ABSTRACT
IUCN Red Lists have been a valuable tool to prioritize conservation plans in endemic neotropical frogs. However, many areas in this region are poorly known in terms of their diversity and endemism. Based on examined museum specimens of the threatened species Pristimantis penelopus we revised its geographic distribution and determined the habitat suitability using niche modeling techniques. Using a mitochondrial fragment of COI gene, we determine the phylogenetic position and the extent of the genetic variation across its distribution in Colombia. We present the first records of P. penelopus for the Cordillera Oriental, the western versant of Cordillera Occidental and the northern portion of the Cauca river basin. Based on the molecular phylogenetic analysis, Pristimantis penelopus belongs to the P. ridens series sensu (Padial et al., 2014). The mean of intraspecific genetic variation is 2.1% and the variation among population ranges between 2.3 and 3.5%. The genetic distance between the western populations and the Magdalena Valley populations suggests a potential phylogeographic break in northwestern Antioquia. We expand the realized distribution by 258 kilometers north, 200 km east and 223 km northwest. Based on our results and according to the IUCN criteria we propose a new category for the species and highlight the need to increase the surveys in poorly known regions to better understand the geographic distribution and conservation status of listed species.
KEY-WORDS:
Colombia; IUCN Red List; Niche modeling; Phylogeography; Terrarana
RESUMEN
Las listas rojas de la IUCN han sido una herramienta fundamental para priorizar planes de conservación de ranas endémicas neotropicales. Sin embargo, muchas áreas en esta región han sido poco estudiadas y el endemismo puede ser un artificio del muestreo. A partir de especímenes de museo de la especie amenazada Pristimantis penelopus revisamos su distribución geográfica y determinamos su hábitat potencial utilizando técnicas de modelo de nicho. A partir de un fragmento del gen mitocondrial COI, determinamos la posición filogenética de la especie y la magnitud de su variación genética a lo largo de su distribución en Colombia Presentamos los primeros registros de P. penelopus en la Cordillera Oriental de los Andes, en la vertiente occidental de la Cordillera Occidental y en el norte de la cuenca del río Cauca. Teniendo en cuenta los análisis filogenéticos moleculares, Pristimantis penelopus pertenece al grupo P. ridens sensu (Padial et al., 2014). La variación genética intraespecífica promedio fue de 2.1% y la variación entre poblaciones estuvo entre 2.3 y 3.5%. La distancia genética entre las poblaciones de la Cordillera Occidental y las poblaciones del Valle del Magdalena sugieren la presencia de un quiebre filogeográfico en el noroccidente de Antioquia. Expandimos la distribución realizada de P. penelopus 258 kilómetros al norte, 200 km al este y 223 km al noroccidente. Teniendo en cuenta estos resultados y los criterios propuestos por la IUCN proponemos una nueva categoría de amenaza para la especie y resaltamos la necesidad de incrementar los muestreo en regions geográficas poco conocidas para comprender mejor la distribución geográfica y el estado de conservación de las especies incluidas en las Listas Rojas.
Palabras-Clave:
Colombia; Listas Rojas de la IUCN; Modelos de Nicho; Filogeografía; Terrarana
INTRODUCTION
The IUCN Red List of Threatened Species highlights which species are at the greatest risk of extinction and seeks to promote their conservation (Collar, 1996; Rodríguez et al., 2006). To determine the conservation status of a species, IUCN uses objective criteria and the expert opinion regarding population declines, geographic range (i.e., extent of occurrence and area of occupancy), number of mature individuals, population size and extinction probability based on quantitative analyses (IUCN, 2001).
In the Neotropical region, evidence about population trends and species distribution of amphibians is scant as many species are only known from few individuals (i.e., the type series). Many regions have not been adequately sampled and information on the distribution range and area of occupancy is anecdotal and fragmented. This limited information hampers the appropriate assessment of the species conservation status. As a consequence, conservation efforts would be incorrectly implemented for species with an equivocal conservation status. Given the scarce information regarding population trends in neotropical amphibians, species description and occurrence records are the most important sources of information for species conservation assessments.
The frog genus Pristimantis Jiménez de la Espada, 1870, a very species-rich lineage with more than 470 described species, is distributed throughout southern Central America and northern South America (Padial et al., 2014). Currently, 162 species are listed in one of the three IUCN threat categories and most of them under the B criterion (reduced extent of presence or area of occupancy; IUCN, 2016). All listed Pristimantis have restricted geographic distribution and face some habitat loss across their range. As a case, Pristimantis penelopus (Lynch & Rueda-Almonacid, 1999) is endemic to Colombia and is listed by IUCN as Vulnerable (VU B1ab (iii)) because its extent of occurrence is less than 20,000 km², it is known from fewer than ten locations, and there is continuing decline in the extent and quality of its habitat on the Cordillera Central of the Colombian Andes (Castro et al., 2004).
Here, we evaluate the extent of occurrence of Pristimantis penelopus (Lynch & Rueda-Almonacid, 1999) based on recent collecting efforts in northwestern Colombia. Based on this new information in addition to genetic analysis and ecological niche modeling, we reassess its conservation category according to the IUCN Red List criteria. Although, new Pristimantis species are described every year, we show that an extensive geographic sampling and an integrative approach (e.g., environmental and genetic) will help to reassess the conservation status of many other threatened species.
MATERIALS AND METHODS
Specimen records
We collated records for P. penelopus from the GBIF database (www.gbif.org) from three biological collections: Instituto de Ciencias Naturales (ICN), Instituto Alexander von Humboldt (IAvH), and Museo de Herpetología Universidad de Antioquia (MHUA). Because records from the MHUA collection represent 73% out of all records for this species, we checked all specimens to corroborate taxonomic identity and we obtained 318 records for P. penelopus. Twenty-two remaining records are deposited at IAvH (17 records) and ICN (five records). In total, 316 confirmed records were obtained and mapped (Fig. 1).
Geographic sampling of Pristimantis penelopus. Colored circles indicate sequenced specimens. Different colors represent the populations used in the genetic analysis (see Figure 2 for color codes).
Laboratory procedures
Samples of 37 individuals from three species: Pristimantis penelopus, Pristimantis erythropleura (Boulenger, 1896), and Pristimantis viejas (Lynch & Rueda-Almonacid, 1999) were sequenced. Samples of P. penelopus come from 15 localities across its range (Fig. 1). Total genomic DNA was extracted from tissue samples (muscle) using the Qiagen DNeasy kit (QIAGEN). A fragment of the mitochondrial genome corresponding to the cytochrome oxidase I gene (COI) was amplified via PCR using the primers dgLCO-dgHCO (Meyer, 2003). All PCR products were sequenced at the Macrogen sequencing facility (www.macrogen.com). Raw sequence chromatographs were edited using Geneious 8.1.4 (Kearse et al., 2012) and aligned with the program MUSCLE using default parameters (Edgar, 2004). All sequences generated in this study were deposited in GenBank (Supplementary Material).
Phylogenetic analysis
We combined previously published COI sequences with the new sequences generated for this study to create a matrix with a total of 109 terminals (Supplementary Material). (Pinto-Sánchez et al., 2014) and later (Padial et al., 2014) incorrectly included in their phylogenetic analyses the voucher AJC1344 (MHUAA48119) as Pristimantis paisa. However, after specimen examination and a molecular analysis, we unambiguously identified this specimen as Pristimantis penelopus. We used their phylogenetic hypotheses to determine the taxon sampling for our genetic analysis. We included available intraspecific sampling for the species most closely related to P. penelopus: P. erythropleura, P. cruentus (Peters, 1873), P. latidiscus (Boulenger, 1898), and P. museosus (Ibáñez et al., 1994). This group represents the P. ridens series sensu (Padial et al., 2014). Based on (Pinto-Sánchez et al., 2014) and our preliminary analyses with the entire genus, we used Pristimantis viejas as the outgroup. We simultaneously inferred the best model of evolution and the Maximum Likelihood tree using the program IQ-TREE (Nguyen et al., 2015). Nodal support was estimated using the ultrafast bootstrap implemented in IQ-TREE (Minh et al., 2013). In addition, we inferred a Bayesian phylogenetic tree using the Markov chain Monte Carlo method implemented in the program BEAST 1.8.3 (Drummond et al., 2012). We implemented the GTR+G+I model of evolution for the entire dataset as suggested by IQ-TREE. We initiated two independent runs from a starting random tree for 10 million generations sampling every 1,000 generations. On each run, the first two million generations were discarded as burn-in and the remaining samples were combined. Nodal support as Bayesian posterior probabilities were annotated on the maximum clade credibility tree.
Population genetic analysis
A haplotype network of P. penelopus samples was obtained using the median joining method (Bandelt et al., 1999) implemented in the program PopART (Leigh, 2016). In PopART, we implemented an AMOVA test (Excoffier et al., 1992) to stablish the molecular variation among geographic region, among localities and within localities. Uncorrected genetic distances among populations and within P. penelopus were estimated in Mega (Kumar et al., 2016). The correlation between geographic and genetic distances was determined using a Mantel test with randomization, which tests for significance of a regression using a randomized permutation procedure to account for the potential non-independence among samples (Wang, 2013). This method was implemented in zt with 10,000 permutations (Bonnet & de Peer, 2002).
Environmental niche modeling
We compiled occurrence records for P. penelopus as we mentioned before and only used for niche modeling those records corroborated by us using morphological and genetic data (53 unique locality records, Table 1). We buffered (~ 225 km of radius) each occurrence points to delimitate the background area to calibrate our niche model. We generated potential geographical distributions for this species using several algorithms (e.g., GAM, GLM, Maxent, Mars, SVM) implemented in the sdm R package (Naimi & Aráujo, 2016) using 10 replicates of subsampling splitting occurrence data in 70% for training and 30% for testing. We used only 11 of 19 bioclimatic variables which were the least correlated between them (bio1, bio2, bio3, bio4, bio8, bio9, bio12, bio15, bio16, bio17, bio19). The best model was selected as the one that maximizes validation metrics of both: Area under the curve (AUC) and the True Skill Statistics (TSS) (Allouche et al., 2006). The best model was one generated by Maxent and we reclassified this to a binary prediction (i.e., presence-absence) using the minimum training presence threshold. We calculated the potential presence area (km²) using ArcMap 10.2 (ESRI, 2011).
Number of individuals, haplotypes and localities by geographic region of Pristimantis penelopus.
RESULTS
The aligned matrix included 688 sites. Phylogenetic relationships within the P. ridens series is in agreement with previous studies (Padial et al., 2014; Pinto-Sánchez et al., 2014). Pristimantis penelopus is recovered as monophyletic and its sister species is P. erythropleura (Fig. 2). Maximum likelihood and Bayesian trees show a phylogeographic structure in P. penelopus where the regions west of Cordillera Occidental, Bolivar, Santander and Antioquia represent different clades. However, two haplotypes from Antioquia are more related to Santander and Bolivar respectively (Fig. 2).
(Left) Maximum clade credibility tree depicting the phylogenetic position of Pristimantis penelopus within the P. ridens series. Numbers on nodes indicate posterior probabilities. Numbers below nodes represent nodal support using the ultrafast bootstrap (see methods). Asterisks indicate nodal support above 95% in both Bayesian and ML methods. (Right) Haplotype network based on 460 bp of the COI region. Numbers of mutational steps are shown on the lines connecting haplotypes. Colors refer to geographic locations shown in Figure 1.
For the intraspecific variation analyses we used a matrix with 460 base pairs of COI obtained from 34 individuals. A total of 42 variable nucleotide sites were found with no insertions or deletions and 18 different haplotypes were identified. All haplotypes were restricted to a single geographic region and we did not find dominant haplotypes. The median-joining network showed a pattern with phylogeographic structure in accordance with the phylogenetic tree (Fig. 2). The AMOVA results for the five geographic regions estimated that 64.7% of the genetic variation occurred among geographic regions, 33.8% among localities and 1.6% within populations. The genetic variation we found follows an isolation by distance pattern (IBD; r = 0.77, P < 0.0001). The mean genetic distance between geographic regions ranged from 2.3% and 3.5% (Table 2).
Uncorrected genetic distances among Pristimantis penelopus populations based on a 688 bp fragment of the mitochondrial gene COI. Numbers on the diagonal represent the intraspecific variation.
The potential geographic distribution of Pristimantis penelopus encompass the central and northern Andes in Colombian and southern Venezuela (Fig. 4). Its distribution area estimated from the environmental niche model is ~ 191535 km². The model has a relative good performance (AUC = 0.93; TSS = 0.77).
Phenotypic variation of Pristimantis penelopus across its distribution. Localitites are shown in Appendix 1.
Potential distribution of Pristimantis penelopus based on ecological niche modeling (red). Yellow dots represents occurrence localities used to calibrate the model. See main text for details.
DISCUSSION
Piristimantis penelopus distribution has been restricted to the sub-Andean forests of the eastern slopes of the Cordillera Central in Antioquia and Caldas Departments (Lynch & Rueda-Almonacid, 1999; Bernal & Lynch, 2008) at elevations between 1,180 to 1,500 m. Several authors (Castro et al., 2004; Stuart et al., 2008; Llano-Mejía et al., 2010) mentioned the presence of P. penelopus in Tolima department but no voucher specimens support its presence. Combining specimen examination, genetic analysis, and niche modeling, we show that Pristimantis penelopus is distributed along the eastern flank of the Cordillera Central in Caldas and Antioquia, the western flank of the Cordillera Oriental in Santander, north of Serranía San Lucas in Bolívar, and the western flank of the Cordillera Central in Antioquia. Our findings expand the known distribution 258 kilometers north, 200 km east and 223 km northwest relative to the closest edge in the previously known distribution range for the species (Fig. 4). Altitudinal distribution is also extended as the original range changed from 1,180-1,500 m to 94-1,720 m.
We present molecular evidence that allows us to clarify the phylogenetic position of Pristimantis penelopus, the interpopulation genetic variation, and the distribution range. Phylogenetic relationships within Pristimantis are still incompletely understood because of the lack of adequate taxon sampling and the high levels of cryptic diversity and taxonomic complexity in this highly diverse clade (Fig. 3; Rivera-Correa & Daza, 2016). Although we used only a fragment of the mitochondrial genome to determine the phylogenetic position of P. penelopus, our results are in agree with the more extensive sampling of taxa and genes in a large Pristimantis phylogeny (i.e.,Pinto-Sánchez et al. 2014). The sister species to P. penelopus is P. erythropleura (Boulenger, 1896). This later species is distributed mainly on the western flank of the Cordillera Occidental of Colombia at elevations from 1,200-2,600 m (Frost, 2016). The genetic variation within Pristimantis species in the P. ridens group are fairly similar (Fig. 2).
The presence of a distinct clade in the lowlands of the western flank on the Cordillera Occidental suggest that a potential phylogeographic break might occur between the Chocó and the Magdalena lowlands, likely due to a conspicuous geographical/ecological barrier hard to delimitate. In some cases, this putative break would result in sister species on both sides of the barrier (e.g., Allobates talamancae and Allobates niputidea, Agalychnis callidryas and Agalychnis terranova). In contrast, western and eastern populations of Pristimantis penelopus correspond to the same species as the observed intraspecific variation is similar to the one found in other Pristimantis species (Crawford et al., 2010; García-R. et al., 2012; Fig. 2). Although we do not have quantitative measures of morphological variation across this break, specimen examination also indicates that populations from both sides represent one single species. More sampling effort in the northern cordilleras Occidental and Central and the Cauca canyon will illuminate the presence and location of this hypothetical phylogeographic break and its role in species divergence.
Pristimantis penelopus is currently categorized as a Vulnerable species (VU B1ab (iii)). Before our study, fewer than ten localities are known for the species and the extent of occurrence (EOO) was less than 20,000 km². In addition, the species has been considered to be rare and restricted to mature forest (Castro et al., 2004). The continued decline in the extent and quality of its habitat across its range was considered an indirect evidence of population decline. Our results show that P. penelopus occurs in at least 26 localities and the EOO is more than seven times larger than originally considered. The species has been found in secondary forests, and all specimens deposited in the MHUA collection provide indirect evidence that the species is not rare. Therefore, based on our results, we suggest that this species is assigned to the Least Concern category according to IUCN guidelines.
CONCLUSION
Our findings highlight the need for thorough sampling in poorly studied regions in the Neotropics. Taxonomic studies using multiple lines of evidence (morphology, genetics and environmental suitability) are necessary to uncover geographic structure and distributional ranges of poorly known amphibians. Lastly, conservation biology should not only deal with enhancing the persistence of local and rare populations but also by exploring undersampled areas that might lead to discover new populations and in turn will decrease the extinction risks of the species across its distribution.
ACKNOWLEDGMENTS
The study was funded by ISAGEN and the Grupo Herpetológico de Antioquia under contract 47/574. We thank E. Alzate, F. Grisales and J.P. Hurtado for providing photographs. C. Muñoz and J. Fang helped in the genetic data generation. We thank students and researchers associated to the MHUA who after the years have collected the largest voucher dataset for this species.
REFERENCES
- ALLOUCHE, O.; TSOAR, A. & KADMON, R. 2006. Assessing the accuracy of species distribution models: Prevalence, kappa and the true skill statistic (TSS). Journal of Applied Ecology, 43:1223-1232. DOI.
- BANDELT, H.; FORSTER, P. & RÖHL, A. 1999. Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution, 16:37-48.
- BERNAL, M.H. & LYNCH, J.D. 2008. Review and analysis of altitudinal distribution of the Andean anurans in Colombia. Zootaxa, 1826:1-25. DOI.
- BONNET, E. & VAN DE PEER, Y. 2002. Zt: A software tool for simple and partial Mantel tests. Journal of Statistical Software, 7:1-12. DOI.
- BOULENGER, G.A. 1896. Descriptions of new reptiles and batrachians from Colombia. Annals and Magazine of Natural History Series, 6(17):16-21.
- BOULENGER, G.A. 1898. An account of the reptiles and batrachians collected by Mr. W.F.H. Rosenberg in Western Ecuador. Proceedings of the Zoological Society of London, 1898:107-126.
-
CASTRO, F.; HERRERA, M.I. & LYNCH, J.D. 2004. Pristimantis penelopus. In: The IUCN Red List of Threatened Species 2004. Available from Available from www.iucnredlist.org/details/56841/0
Accessed in: 10/2016.
» www.iucnredlist.org/details/56841/0 - COLLAR, N.J. 1996. The reasons for Red Data Books. Oryx, 30:121-130. DOI.
- CRAWFORD, A.J.; LIPS, K.R. & BERMINGHAM, E. 2010. Epidemic disease decimates amphibian abundance, species diversity, and evolutionary history in the highlands of central Panama. Proceedings of the National Academy of Sciences, 107:13777-13782.
- DRUMMOND, A.J.; SUCHARD, M.A.; XIE, D. & RAMBAUT, A. 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7 Molecular Biology and Evolution, 29:1969-1973. DOI.
- EDGAR, R.C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32:1792-1797. DOI.
- ESRI. 2011. ArcGIS Desktop: Release 10. Available from Environmental Systems Research Institute. Redlands, USA.
- EXCOFFIER, L.; SMOUSE, P.E. & QUATTRO, J.M. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics, 131:479-491.
-
FROST, D.R. 2016. Amphibian species of the world: an online reference. Version 6.0. (Junuary 2017) Available from http://research.amnh.org/herpetology/amphibia/index.html New York, American Museum of Natural History.
» http://research.amnh.org/herpetology/amphibia/index.html - GARCÍA-R., J.C.; CRAWFORD, A.J.; MENDOZA, Á.M.; OSPINA, O.; CARDENAS, H. & CASTRO, F. 2012. Comparative phylogeography of direct-developing frogs (Anura: Craugastoridae: Pristimantis) in the southern Andes of Colombia. PLoS One, 7:e46077.
- IBÁÑEZ, R.; JARAMILLO, C.A. & AROSEMENA, F.A. 1994. A new species of Eleutherodactylus (Anura: Leptodactylidae) from Panamá. Amphibia-Reptilia, 15:337-341. DOI.
- IUCN. 2001. IUCN Red List Categories and Criteria: Version 3.1. Gland and Cambridge, IUCN Species Survival Commission.
-
IUCN. 2016. The IUCN Red List of Threatened Species. Version 2016-3. (January 2017). Available from www.iucnredlist.org
» www.iucnredlist.org - KEARSE, M.; MOIR, R.; WILSON, A.; STONES-HAVAS, S.; CHEUNG, M.; STURROCK, S.; BUXTON, S.; COOPER, A.; MARKOWITZ, S.; DURAN, C.; THIERER, T.; ASHTON, B.; MEINTJES, P. & DRUMMOND, A. 2012. Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics, 28:1647-1649. DOI.
- KUMAR, S.; STECHER, G. & TAMURA, K. 2016. MEGA7: Molecular evolutionary genetics analysis. Version 7.0 for bigger datasets. Molecular Biology and Evolution, 33(7):1870-1874.
-
LEIGH, J. 2016. PopArt Version 1.7. Available from http://popart.otago.ac.nz
» http://popart.otago.ac.nz - LLANO-MEJÍA, J.; CORTÉS-GÓMEZ, A.M. & CASTRO-HERRERA, F. 2010. Lista de anfibios y reptiles del departamento del Tolima, Colombia. Biota Colombiana, 11:89-106.
- LYNCH, J.D. & RUEDA-ALMONACID, J.V. 1999. New species of frogs from low and moderate elevations from the Caldas transect of the eastern flank of the Cordillera Central. Revista de la Academia Colombiana de Ciencias Exactas, Físicas y Naturales, 23:307-314.
- MEYER, C.P. 2003. Molecular systematics of cowries (Gastropoda: Cypraeidae) and diversification patterns in the tropics. Biological Journal of the Linnean Society, 79:401-459. DOI.
- MINH, B.Q.; NGUYEN, M.A.T. & VON HAESELER, A. 2013. Ultrafast approximation for phylogenetic bootstrap. Molecular Biology and Evolution, 30:1188-1195.
- NAIMI, B. & ARAÚJO, M. 2016. Sdm: a reproducible and extensible R platform for species distribution modeling. Ecography, 39:368-375.
- NGUYEN, L.T.; SCHMIDT, H.A.; VON HAESELER, A. & MINH, B.Q. 2015. IQ-TREE: A fast and effective stochastic algorithm for estimating maximum likelihood phylogenies. Molecular Biology and Evolution, 32:268-274. DOI.
- PADIAL, J.M.; GRANT, T. & FROST, D. 2014. Molecular systematics of terraranas (Anura: Brachycephaloidea) with an assessment of the effects of alignment and optimality criteria. Zootaxa, 3825:1-132. DOI.
- PETERS, W.C.H. 1873. Über eine neue Schildrötenart, Cinosternon Effeldtii und einige andere neue oder weniger bekannte Amphibien. Monatsberichte der Königlichen Preussische Akademie des Wissenschaften zu Berlin, 1873:603-618.
- PINTO-SÁNCHEZ, N.R.; CRAWFORD, A.J. & WIENS, J.J. 2014. Using historical biogeography to test for community saturation. Ecology Letters, 17:1077-1085. DOI.
- RIVERA-CORREA, M. & DAZA, J.M. 2016. Molecular phylogenetics of the Pristimanstis lacrimosus species group (Anura: Craugastoridae) with the description of a new species from Colombia. Acta Herpetologica, 11:31-45. DOI.
- RODRÍGUEZ, A.S.; PILGRIM, J.D.; LAMOREUX, J.F.; HOFFMANN, M. & BROOKS, T.M. 2006. The value of the IUCN Red List for conservation. Trends in Ecology & Evolution, 21:71-76. DOI.
- STUART, S.N.; HOFFMANN, M.; CHANSON, J.; COX, N.; BERRIDGE, R.; RAMANI, P. & YOUNG, B. 2008. Threatened Amphibians of the World. Barcelona, Lynx Editions. 151p.
- WANG, I.J. 2013. Examining the full effects of landscape heterogeneity on spatial genetic variation: A multiple matrix regression approach for quantifying geographic and ecological isolation. Evolution, 67:3403-3411. DOI.
SUPPLEMENTARY MATERIAL
Voucher information of the Pristimantis penelopus specimens used in this study.
GenBank accession numbers for the COI fragment used in this study. Species from the Pristimantis ridens series were obtained from Crawford et al. (2013; http://onlinelibrary.wiley.com/doi/10.1111/1755-0998.12054/suppinfo).




