure 3 Part 1: Summarized Bayesian tree of the all samples obtained with COI data. The asterisks in the node branches represent the pos- terior probability greater than 95%. The taxa names were inferred by morphological identification. The vertical columns represent the status of the identification of the MOTUs, where the left side represents the GMYC analysis identifications, while the right side represents the BIN analysis identification. White regions in the columns represent MOTUs identified by all methodologies. Black marks represent MOTUs with conflicts in the identifications by at least one methodology. Red marks represent the clusters with morphospecies that are no reciprocally monophyletic (these clusters are shown in detail in Figure 4).

Unauthenticated Download Date | 12/4/15 10:38 PM DNA barcode fish Ribeira de Iguape coastal rivers Brazil 123

Figure 3 Part 2: Legend is the same of the Figure 3 part 1.

Unauthenticated Download Date | 12/4/15 10:38 PM 124 J. Monteiro Henriques, et al. results, these two groups have well-structured clusters, the existence of more than one MOTU within each although these clusters are not composed of only one morphospecies. morphospecies (Figure 4). However, some morphospecies such Symbranchus marmoratus (genetic divergence ranging from 0 to 5.5%, average 3,4%), Listrura camposi 4 Discussion (ranging from 0 to 16.1%, average 8.7%), Characidium From the data obtained in this study, it is possible to lanei (ranging from 0 to 4.6%, average 2.8%), Cyphocharax confirm the importance of interdisciplinary approaches santacatarinae (ranging from 0 to 17.9%, average 5.8%), in order to obtain more accurate data on species and Mimagoniates microlepis (ranging from 0 to 5.1%, identification. A preliminary morphological analysis average 1.6%) have two or more structured clusters identified 89 species in our database. Comparison with the (Figure 5), and in these cases, genetics analyses revealed genetic data revealed that the number of species varied

Figure 4 Detail of the clusters with morphospecies not monophyletic (with red marks in Figure 3). Asterisks represent nodes with a posterior probability higher than 95%. (A) Details of the Gymnotus clade. (B) Details of the Pimelodella clade. (C) Details of the Phalloceros clade. Note: In B, the scale bars represent 0.001 (or 0.1%) of GTR genetic divergence, while in A and C, the scale bar represents 0.01 (or 1%).

Unauthenticated Download Date | 12/4/15 10:38 PM DNA barcode fish Ribeira de Iguape coastal rivers Brazil 125

Figure 5 Detail of the clusters with monophyletic morphospecies (with black marks in Figure 3). Asterisks represent nodes with a posterior probability higher than 95%. (A) Details of Characidium lanei clade. (B) Details of Cyphocarax santacatarinae clade. (C) Details of Listrura camposi clade. (D) Details of Mimagoniates microlepis clade. (E) Details of Symbranchus marmoratus clade. Note: The scale bars represent 0.01 (or 1%) of GTR genetic divergence. from 89 (minimum number of concordant MOTUs in all To compare the data obtained in this study with the analyses) to 109 (maximum number of MOTUs obtained in previous studies on molecular species identification by at least one analysis). For this dataset, the GMYC analysis DNA Barcode, we performed a genetic analysis based on was the one that corroborated more with morphological the K2P model. Thus, we estimated the average intra/ data, but the fact that one of these methodologies point inter-specific genetic divergence of 0.81% and 10.94%, the existence of a new taxon is a reason for further work respectively, where divergence between species is 13.5-fold with the group by specialists and thereby improve the greater than divergence within them. Other studies using understanding of this diversity. the technique of DNA barcode for species identification Ratnasingham and Herbert [34] point out four main according to the geographical region show lower values reasons for the differences between MOTUs identification to those found in the present study, as follows: Valdez- by different approaches: taxonomic error, sequence Moreno et al. [18] and Pereira et al. [28], who studied contamination, deficits in the methodology, or the neotropical fishes from Central Mexico and South inability of the sequence variation of COI to identify America, obtained intra/inter-specific genetic distance species because of introgression or their young age. In relationships of 0.45%/5.1% and 0.3%/6.8%, respectively. some cases, as discussed below, these explanations may Although DNA barcode is accurate in identifying most apply. of the studied species, groups with a complex taxonomic

Unauthenticated Download Date | 12/4/15 10:38 PM 126 J. Monteiro Henriques, et al. history present discordance in MOTUs delimitation [25,50] a complex history of evolution of the basins involved in because some species are not distinguished by genetic this study [16,52–54]. approaches due to lack of reciprocal monophyly between them. All studied groups that we found problems to delimit the species and show overlap between intraspecific and 5 Conclusion interspecific values are present at the figure 3 (red marks The last icthyofauna checklist of the Ribeira de Iguape and black marks). Basin and coastal rivers from São Paulo identified 96 In the case of Gymnotus and Phalloceros, sample species [52], and in the present study, we found a similar identification at the species level is a difficult task value of 89 to 109 MOTUs, corroborating the hypothesis because of the similarity in morphological characteristics, that DNA barcode can be an effective tool for species especially in juvenile forms [15]. However, these two identification. However, the present data indicate the genera have well-defined internal clusters in our analyses, existence of unnamed species and species complexes that suggesting that a regional taxonomic revision is still require further investigation. necessary to check the problem of species delimitation in these genera. Acknowledgments: The authors wish to thank Oswaldo However, this explanation cannot be used to account T. Oyakawa for the help with the morphological for the non-reciprocal monophyly observed in Pimelodella. identifications, Renato Devidé, Bruno F. Melo, and Fábio F. The species P. transitoria and P. kronei are apparently Roxo for their help during the sampling expeditions, and easily distinguished morphologically because P. kronei BIO (Biodiversity Institute of Ontario) at the University show characteristics of cave fish (little pigmentation and of Guelph for the exchange program and sequencing no eyes), while P. transitoria is a typical surface fish [15]. facilities. To better investigate the relationship between these two This research was financially supported by Brazilian species, other types of markers are necessary. agencies such as FAPESP (Fundação de Amparo a Pesquisa The different methodologies of analysis also indicate do Estado de São Paulo) and CNPq (Conselho Nacional de another divergence. In some groups, such Synbranchus Desenvolvimento Científico e Tecnológico). marmoratus, Cyphocharax santacatarinae, Characidium lanei, Listrura camposi and Mimagoniates microlepis, we observed a genetic divergence between haplotypes, References and both genetic analyses (BIN and GMYC) suggest the existence of more than one MOTUs in the same [1] Frankham R., Genetics and conservation biology, C. R. Biol., morphospecies. This result is similar to that found in 2003, 326, 22–29. doi:10.1016/S1631-0691(03)00023-4 [2] Frankham R., Genetics and extinction, Biol. Conserv., 2005, other studies of fish barcodes [25,28,50]. According to 126, 131–140. doi:10.1016/j.biocon.2005.05.002 De Queiroz [51], sometimes when a speciation process [3] Frankham R, Briscoe D.A., Ballou J.D., Introduction to occurs, morphologic differentiations take time to Conservation Genetics Cambridge University Press, Cambridge, become perceptible, but with evolution, mutations can 2002. accumulate and genetic discriminations is possible [4] Carstens B.C., Pelletier T.A., Reid N.M., Satler J.D., How to without morphological differentiation; these entities are fail at species delimitation, Mol. Ecol., 2013, 22, 4369–83. doi:10.1111/mec.12413 also known as cryptic species. Evolution can act in the [5] Tautz D., Arctander P., Minelli A., Thomas R.H., Vogler A.P., DNA morphology of the species for the maintenance of body points the way ahead in taxonomy, Nature, 2002, 418, 479. shape and the result of this selective pressures is the doi:10.1038/418479a maintaining of the shape, while genetically the lineages [6] Hebert P.D.N., Ratnasingham S., deWaard J.R., Barcoding can have a high rate of evolution for COI, which could give animal life: cytochrome c oxidase subunit 1 divergences among closely related species, Proc. Biol. Sci., 2003, 270, S96–9. an illusion that the time of divergence is greater than it doi:10.1098/rsbl.2003.0025 really is [11], and the reverse is also true, once COI may [7] Hebert P.D.N., Cywinska A., Ball S.L., deWaard J.R., Biological evolve slowly among different lineages, as well as, groups identifications through DNA barcodes, Proc. Biol. Sci., 2003, can evolve faster morphologically. 270, 313–21. doi:10.1098/rspb.2002.2218 Some of the morphospecies in this study are found [8] Lipscomb D., Platnick N., Wheeler Q., The intellectual content in all independent drainages (Ribeira basin and all small of taxonomy: a comment on DNA taxonomy, Trends Ecol Evol, 2003, 18, 65–66. doi:10.1016/S0169-5347(02)00060-5 drainages along the coast that were sampled); among [9] Tautz D., Arctander P., Minelli A., Thomas R.H., Vogler A.P., A these, some have a high genetic divergence, while others plea for DNA taxonomy, Trends Ecol. Evol., 2003,18, 70–74. have a low genetic divergence. This may be a reflection of doi:10.1016/S0169-5347(02)00041-1

Unauthenticated Download Date | 12/4/15 10:38 PM DNA barcode fish Ribeira de Iguape coastal rivers Brazil 127

[10] Ebach M.C., Holdrege C., DNA barcoding is no substitute for Mol. Ecol. Resour., 2011, 11, 60–71. doi:10.1111/j.1755- taxonomy, Nature, 2005, 434, 697. doi:10.1038/434697b 0998.2010.02894.x [11] Waugh J., DNA barcoding in animal species: progress, potential [25] Pereira L.H.G., Maia G.M.G., Hanner R., Foresti F., Oliveira C., and pitfalls, Bioessays, 2007, 29, 188–97. doi:10.1002/ DNA barcodes discriminate freshwater fishes from the Paraíba bies.20529 do Sul River Basin, São Paulo, Brazil, Mitochondrial DNA, 2011, [12] Wiemers M., Fiedler K., Does the DNA barcoding gap exist? - a S1, 71–79. doi:10.3109/19401736.2010.532213 case study in blue butterflies (Lepidoptera: Lycaenidae), Front. [26] Ribeiro A., Caires R., Mariguela T., Pereira L., Hanner R., Zool., 2007, 4, 8. doi:10.1186/1742-9994-4-8 Oliveira C., DNA barcodes identify marine fishes of Sao Paulo [13] Castro R.M.C., Menezes N.A., Estudo diagnóstico da State, Brazil, Mol. Ecol. Resour., 2012, 12, 1012–1020. diversidade de peixes do Estado de São Paulo. Biodiversidade [27] Rosso J.J., Mabragaña E., González Castro M., Díaz de Astarloa do Estado de São Paulo, Brasil: Síntese do conhecimento ao J.M., DNA barcoding Neotropical fishes: Recent advances final do século XX. WinnerGraph, 1998, pp. 1–13. from the Pampa Plain, Argentina, Mol. Ecol. Resour., 2012, 12, [14] Bizerril C.R.S.F., Lima N., Levantamento da ictiofauna da bacia 999–1011. doi:10.1111/1755-0998.12010 do Rio Ribeira, Brasil, Acta Biol. Leopoldensia, 2000, 22, [28] Pereira L.H.G., Hanner R., Foresti F., Oliveira C., Can DNA 103–110. barcoding accurately discriminate megadiverse Neotropical [15] Oyakawa O.T., Akama A., Mautari K.C., Nolasco J.C., Peixes de freshwater fish fauna?, BMC Genet., 2013, 14, 20. riachos da Mata Atlântica: nas unidades de conservação do doi:10.1186/1471-2156-14-20 Vale do Rio Ribeira de Iguape no Estado de São Paulo, Editora [29] Fouquet A., Gilles A., Vences M., Marty C., Blanc M., Gemmell Neotropica, São Paulo, 2006. N.J., Underestimation of species richness in neotropical frogs [16] Ribeiro A.C., Lima F.C.T., Riccomini C., Menezes N. A., Fishes revealed by mtDNA analyses, PLoS One, 2007, 2. doi:10.1371/ of the Atlantic Rainforest of Boracéia: Testimonies of the journal.pone.0001109 Quaternary fault reactivation within a Neoproterozoic tectonic [30] Allcock A.L., Barratt I., Eléaume M., Linse K., Norman M.D., province in Southeastern Brazil, Ichthyol. Explor. Freshwaters, Smith P.J., et al., Cryptic speciation and the circumpolarity 2006, 17, 157–164. debate: A case study on endemic Southern Ocean octopuses [17] Hubert N., Hanner R., Holm E., Mandrak N.E., Taylor E., Burridge using the COI barcode of life, Deep Res. Part II Top Stud. M., et al., Identifying Canadian freshwater fishes through Oceanogr., 2011, 58, 242–249. doi:10.1016/j.dsr2.2010.05.016 DNA barcodes, PLoS One, 2008, 3. doi:10.1371/journal. [31] Vanhaecke D., de Leaniz C.G., Gajardo G., Young K., Sanzana pone.0002490 J., Orellana G., et al., DNA barcoding and microsatellites help [18] Valdez-Moreno M., Ivanova N. V., Elías-Gutiérrez M., species delimitation and hybrid identification in endangered Contreras-Balderas S., Hebert P.D.N., Probing diversity in galaxiid fishes, PLoS One, 2012, 7. doi:10.1371/journal. freshwater fishes from Mexico and Guatemala with DNA pone.0032939 barcodes, J. Fish Biol., 2009,74, 377–402. doi:10.1111/j.1095- [32] Zhou C., Kandemir I., Walsh D.B., Zalom F.G., Lavine L.C., 8649.2008.02077.x Identification of lygus hesperus by DNA barcoding reveals [19] Lara A., Ponce de León J.L., Rodríguez R., Casane D., Côté insignificant levels of genetic structure among distant and G., Bernatchez L., et al., DNA barcoding of Cuban freshwater habitat diverse populations, PLoS One, 2012, 7. doi:10.1371/ fishes: evidence for cryptic species and taxonomic conflicts, journal.pone.0034528 Mol. Ecol. Resour., 2010, 10, 421–30. doi:10.1111/j.1755- [33] Pons J., Barraclough T.G., Gomez-Zurita J., Cardoso A., Duran 0998.2009.02785.x D.P., Hazell S., et al., Sequence-based species delimitation for [20] April J., Mayden R.L., Hanner R.H., Bernatchez L., Genetic the DNA taxonomy of undescribed insects, Syst. Biol., 2006, calibration of species diversity among North America 55, 595–609. doi:10.1080/10635150600852011 ’ s freshwater fi shes, Proc. Nat. Acad. Sci., 2011, 108 [34] Ratnasingham S., Hebert P.D.N., A DNA-Based Registry for All (26), 10602-10607. doi:10.1073/pnas.1016437108/-/ Animal Species: The Barcode Index Number (BIN) System, PLoS DCSupplemental.www.pnas.org/cgi/doi/10.1073/ One, 2013, 8. doi:10.1371/journal.pone.0066213 pnas.1016437108 [35] Ivanova N. V.., Dewaard J.R., Hebert P.D.N., An inexpensive, [21] Aquilino S.V.L., Tango J.M., Fontanilla I.K.C., Pagulayan R.C., automation-friendly protocol for recovering high-quality DNA, Basiao Z.U., Ong P.S., et al., DNA barcoding of the ichthyofauna Mol. Ecol. Notes, 2006, 6, 998–1002. doi:10.1111/j.1471- of Taal Lake, Philippines, Mol. Ecol. Resour., 2011, 11, 612–619. 8286.2006.01428.x doi:10.1111/j.1755-0998.2011.03000.x [36] Ward R.D., Zemlak T.S., Innes B.H., Last P.R., Hebert P.D.N., DNA [22] Asgharian H., Sahafi H.H., Ardalan A.A., Shekarriz S., Elahi barcoding Australia’s fish species, Philos. Trans. R. Soc. Lond. E., Cytochrome c oxidase subunit 1 barcode data of fish of the B Biol. Sci., 2005, 360, 1847–1857. doi:10.1098/rstb.2005.1716 Nayband National Park in the Persian Gulf and analysis using [37] Ivanova N. V., Zemlak T.S., Hanner R.H., Hebert P.D.N., Universal meta-data flag several cryptic species, Mol. Ecol. Resour., 2011, primer cocktails for fish DNA barcoding, Mol. Ecol. Notes, 2007, 11, 461–72. doi:10.1111/j.1755-0998.2011.02989.x 7, 544–548. doi:10.1111/j.1471-8286.2007.01748.x [23] De Carvalho D.C., Oliveira D.A.A., Pompeu P.S., Leal C.G., [38] Miya M., Nishida M., Use of mitogenomic information in Oliveira C., Hanner R., Deep barcode divergence in Brazilian teleostean molecular phylogenetics: a tree-based exploration freshwater fishes: the case of the São Francisco River basin, under the maximum-parsimony optimality criterion, Mitochondrial DNA, 2011, S1, 80–86. doi:10.3109/19401736.20 Mol. Phylogenet. Evol., 2000, 17, 437–455. doi:10.1006/ 11.588214 mpev.2000.0839 [24] Lakra W.S., Verma M.S., Goswami M., Lal K.K., Mohindra [39] Melo B.F., Benine R.C., Mariguela T.C., Oliveira C., A new V., Punia P., et al., DNA barcoding Indian marine fishes, species of Tetragonopterus Cuvier, 1816 (Characiformes:

Unauthenticated Download Date | 12/4/15 10:38 PM 128 J. Monteiro Henriques, et al.

Characidae: Tetragonopterinae) from the rio Jari, Amapá, [47] Rambaut A., Drummond A.J., TreeAnnotator version 1.6.2, 2011. northern Brazil, Neotrop. Ichthyol., 2011, 9, 49–56. [48] R Development Core Team, R. A. Lang Environ. Stat. Comput., doi:10.1590/S1679-62252011000100002 2013. [40] Hajibabaei M , deWaard J.R., Ivanova N. V., Ratnasingham [49] Meyer C.P., Paulay G., DNA barcoding: Error rates based S., Dooh R.T., Kirk S.L., et al., Critical factors for on comprehensive sampling, PLoS Biol., 2005, 3, 1–10. assembling a high volume of DNA barcodes, Philos. doi:10.1371/journal.pbio.0030422 Trans. R. Soc. Lond. B Biol. Sci., 2005, 360, 1959–67. [50] Dettai A., Lautredou A.C., Bonillo C., Goimbault E., Busson F., doi:10.1098/rstb.2005.1727 Causse R., et al., The actinopterygian diversity of the CEAMARC [41] Drummond A.J., Ashton B., Buxton S., Cheung M., Cooper A., cruises: Barcoding and molecular taxonomy as a multi-level Duran C., et al. Geneious v5.4, http://www.geneious.com. tool for new findings, Deep Res. Part II Top Stud. Oceanogr., [42] Edgar R.C., MUSCLE: a multiple sequence alignment method 2011, 58, 250–263. doi:10.1016/j.dsr2.2010.05.021 with reduced time and space complexity, BMC Bioinformatics, [51] De QueirozK., Species concepts and species delimitation, Syst. 2004, 5, 113. doi:10.1186/1471-2105-5-113 Biol., 2007, 56, 879–86. doi:10.1080/10635150701701083 [43] Tamura K., Stecher G., Peterson D., Filipski A., Kumar S., [52] Menezes N.A., Weitzman S.H., Oyakawa O.T., Lima F.C.T., Castro MEGA6: Molecular Evolutionary Genetics Analysis version 6.0., R.M.C., Weitzman M.J., Peixes de água doce da Mata Atlântica, Mol. Biol. Evol., 2013, 30, 2725–9. doi:10.1093/molbev/mst197 Museu de Zoologia/USP, 2007. [44] Posada D., Crandall K.A., MODELTEST: testing the model of DNA [53] Menezes N.A, Ribeiro A.C., Weitzman S., Torres R. A., substitution, Bioinformatics, 1998, 14, 817–818. doi:10.1093/ Biogeography of Glandulocaudinae ( Teleostei : Characiformes : bioinformatics/14.9.817 Characidae ), Zootaxa, 2008, 48, 33–48. [45] Drummond A.J., Rambaut A. ,BEAST: Bayesian evolutionary [54] Buckup P.A., The Eastern Brazilian shield. In: Albert JE, Reis analysis by sampling trees, BMC Evol. Biol., 2007, 7, 214. RE, editors. Historical biogeography of Neotropical freshwater doi:10.1186/1471-2148-7-214 fishes, Berkeley, University of California Press, 2011. [46] Rambaut A., Suchard M., Xie D., Drummond A., Tracer | BEAST, 2014. Supplemental Material: The online version of this article (DOI: 10.1515/dna-2015-0015) offers supplementary material.

Unauthenticated Download Date | 12/4/15 10:38 PM