Conservation Biogeography of Amphibians in Mauritania: Integrating DNA Barcoding and Spatial Analyses for the Identification Of

Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Ana Marta Martinho Sampaio Mestrado em Biodiversidade, Genética e Evolução Departamento de Biologia 2016/2017

Orientador José Carlos Brito, Investigador Principal, CIBIO

Coorientador Guillermo Velo-Antón, PhD, CIBIO Todas as correções determinadas pelo júri, e só essas, foram efetuadas.

O Presidente do Júri,

Porto, ______/______/______

To my beloved parents, Ana and Luís.

FCUP i Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Acknowledgments

To the best supervising duo, José Carlos Brito and Guillermo Velo-Antón, cheers! It was an immense pleasure to work with two creative and committed researchers. I felt supported at every step of the process. I hope both of you see your work recognized and valued as it should. I wish you nothing but the very best.

To the CTM members and all the ones who gave me precious tips for grasping the ghost from the underworld of lab work: Susana Lopes, Diana Castro, Jolita Dilyte, Joana Veríssimo, particularly Patrícia Ribeiro and Sofia Mourão who for sure have a spot reserved in heaven.

Thank you, Filipa Martins, for shedding some light when it came to genetics. Years later, it’s great fun to see how much you’ve grown! I’m grateful to you and Cris for your immense hospitality during my early times in the Far North.

To Wolfgang Böhme and Alberto Sánchez-Vialas who contributed with samples. Thank you, Alberto, for the 48-hour taxonomic marathon.

To BIODESERTS, for all the incredible field and lab work that preceded this thesis and made this work possible. It has been a pleasure.

To all the people out there who create and maintain free servers and software for strangers like me to use, my sincerest acknowledgements. To Alexandra Elbakyan for founding and struggling to maintain the invaluable source of information that is Sci- Hub.

To my Porto crew: My flatmates and dearest friends, Claudia Chickpea Meneghesso and Fermin Boludo Otero, you filled my heart with joy during my stay in Porto. With you I laughed the perfect laughter. Thank you for taking such good care of me. Fleur Van deVille, for the tasty homemade warming meals and sweet sharing of stories, concerns and laughers. Miss Buchadas, apart from the coolest surname ever, you were my partner in crime for almost two years of holy moments. Thank you for the complicity. Thank you for Tarkovsky. FCUP ii Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

To my role models, who also happen to be my beloved friends. Andreia Penado, Ricardo Rocha, Diana Rodrigues, Filipa Machado, Cátia Caeiro, Inês Silva, Nídia Fernandes, Aurora Santos, Maggie Wood, Nahla Mahmoud, Ana Bontorim and Ipek Ozdemir I’ve been seeing you all blossoming into outstanding, strong, ethical-driven, genuine individuals. When I grow up I want to carry bits of you with me.

To Teresa Franco and Tiago Jorge, the guardians of my memories, we will get old together either you want it or not.

Thank you, family. Thank you, Salomé, although you still don’t understand it.

Funding for the collection of the samples used in the current M.Sc. thesis were provided by National Geographic Society (grants CRE-7629-04, CRE-8412-08, GEFNE-53-12), Mohammed bin Zayed Species Conservation Fund (grants 11052709, 11052707, 11052499, 13257467), Rufford Foundation (SG-15399-1), Fundação para a Ciência e Tecnologia (PTDC/BIA-BEC/099934/2008, PTDC/BIA-BIC/2903/2012), and FEDER through COMPETE Operational Programme for Competitiveness Factors (FCOMP-01-0124-FEDER-008917, FCOMP-01-0124-FEDER -028276).

FCUP iii Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Resumo

A biogeografia da conservação é uma disciplina que recorre a conceitos teóricos e a metodologias da biogeografia para responder a questões conservacionistas, o que lhe concede potencial para combater duas grandes lacunas de conhecimento da biodiversidade, denominadas em inglês, Linnaean shortfall e Wallacean shortfall. A primeira caracteriza o desfasamento que entre a diversidade que realmente existe e a que está formalmente descrita, enquanto que a segunda denota a escassez de conhecimento sobre a geografia da vida. O presente estudo tem como objetivo reduzir estas lacunas de conhecimento nos anfíbios na região de transição biogeográfica da Mauritânia. Utilizam-se dados resultantes do barcoding de ADN e da modelação ecológica espacial para responder às seguintes questões: 1) Quanta diversidade de anfíbios está presente na Mauritânia? 2) Como está essa diversidade geograficamente distribuída? 3) Quais os fatores ambientais que se correlacionam com a distribuição da diversidade de anfíbios? e 4) Onde estão localizados os corpos de água prioritários para a conservação dos anfíbios? Ao contrário da maioria dos estudos de barcoding em animais que se focam exclusivamente no marcador mitocondrial COI, adicionou-se uma segunda linha de evidência independente com o marcador RAG1 e identificámos 15 taxa. Registaram-se 14 taxa formalmente descritos com presença na Mauritânia e detetou-se potencial diversidade críptica em Hoplobatrachus occipitalis que revelou elevada diversidade intraespecífica. Destaca-se o primeiro registo para a Mauritânia, de dois novos géneros (Amnirana e Leptopelis) e duas novas espécies (Kassina fusca e Tomopterna milletihorsini). Construiu-se a primeira base de dados de barcoding para os anfíbios da Mauritânia a partir de 418 indivíduos sequenciados. Identificaram-se seis fatores ambientais correlacionados com a distribuição da riqueza de anfíbios, proximidade à savana e às zonas de várzea de cascalho e distanciamento das dunas e das várzeas de cascalho arenoso. Ao contrário das águas temporárias, as águas permanentes correlacionaram-se negativamente com a riqueza de anfíbios. Os resultados obtidos sublinham a importância dos habitats terrestres e da rede hidrográfica sazonal na manutenção da diversidade de anfíbios na Mauritânia. De acordo com o modelo ecológico, a riqueza de anfíbios segue um gradiente latitudinal, aumentando em direção ao sul, com localidades variando entre zero e 13 taxa. Dois pontos quentes (hotspots em inglês) de biodiversidade foram detetados no sudeste da Mauritânia, numa área rica em zonas húmidas. No entanto, estas zonas ricas em diversidade têm sido substituídas por agricultura, estão desprotegidas, e não estão contempladas na FCUP iv Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units rede de espaços protegidos atualmente em discussão para implementação na Mauritânia. Este estudo oferece uma primeira avaliação da diversidade de anfíbios na Mauritânia e destaca o uso do barcoding de ADN para inferência de padrões biogeográficos à escala regional. Palavras-chave: anfíbios, barcoding, biogeografia da conservação, diversidade críptica, modelação ecológica, Linnaean shortfall, RAG1, Saara, Sahel, SIG, Wallacean shortfall, zona de transição

FCUP v Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Abstract

By borrowing theoretical and methodological frameworks from biogeography to solve conservation problems, conservation biogeography holds the potential for challenging two major knowledge gaps in biodiversity: the Linnaean and the Wallacean shortfalls. The first stands for the mismatch between the existing and the formally described diversity whilst the second denotes the missing information on the geography of life. The present study aimed at reducing these knowledge shortfalls in the amphibians occurring in the biogeographic transition zone in Mauritania. Data derived from DNA barcoding and ecological spatial modelling were combined for addressing the following questions: 1) How much amphibian diversity is present in Mauritania? 2) How is this diversity geographically distributed? 3) Which environmental factors correlate with amphibian diversity distribution? and 4) Where are priority water-bodies for the conservation of amphibian diversity located? Unlike in most animal barcoding studies narrowed to the mitochondrial COI, a second independent line of evidence from the nuclear marker RAG1 was added and identified a total of 15 taxa. Fourteen formally described taxa were registered and potential cryptic diversity was detected in Hoplobatrachus occipitalis which displayed high levels of intraspecific diversity. This thesis provides the first record for Mauritania of two new genera (Amnirana and Leptopelis) and two new species (Kassina fusca and Tomopterna milletihorsini). The first DNA barcoding library for amphibians of Mauritania was constructed from 418 sequenced individuals. Six environmental correlates of amphibian richness distribution were identified, proximity to savannah and gravel floodplains, and greater distances to yellow dunes and gravel and sand floodplains. Permanent waters related negatively with amphibian richness as opposed to seasonal waters. Results obtained highlighted the importance of a suitable terrestrial habitat and an ephemeral hydrographic network for sustaining amphibian diversity in Mauritania. According to the predictive ecological model, amphibian richness increases southwards following a latitudinal gradient ranging from zero to 13 co-existing taxa per locality. Two major diversity hotspots were detected in south-eastern Mauritania, in a wetland-rich area. Yet, these biodiverse regions have been giving way to agriculture, and no protected areas are implemented nor are they under discussion for implementation in these areas. This study provides a preliminary assessment of amphibian diversity in Mauritania and highlights the use of DNA barcoding to infer biogeographic patterns at a regional scale. FCUP vi Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Keywords: amphibians, barcoding, conservation biogeography, cryptic diversity, ecological modelling, GIS, Linnaean shortfall, RAG1, transition zone, Sahara, Sahel, Wallacean shortfall

FCUP vii Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Table of contents

Acknowledgments ...... i Resumo ...... iii Abstract ...... v Table of contents ...... vii Table index ...... ix Figure index ...... x Abbreviation index ...... xii 1. Introduction ...... 1 1.1 Conservation biogeography ...... 1 1.1.1. Linnaean shortfall – DNA barcoding ...... 1 1.1.2. Wallacean shortfall – spatial ecological modelling ...... 5 1.2. Biodiversity in arid regions ...... 7 1.2.1. The transition zone in Mauritania ...... 8 1.2.2. Amphibians of Mauritania ...... 10 2. Objectives ...... 14 3. Methods ...... 15 3.1. Study area ...... 15 3.2. Fieldwork ...... 15 3.3. DNA extraction, amplification and sequencing ...... 17 3.4. Phylogenetic reconstruction...... 19 3.5. Sequence similarity analyses and barcoding gap ...... 19 3.6. Delimitation of phylogenetic units ...... 20 3.7. Spatial ecological modelling ...... 21 4. Results ...... 24 4.1. Phylogenetic reconstruction...... 24 4.2. Barcoding accuracy, genetic distances and barcoding gap ...... 28 4.3. Delimitation of phylogenetic units ...... 31 4.4. Spatial ecological modelling ...... 33 5. Discussion ...... 39 5.1. Systematics and DNA barcoding of amphibians of Mauritania ...... 39 5.1.1. Systematics ...... 40 5.1.2. Barcoding performance ...... 41 5.2. Amphibian diversity and conservation in Mauritania ...... 42 5.2.1. Environmental correlates of amphibian diversity ...... 42 FCUP viii Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

5.2.2. Predicting amphibian distribution ...... 44 5.3. Concluding remarks ...... 46 6. References ...... 47 Appendix ...... 68

FCUP ix Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Table index

Table 1. List of amphibian species recorded in Mauritania, their distribution and population status ...... 122 Table 2. Environmental variables used for developing ecological models of amphibian richness distribution in Mauritania...... 22 Table 3. Summary statistics of uncorrected pairwise distances within species and genera in COI and RAG1 ...... 28 Table 4. Coefficient values for all the predictors returned by the final model and cross- validation process relating amphibian richness with environmental predictors in Mauritania ...... 35 Supplementary Table 1. Amphibian samples amplified for at least one marker. ……68 Supplementary Table 2. Amphibian samples retrieved from GenBank for both markers...... 82 Supplementary Table 3. Correlation matrix between the 12 environmental variables...... 83 Supplementary Table 4. Support values for clusters obtained from GMYC single- threshold method applied to COI and RAG1 gene trees...... 855

FCUP x Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure index

Figure 1. Schematic representation of DNA barcoding gap (Meyer & Paulay, 2005). Intraspecific and interspecific variability are shown in red and yellow, respectively. (A) Discrete distributions without overlap, showing a barcoding gap. (B) Discrete distributions significantly overlapping...... 4 Figure 2. Predictive model building process (Guisan & Zimmermann, 2000)...... 6 Figure 3. Transition zone in Mauritania. Geographic context of the extent of Sahara and Sahel ecoregions in Africa. Distribution of the montane systems and transition between Sahara and Sahel in Mauritania...... 9 Figure 4. Distribution of the montane systems and hydrographic network in the study area...... 15 Figure 5. Study area depicting 61 sampling stations, 36 observation point localities and 418 sequenced individuals from 150 point localities...... 16 Figure 6. Identified and nonidentified taxa sampled...... 17 Figure 7. Distribution of random absences relative to the sampling stations across the study area...... 23 Figure 8. Phylogenetic reconstruction inferred from COI extended dataset I including sequences from GenBank (family names in grey)...... 25 Figure 9. Phylogenetic reconstruction inferred from RAG1 extended dataset I including sequences from GenBank (family names in grey)...... 26 Figure 10. Phylogenetic reconstruction inferred from COI and RAG extended datasets I including sequences from GenBank (family names in grey)...... 27 Figure 11. Line-plot of the barcoding gap for the 163 COI haplotypes...... 29 Figure 12. Line-plot of the barcoding gap for the 163 COI haplotypes after splitting H. occipitalis in two clusters...... 29 Figure 13. Line-plot of the barcoding gap for the 127 RAG1 haplotypes...... 30 Figure 14. Line-plot of the barcoding gap for the 127 RAG1 haplotypes after splitting H. occipitalis in two clusters...... 30 Figure 15. Phylogenetic reconstruction inferred from COI and RAG1 sequences for Mauritanian amphibians...... 32 Figure 16. Study area depicting species richness recorded in all sampling stations. .. 33 Figure 17. Predictors from the final model originated by data dredging and its relationship with the observed amphibian richness. R2 is shown...... 34 Figure 18. Predictive accuracy of the chosen model assessed by cross-validation. Relationship between predicted and observed amphibian richness returned by training FCUP xi Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units and validation subsets in ten replicates. Lines fitted by simple regression and respective R2 values are shown...... 36 Figure 19. Amphibian richness map of Mauritania and inset maps showing in detail: (A) Ngouye Classified Forest where the highest observed richness (8 species). (B) Detail of the Kolimbiné river floodplains with high predicted richness (C) Detail of Nioût river floodplains with high predicted richness. (D) Relationship between observed values and model predictions. R2 is shown...... 38 Supplementary Figure 1. Latest phylogenetic reconstruction for extant amphibians. Maximum likelihood tree built from 2781 species. Adapted from Pyron & Wiens, 2011…………………………………………………………………………………...... 81 Supplementary Figure 2. Spatial variation of the environmental variables in the study area ...... 84

FCUP xii Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Abbreviation index

AICc – corrected Akaike Information Criterion ANN – Artificial Neural Network bp – base pairs BEAST – Bayesian Evolutionary Analysis Sampling Trees BIC – Bayesian Information Criterion BIN –Barcode Index Number BLAST – Basic Local Alignment Search Tool BOLD – Barcode of Life Data Systems BRT – Boosted Regression Trees CBOL – Consortium for the Barcode of Life CIPRES – Cyberinfrastructure for Phylogenetic Research CITES – Convention on International Trade in Endangered Species of Wild Fauna and Flora COI – Cytochrome C oxidase subunit I dd – double-distilled DEM – Digital Elevation Model ESS – Effective Sample Size GEM – General Ecosystem Model GIS – Geographical Information System GMYC – Generalized Mixed Yule Coalescent GLM – Generalized Linear Model iBOL – International Barcode of Life IUCN – International Union for Conservation of Nature MaxEnt –Maximum Entropy MCMC – Markov Chain Monte Carlo MNDWI – Modified Normalised Difference Water Index mPTP – Multi-rate Poisson Tree Processes NDWI – Normalised Difference Water Index NUMT – Nuclear mitochondrial pseudogene PSC – Phylogenetic Species Concept RAG1 – Recombination activating gene 1 SPIDER – Species Identity and Evolution in R SPLITS – Species’ Limits by Threshold Statistics TRI – Terrain Ruggedness Index WGS – World Geodetic System

FCUP 1 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

1. Introduction 1.1 Conservation biogeography Conservation biogeography is a blossoming field aimed at addressing conservation concerns under the study of the dynamics of taxa diversity distribution (Richardson, 2012; Whittaker et al., 2005). This field has recently been substantially boosted by the advent of technical advances in Geographical Information System (GIS) data generation and analysis, and molecular biology (Dawson et al., 2016). These approaches allow investigating spatial patterns (e.g. GIS based modelling of species distributions) and historical processes of biological units, which can be objectively defined (Funk et al., 2012) and subsequently their past, present and future distributions more accurately predicted. Overall, researchers can quantitatively address which species are significantly threatened by extinction, and thereby evaluate conservation measures (Richardson, 2012; Whittaker et al., 2005). Conservation biogeographers are now equipped with an remarkable set of tools, but our knowledge on biodiversity is still plagued by inadequacies and gaps that bias the available data for accurately describing and predicting its distributional patterns (Hortal et al., 2015). Reducing such bias is of paramount importance for monitoring the biological effects of global change in the midst of an increasing extinction rate (Dirzo et al., 2014; Hortal et al., 2015; Stuart, 2004; Whittaker et al., 2005), so conservation actions can be executed where and when deemed necessary and appropriate. Among several key shortfalls in biodiversity data knowledge, two are the scope of conservation biogeography (Hortal et al., 2015; Whittaker et al., 2005). The Linnaean shortfall denotes the discrepancy between the numbers of species that actually exist and those yet to be formally described and catalogued (Brown & Lomolino, 1998). Although first and foremost a systematics enterprise, conservation biogeography is hampered by our very little understanding of the diversity itself, adding to the biased knowledge on the geography of living organisms, dubbed Wallacean shortfall (Lomolino, 2004). Technical developments are allowing the scientific community to accelerate the collection of data. The following sections aim at introducing two methodological approaches that can challenge these shortfalls.

1.1.1. Linnaean shortfall – DNA barcoding DNA barcoding emerges as a quick and cost-effective method for specimen’s identification and aiding species discovery, thereby reducing the magnitude of the FCUP 2 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Linnean shortall. It uses molecular markers to amplify short and highly variable DNA sequences (DNA barcode) from standardized regions of the genome, that identify each living organism (Hebert et al., 2003a). The idea was introduced in 2003 by Hebert and colleagues and proposed as a global method to identify animal species by creating cytochrome C oxidase subunit I (COI) profiles out of a ca. 650 bp fragment (Hebert et al., 2003a). The 5´ COI region was demonstrated to be sufficiently conserved within species while variable enough between them to provide for an accurate discrimination of most organisms (Hebert et al., 2003a, 2003b). Despite the fact that DNA barcoding advocated from the early beginning for the use of a global standard fragment, surrogate markers have been developed for many organisms where COI lacks resolution, particularly in plants, protists and fungi (Hajibabaei et al., 2007). DNA barcoding was also envisioned as part of a solution to overcome a decline of taxonomic expertise, increase the chances of dealing with morphological misidentifications, and resolve adult and larval stages within species (Hebert et al., 2003a). It is tackling the paucity of information in lesser known taxa, such as in marine crustaceans (Raupach et al., 2015), millipedes (Wesener & Conrad, 2016) algae (Leliaert et al., 2014), and fungi (Yahr et al., 2016). From a conservation perspective, it holds the potential for flagging endemic (Hosein et al., 2017) and cryptic diversity (Dincă, et al., 2011; Vasconcelos et al., 2016). It became the new method for the global inventory of life and continues to generate a growing body of literature (Hebert et al., 2016) as well as a substantial media coverage (Geary et al., 2016). Moreover, it became a tool for implementing citizen science and motivating students to pursue STEAM (science, technology, engineering, arts and mathematics) related careers (Henter et al., 2016). Its growing importance has led to the establishment of worldwide barcoding initiatives, such as the Cold Code (Murphy et al., 2013), a plan for barcoding amphibians and reptiles established under the umbrella of International Barcode of Life (iBOL) project [http://www.ibol.org]. iBOL works in collaboration with citizens, organisations and researchers around the world to build a publicly searchable comprehensive databank of DNA barcodes dubbed BOLD (Ratnasingham & Hebert, 2007) – Barcode of Life Data Systems [http://www.boldsystems.org]. Institutions investing in DNA barcoding can also join the Consortium for the Barcode of Life (CBOL) aimed at promoting and developing DNA barcoding through workshops, networks, conferences and training [http://www.barcodeoflife.org]. BOLD harbours DNA barcode reference libraries that are built from identified and vouchered specimens and aim at ensuring reproducibility, traceability and reliability of the taxonomic process (Hubert & Hanner, 2015; FCUP 3 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Ratnasingham & Hebert, 2007). It also provides a workbench for analytical procedures (Ratnasingham & Hebert, 2007). A pipeline has also been developed to cluster DNA sequences of closely related species deposited in BOLD into BINs (Barcode Index Numbers) for the entire animal kingdom (Ratnasingham & Hebert, 2013). A sample ID number and a taxonomic assignment are required to submit new sequences, but records must meet specific requirements to achieve a formal barcode status such as possessing a minimum length of 500 bp in the case of COI (Ratnasingham & Hebert, 2007).

At the heart of the DNA barcoding lies the assumption that a barcoding gap exists between two ranges of genetic variation: intraspecific polymorphism and interspecific divergence (Figure 1) (Hebert et al., 2003a, 2003b, 2004; Meyer & Paulay, 2005). In other words, the greatest intraspecific genetic distance will be lower than the smallest interspecific genetic distance. As such, accuracy, performance of a given barcode in discriminating species, will depend largely on the separation and extent of these two ranges of variability in the molecular marker (Meyer & Paulay, 2005). After retrieving the barcode from the specimen, its accuracy should be measured by comparing the query sequence to a comprehensive molecular databank in order to assign the unknown sampled specimen to a known species (Ratnasingham & Hebert, 2007). Because these reference libraries are still inexistent for most species, a plethora of methodological strategies are being employed/developed to accurately identify specimens and to assist species discovery: distance-based (Aliabadian et al., 2009; Vasconcelos et al., 2016), phylogeny-based (Vasconcelos et al., 2016), diagnostic (DasGupta et al., 2005) and statistical (Nielsen et al., 2006). Associated to these, species delimitation methods are commonly used in barcoding studies (Talavera et al., 2013; Zhang et al., 2013a). However, in incipient species that have not been yet fully separated by the coalescent, these two ranges of variability can overlap (Rosenberg, 2003) and therefore, no barcoding gap would be observed (Figure 1). Moreover, cases of paraphyly and polyphyly may arise due to hybridisation, introgression and incomplete lineage sorting following recent speciation (Funk & Omland, 2003), which introduces complexities when attempting to differentiate species through barcoding techniques since phylogeny-based methods require reciprocal monophyly. Absence of a barcoding gap, or reciprocal monophyly, can likewise result from misidentification (Mutanen et al., 2016), inaccurate reference taxonomy (Mutanen et al., 2016) undersampling (Meyer & Paulay, 2005) or presence of cryptic species (Mutanen et al., 2016). FCUP 4 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Researchers are now beginning to extend the number of independent molecular markers in DNA barcoding studies (Dupont et al., 2016; Simeone et al., 2013) allowing to tackle non-reciprocal monophyly issues (Funk & Omland, 2003; Vences et al., 2005a) and to move away from one of the major shortcomings of earlier barcoding studies, the assumption that gene trees and species trees are identical (Degnan & Rosenberg, 2009; Mallo & Posada, 2016).

From identifying singles specimens, DNA barcoding evolved to identifying entire communities through high-throughput DNA sequencing from mass collections of organisms or from environmental DNA, a technique called metabarcoding (Taberlet et al., 2012). Since metabarcoding demands a considerable investment to build high quality reference libraries, standard DNA barcoding and databases like iBOL with well- curated collections of specimens, still often offer the best solution for species identification in metabarcoding studies (Taberlet et al., 2012). The research community currently uses DNA barcoding and metabarcoding approaches in numerous ecological studies that go beyond taxonomy and biodiversity inventories, often revealing conservation and socio-economic applications (Adamowicz, 2015). Examples include border biosecurity such as detection of regulated pests (Hodgetts et al., 2016) and FCUP 5 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units invasive organisms (Armstrong & Ball, 2005), exposing and controlling illegal trade of CITES species (Williamson et al., 2016) and unregulated fishing (Helyar et al., 2014), study of food webs structure (McLean et al., 2016; Roslin et al., 2016; Smith et al., 2011), characterization of host-associated communities (Baker et al., 2016), and authentication of medicinal plants (Chen et al., 2014).

1.1.2. Wallacean shortfall – spatial ecological modelling Ecological spatial modelling is a mathematical modelling approach for understanding, predicting (current conditions) and forecasting (future conditions) ecological relationships in a geographical space, based on GIS and/or remote sensing data (Jørgensen & Fath, 2011). It helps disentangling biodiversity spatial patterns at different grain sizes (e.g. Brito et al., 2016; Vale et al., 2015) and allows to generate information on species occurrence data at a fine spatial grain otherwise hard to obtain and essential for informing conservation action (Jetz et al., 2012). In this regard, spatial ecological modelling offers a powerful tool for conservation biogeography to reduce the Wallacean shortfall.

As simplified versions of a complex reality, ecological mathematical models were borrowed from physical systems to synthetize our knowledge on the natural world (Jørgensen & Bendoricchio, 2001). They were first boosted by the increasing power of computer tools and later matured with the inclusion of fixed modelling procedures and more ecological knowledge (Jørgensen & Bendoricchio, 2001). Since then, different statistical and computational frameworks have been developed to grasp the overwhelming complexity of ecosystems in many different subfields of ecology and environmental management (Jørgensen & Fath, 2011). Spatial modelling has been profiting from these advances by incorporating a variety of techniques with examples including Artificial Neural Networks (ANN; Fischer, 2006), Generalized Linear Models (GLM) and extensions (Guisan et al., 2002), Maximum Entropy (MaxEnt; Phillips et al., 2006), Boosted Regression Trees (BRT; Friedman, 2001) and General Ecosystem Model (GEM; Harfoot et al., 2014).

Building a predictive spatial model begins with the identification of a problem or a natural system (Jørgensen & Fath, 2011) and, depending on the author, can be summarized in five major steps as schematised in Figure 2: (1) conceptual model, (2) statistical formulation, (3) calibration, (4) predictions and (5) evaluation. FCUP 6 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 2. Predictive model building process (Guisan & Zimmermann, 2000).

Conceptualization begins by laying a theoretical framework regarding the general patterns in the distribution of organisms (Guisan & Zimmermann, 2000). Then, it follows the choice on whether describing an empirical correlation (correlative models) or a causal link (mechanistic models) between two or more ecological variables. This stage implies understanding the trade-offs between the reality, precision and transferability (generalization) of the model’s predictive ability (Guisan & Zimmermann, 2000). An important part of the conceptualization step involves defining what to model: individual species and lineages (e.g. Rosauer et al., 2015) or properties of entire communities (Ferrier & Guisan, 2006). When modelling properties of communities, such as species richness, three distinct approaches can be considered: “assemble first, predict later” in which richness is estimated directly using predictors; “predict first, assemble later” in which richness is estimated from modelled species’ distributions; and “assemble and predict together” that implies modelling all species simultaneously (Ferrier & Guisan, 2006). Conceptualization ends with a set of meaningful predictive variables and an appropriate spatial scale (Guisan & Zimmermann, 2000). Ideally, the conceptual model should determine the collection of the data, but in reality, models are often developed from available data due to logistic and economic limitations (Jørgensen & Fath, 2011). The second step of the model building process leads to the translation of our knowledge on a natural system into a mathematical formula, through the selection of an optimal statistical approach (Guisan & Zimmermann, 2000). FCUP 7 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

The third step involves fitting the model to the data by comparing its output (prediction values) to the observations. It is aimed at increasing the predictive power and accuracy of the model. The initial set of predictors is often downsized which can be done automatically, for instance, through data dredging (Barton, 2016). So far, we have acquired an understanding of a certain natural system and formalized it as a model, that can be used for making predictions of current conditions. In spatial modelling, the forth step involves predicting the distribution of living organisms within the modelled area which can then, be mapped, for instance, by implementing the model in a GIS environment (Guisan & Zimmermann, 2000). The model building process ends with a final validation step aimed at evaluating the predictive power of the model, for example, through cross-validation. In other words, the goal is not to test if a model is true or false, but to confirm its behaviour under a range of conditions represented by the data (Guisan & Zimmermann, 2000).

Ecological spatial modelling has been used by both the research community and conservation practitioners to deal with a vast array of topics like invasive biology (Buchadas et al., 2017), climate change research (Penado et al., 2016), landscape connectivity (Roscioni et al., 2014), human-wildlife conflicts (Braunisch et al., 2011; Roscioni et al., 2014), ecological succession (Mittanck et al., 2014) and ecosystem services (Nelson et al., 2009).

1.2. Biodiversity in arid regions With extreme aridity indices (annual precipitation/potential evapotranspiration rates) falling below 0.20 and subjected to infrequent and intense pulses of precipitation (Ward, 2016), warm deserts and arid regions are harsh abiotic environments that push nature into unique adaptations to climate extremes (Brito et al., 2014). Precisely because of the possibility of already inhabiting close to physiological limits, desert- adapted species might be more strongly impacted by climate change (Vale & Brito, 2015). A scenario that is aggravated by the high vulnerability of drylands to global environmental changes (IPCC, 2014; Loarie et al., 2009). Among all arid regions, the Sahara-Sahel stands out with a set of idiosyncrasies potentially leading to cryptic diversity and localised biodiversity hotspots (Brito et al., 2014) critical for conservation (Brito et al., 2016; Vale et al., 2015). The boundary between the Sahara and the Sahel is situated in a crossroad between the Palaearctic and Afro-tropic terrestrial biogeographical realms (Dinerstein et al., 2017), thereby FCUP 8 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units comprising several distinct terrestrial ecosystems and phytogeographic regions (Sayre et al., 2013). This transition translates into high levels of species richness as expected whenever areas with distinct important biodiversity elements intersect (Spector, 2002). The Sahara-Sahel spreads over 10 countries that have been staging political and socio-economic instability in different degrees, often associated to terrorism (IEP, 2016). Besides the obvious tragic consequences for local populations that largely live in low human development (UNDP, 2016), the current situation is having a pervasive impact on wildlife and hampering basic scientific research (Brito et al., 2014; Durant et al., 2012).

1.2.1. The transition zone in Mauritania Mauritania lies in the transition zone between the Palaearctic and Afro-tropic realms. A system of mountains and plateaux with associated water-bodies surrounded by sand seas, typically punctuates the landscape (Brito et al., 2016; Le Houérou, 1997). Four major massifs, Adrar Atar in the Sahara, Tagant, Assaba and Afollé in the Sahel, constitute the mountain system of the country (Figure 3). Predominantly in the Saharan desert, these water-bodies occur essentially seasonally and isolated in the form of small mountain rocky pools locally known as gueltas (between 0.001 and 1.0ha) (Campos et al., 2012; Cooper et al., 2006). They are typically deep, located upstream mountain valleys and supplied by torrential rains during the short raining season between July and September (Brito et al., 2011; Cooper et al., 2006). As we move further south into the Sahel, the hydrographic network begins to be increasingly connected; riverbeds (wadis) and floodplains (tâmoûrts) become more common (Campos et al., 2012). Tâmoûrts form at the foothills of massifs, are more shallow than gueltas and therefore tend to dry outside the rainy season (Cooper et al., 2006), and are supplied by the wadis formed during torrential rainfalls (Campos et al., 2012). The region also displays a latitudinal variation in the species distribution patterns. The diversity of both fauna and flora is widespread in the south and increasingly confined and scattered towards north (Brito et al., 2014; Le Houérou, 1997). Mountains and associated hydrographic sub-basins (Figure 4) form localised biodiversity hotspots, where both endemic and relict taxa with several biogeographic affinities occur (Brito et al., 2014). Adrar Atar and Tagant are particularly rich, with the former hosting the highest number of Saharan vertebrate endemics (Brito et al., 2014). Particularly in the Sahara, many water dependent species persist in isolated populations within restricted habitats, mainly associated to gueltas that constitute refugia for both endemic and FCUP 9 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units range-limit populations (Brito et al., 2011, 2014; Trape, 2009; Vale et al., 2015; Velo- Antón et al., 2014) and where cryptic diversity is being uncovered (Brito et al., 2016). The Mauritanian hydrographic network has been suggested to play a crucial role for the persistence of crocodile (Crocodylus suchus) populations in the region, by allowing dispersal movements along the wadis connecting populations otherwise completely isolated (Velo-Antón et al., 2014). Extreme temperatures and droughts coupled with human disturbance are threatening these inland waters, yet, 80% of the gueltas identified as priority for conservation in Mauritania do not hold any legal protection status (Vale et al., 2015). Moreover, facing an accelerating water cycle, where wet areas become wetter and dry areas become drier (Sutherland et al., 2013), the Sahel is expected to experience amongst the highest drought recovery times of the planet (Schwalm et al., 2017). Species loss could, therefore, be more acute than hypothesized (Sutherland et al., 2013).

Figure 3. Transition zone in Mauritania. Geographic context of the extent of Sahara and Sahel ecoregions in Africa. Distribution of the montane systems and transition between Sahara and Sahel in Mauritania.

FCUP 10 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

1.2.2. Amphibians of Mauritania Amphibian distribution in Mauritania is widespread in the Sahel but rather restricted in the southern Sahara where population isolates are confined to the wadis, gueltas and springs, with important areas occurring in the mountains and some could temporarily connect during the rainy season (Padial et al., 2013). Because of this distribution pattern, amphibians occurring in the dryer Sahara are hypothesized to result from the dispersion of species coming from the Sahelian savannah. Priority areas for amphibian conservation are thought to be along the hydrographic network on the more explored massifs of Adrar Atar and Tagant (Padial et al., 2013). The last one lies midway between the two realms, exhibiting transitional environmental features and therefore, containing more suitable habitats for amphibians (Padial et al., 2013). On the other hand, the southern less known mountains of Assaba and Afollé, as well as the Senegal river basin might also prove to be equally important (Padial et al., 2013). Gueltas in Assaba recorded so far the highest amphibian richness of all four major massifs associated gueltas (Vale et al., 2015), meaning the unexplored associate river basins and tâmoûrts hold the potential of being equally rich. The undersampled savannah bordering Senegal and Mali in the south of the country is also expected to be diverse (Padial et al., 2013). Data on the distribution and diversity of amphibians of Mauritania has been highlighted as a priority research need for the region (Padial et al., 2013). Yet, only 12 anuran species have been reported in the country (Table 1), with the first record of a newly confirmed species (Ptychadena schillukorum) being from 2017 (Sánchez-Vialas et al., 2017). This suggests an underestimation of the Mauritanian amphibian diversity, and more species are hypothesized to exist given that large southern areas remain unexplored and taxonomy of these species is seldom reliable (Padial et al., 2013). Only three species have been recorded across the Saharan realm: Hoplobatrachus occipitalis, Sclerophrys xeros, and Tomopterna cryptotis (Table 1). Further south, the Sahel exhibits higher diversity and comprises all species present in the Sahara (Padial et al., 2013). Among the true toads, family Bufonidae, three species belonging to genus Sclerophrys are known to occur in Mauritania. Potential breeding sites of S. regularis include a vast variety of water-bodies, from permanent to temporary, from lotic to lentic water systems (Rödel, 2000). Reproduction in S. xeros and S. pentoni seems to be more associated to ponds instead (Rödel, 2000). H. occipitalis from the Dicroglossidae family, is essentially an aquatic species; its tadpoles are capable of surviving in shallow ponds at high temperatures and can prey on S. xeros tadpoles (Rödel, 2000). The ground-dwelling Kassina senegalensis is the sole representative of the hyper diverse FCUP 11 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units clade Afrobatrachia and so far, breeding is known to occur in lentic systems (Fleischack & Small, 1978). The species rich family of puddle frogs Phrynobatrachidae, counts with records of Phrynobatrachus natalensis, with breeding sites associated to lentic systems, mainly ponds and puddles (Rödel, 2000). Four species from the family Ptychadenidae have been recorded in Mauritania. They belong to the genus Ptychadena, that comprises species that frequently occur in syntopy (Bwong et al., 2009; Rödel, 2000) and commonly breed in lentic water-bodies with associated vegetation (Rödel, 2000). The family Pyxicephalidae is represented in Mauritania by two ground-dwelling species, Pyxicephalus edulis and Tomopterna cryptotis, the latter is thought to be mostly nocturnal (Rödel, 2000). To date, no species hold an unfavourable conservation status (Table 1), yet it could change as knowledge on the Saharan isolates is being gathered.

FCUP 12 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Table 1. List of amphibian species recorded in Mauritania, their distribution and population status.

Species Species Species Distribution and Population Status Taxonomic Status Family Notes (Rödel, 2000) (Padial et al., 2013) (IUCN, 2015) (Padial et al., 2013) (Rödel, 2000)

Sclerophrys Scattered localities across the Sahelian Bufonidae Bufo pentoni “Bufo” pentoni Stable - pentoni savannah; locally abundant

Scattered localities across the Sahelian Amietophrynus Sclerophrys Unstable; possible Bufo regularis savannah and along the coast; locally - regularis regularis species complex abundant

Amietophrynus Sclerophrys Unstable; possible Bufo xeros Most Saharan waterbodies; locally abundant - xeros xeros species complex

Most Saharan waterbodies and across the Hoplobatrachus Hoplobatrachus Hoplobatrachus Unstable; possible Dicroglossidae Sahelian savannah; locally abundant but - occipitalis occipitalis occipitalis species complex possible local extinctions have occurred

Kassina Kassina Kassina Scattered localities across the Sahelian Unstable; possible Hyperoliidae - senegalensis senegalensis senegalensis savannah; scarce species complex

Never reported; considered Hyperolius - - - - by IUCN as likely in nitidulus Guidimaka province Never reported; considered Phrynomantis Microhylidae - - - - by IUCN as likely in microps Guidimaka province Never reported; considered Phrynobatrachus Phrynobatrachidae - - - - by IUCN as likely along the francisci Senegal river

Phrynobatrachus Phrynobatrachus Phrynobatrachus Known only from a single locality in the Unstable; possible - cf. natalensis natalensis natalensis Sahelian savannah; no data on abundance species complex

Ptychadena Known from two localities in the Sahelian Ptychadenidae Ptychadena bibroni - Stable Excluded by IUCN bibroni savannah; no data on abundance

Ptychadena Ptychadena Ptychadena Known only from a single locality in the Unstable; possible - mascareniensis mascareniensis mascareniensis Sahelian savannah; no data on abundance species complex FCUP 13 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Species Species Species Distribution and Population Status Taxonomic Status Family Notes (Rödel, 2000) (Padial et al., 2013) (IUCN, 2015) (Padial et al., 2013) (Rödel, 2000) Never reported; considered Ptychadena - - - - by IUCN as likely along all pumilio southern Mauritania

Ptychadena Ptychadena Ptychadena Known from two localities in the Sahelian Stable - trinodis trinodis trinodis savannah; no data on abundance

Pyxicephalus Pyxicephalus Pyxicephalus Scattered localities across the Sahelian Unstable; possible Pyxicephalidae - edulis edulis edulis savannah; probably locally abundant species complex

Scattered localities across the Sahelian Tomopterna Tomopterna Tomopterna Unstable, possible savannah; also recorded from the coast and - cryptotis cryptotis cryptotis species complex in some Saharan waterbodies; rare

FCUP 14 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

2. Objectives

This study addresses the following questions: 1) How much amphibian diversity is present in Mauritania? 2) How is this diversity geographically distributed? 3) Which environmental factors correlate with amphibian diversity distribution? and 4) Where are priority water-bodies for the conservation of amphibian diversity located? Overall, this study will allow us to conduct a preliminary conservation assessment of Mauritanian amphibians by identifying hotspots of diversity. Our goal is to detect key locations for conservation of autochthonous amphibian species occurring in water-bodies of Mauritania by assessing the environmental correlates of their distributions.

Specifically, we aim to: (a) Identify amphibian species and possibly cryptic diversity, using two independent genetic markers, the mitochondrial COI and the nuclear RAG1; (b) Build a barcode library for the region providing a useful and standardized species identification tool, improving the poor taxonomic knowledge of Mauritanian amphibians; (c) Quantify amphibian richness and assess its distribution patterns; (d) Identify environmental correlates of amphibian richness distribution; (e) Identify diversity hotspots.

This work will contribute directly for building the first DNA barcode library for the amphibian species of the region and for an atlas of the distribution of the herpetofauna of Mauritania...... FCUP 15 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

3. Methods 3.1. Study area The study area comprises Mauritania, Northwest of Africa, where tissue samples have been collected from individuals captured along the hydrographic network, in lowlands and mountains (Figure 4). The mountain of Adrar Atar, Tagant, Afollé and Assaba harbour endemic and relict vertebrate and invertebrate taxa (Brito et al., 2014). The central-southern Mauritania lies within the Sahel ecoregion and exhibits increased water availability, which includes the Senegal river in the southern border of the country (Figures 3 and 4; Campos et al., 2012).

Figure 4. Distribution of the montane systems and hydrographic network in the study area.

3.2. Fieldwork A total of 418 samples were collected by researchers and collaborators of BIODESERTS – Biodiversity of Deserts and Arid Regions – research group during 16 field expeditions between 2003 and 2016 (Appendix: Supplementary Table 1). Amphibians were sampled using dip-nets, from across the Sahara-Sahel. Tissue samples were collected by toe and tail-clipping of adult specimens and tadpoles respectively, and stored them in 95% ethanol. Locality data for all samples was geo- referenced in the field with a Global Positioning System (GPS) on the WGS84 datum. Samples were collected for genetic analyses over 195 point localities, registered observations and collect specimens over 38 point localities (Figure 5). For ecological FCUP 16 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units modelling, we used richness values from 61 sampling stations located in Mauritanian selected taking into account a similar sampling effort (Figure 5). Whenever possible, specimens were assigned to formally described species through careful examination of vouchers, photographs and location data. For this purpose, we resorted to taxonomic expertise which allowed for the morphological identification of the diagnostic characters. We followed the latest amphibian phylogenetic reconstruction from Pyron and colleagues (Appendix: Supplementary Figure 1; Pyron & Wiens, 2011) for reference and identified 13 formally described amphibian species plus one identified only at the genus level (Phrynobatrachus sp.) (Figure 6). Our dataset included putatively seven species from five Afro-tropic endemic genera: Kassina (K. fusca and K. senegalensis) Leptopelis (L. bufonides), Phrynobatrachus sp., Ptychadena (P. bibroni, P. schillukorum and P. trinodis), Pyxicephalus (P. edulis) and Tomopterna (T. milletihorsini) (Frost et al., 2006; Pyron & Wiens, 2011; Stuart et al., 2008). Non- endemic genera in the Afro-tropics occurring in Mauritania include Hoplobatrachus (H. occipitalis) (Stuart et al., 2008) and Sclerophrys (S. pentoni, S. regularis and S. xeros) (Frost, 2017). Amnirana genus is the only member of the essentially Palaearctic family Ranidae occurring in the Afro-tropic realm (Stuart et al., 2008) and is represented in Mauritania by A. galamensis.

Figure 5. Study area depicting 61 sampling stations, 36 observation point localities and 418 sequenced individuals from 150 point localities.

FCUP 17 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 6. Identified and nonidentified taxa sampled.

3.3. DNA extraction, amplification and sequencing We performed DNA extraction and purification from tissue samples using Easy Spin extraction and purification kit. We verified the quality and approximate quantity by electrophoresis in TBE (Tris-Borate-EDTA buffer) 0.5x at 300V, using 0.8% agarose gel stained with gel red (Biotium). We then visualised the gel through UV radiation in a BioRad Universal Hood II Quantity One 4.4.0. Whenever necessary, we diluted the extractions by adding ultra-pure water before proceeding to PCR. The COI marker was ampliﬁed using the primer pair Chmf4 (5′-TYT CWA CWA AYC AYA AAG AYA TCG C- 3′) and Chmr4 (5′-ACY TCR GGR TGR CCR AAR AAT CA-3′) (Che et al., 2012). We selected a second slower evolving marker (San Mauro et al., 2004a), the nuclear recombination-activating gene 1 (RAG1), frequently used in amphibian phylogenetic FCUP 18 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units reconstructions and proven to be useful at inter-specific levels (Frost et al., 2006; Hoegg et al., 2004; Irisarri et al., 2012; Portik & Blackburn, 2016; Pyron & Wiens, 2011; Roelants et al., 2007; San Mauro et al., 2004; van der Meijden et al., 2004; Wiens, 2007). RAG1 was amplified using Amp-RAG1 F (5′-AGC TGC AGY CAR TAC CAY AAR ATG TA-3′) and Amp-RAG1 R1 (5′-AAC TCA GCT GCA TTK CCA ATR TCA CA- 3′) (San Mauro et al., 2004b). We performed the PCR of both genes in a 10-µl volume reaction containing 1µl of DNA, 3µl of ddH2O, 5µl of Taq polymerase and 0.5µl of each primer. PCR conditions for COI (a) and RAG1 (b) respectively, were as follows: (a) initial denaturation step with 5 min at 95ºC; 35 cycles of denaturation for 1 min at 94ºC, annealing for 1 min at 52ºC and extension for 1 min at 72ºC; final extension at 72ºC for 10min; (b) initial denaturation step with 5 min at 94 ºC; 35 cycles of denaturation for 1 min at 94ºC, annealing for 1 min at 54ºC and extension for 1.30 min at 72ºC; final extension at 72ºC for 7 min. We confirmed the presence of the PCR products by 2% gel electrophoresis. Purification and Sanger Sequencing protocols of PCR products were outsourced to GeneWiz. We inspected and verified the sequence chromatograms and alignments using Geneious v.4.8.5 (http://www.geneious.com/; Kearse et al., 2012). Aside from a gap penalty set at 100, we used the default settings of the Geneious alignment algorithm to produce global alignments for each gene. To detect the presence of nuclear mitochondrial pseudogenes (NUMTs), we searched for ambiguities and translated all aligned sequences into amino acids and looked for stop codons and frameshift mutations (Bensasson et al., 2001). RAG1 polymorphic sites were encoded with the IUPAC ambiguity code. We submitted the sequences to a BLAST (Basic Local Alignment Search Tool) in GenBank and BOLD databases, to confirm that the targeted regions had been amplified. We sequenced all the 418 samples with 606bp for the mitochondrial COI and selected 187 nuclear sequences with 854bp in order to represent all mitochondrial lineages (Appendix: Supplementary Table 1). We identified haplotypes with FaBox v.1.41 (Villesen, 2007) and removed sequences from the subsequent analyses for computational reasons. We then built two datasets for each marker: i) 186 COI and 151 RAG1 haplotypes from samples collected throughout the Sahara-Sahel, herein datasets I, and ii) 163 COI and 127 RAG1 haplotypes from samples collected only in Mauritania, herein datasets II.

FCUP 19 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

3.4. Phylogenetic reconstruction We used datasets I and extended it to included 21 species retrieved from GenBank, representing families missing in the study area, following the phylogenetic reconstruction from Pyron and colleagues (Pyron & Wiens, 2011) (Appendix: Supplementary Table 2). The aim was to obtain a better resolved topology, for solving and comparing phylogenetic relationships between species. We employed a Bayesian inference approach implemented in BEAST v.2.4.6 (Bouckaert et al., 2014) provided by CIPRES science gateway v.3.3 (Miller et al., 2010) to construct all trees. We applied the closest nucleotide substitution models available in BEAST as suggested by ModelFinder (Kalyaanamoorthy et al., 2017) implemented in IQ-Tree web server (Trifinopoulos et al., 2016), and selected according to BIC (Bayesian Information Criterion). We diagnosed chains convergence by inspecting the trace plots and effective sample sizes (ESS) of the parameters in Tracer v.1.6 (Rambaut et al., 2003) and combined the trees with LogCombiner 2.4.5 and TreeAnnotator 2.4.5. We used Figtree v.1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/) to visualize and edit consensus trees. As some of the subsequent methods require a corrected rooted tree, we chose Ascaphus truei as an outgroup taxon (Irisarri et al., 2012). We obtained a gene tree for each marker. We selected the following settings (otherwise by default): COI tree site model TN93+I+G and RAG1 tree site model HKY+I+G; strict clock model; Yule tree prior; ingroup enforced monophyly; MCMC length of 1000 000 000 steps and logging parameters every 10 000 000 step, combining three independent runs and a 10% burn-in. We ran an analysis with a partitioned dataset including COI and RAG1 haplotypes and selected the following settings (otherwise by default): COI tree site model HKY+I+G and RAG1 tree site model TVMe+I+G; strict clock model; Yule tree prior; ingroup enforced monophyly; MCMC length of 100 000 000 steps and logging parameters every 10 000 000 step, combining three independent runs and a 10% burn-in. Although not appropriate for inter-species datasets, strict clocks were enforced in these analyses to reach chain convergence.

3.5. Sequence similarity analyses and barcoding gap The following analyses are implemented in the R (R Core Team, 2017) package SPIDER v.1.3 (Brown et al., 2012) and were run on datasets II. We produced an uncorrected pairwise distance matrix for each marker, following previous recommendations on model choices for constructing matrices from low genetic FCUP 20 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units distances (Nei & Kumar, 2000). We then performed two query identification analyses to evaluate the identification performance of our COI barcodes, and to distinguish between successful, ambiguous, misidentified sequences or no match (Meier et al., 2006): Meier's best close match and BOLD identification criteria. Both imply the use of distance thresholds that we defined as 10%, as previously proposed for amphibians (Vences et al., 2005a). Meier's best close match finds the closest individual to the query sequence (Austerlitz et al., 2009; Meier et al., 2006) while BOLD identification criteria emulates the method of specimen identiﬁcation used by BOLD. To evaluate the presence of barcoding gap at species and genus level in both COI and RAG1 datasets, we used the statistics maxInDist (furthest intraspecific distance) and nonConDist (smallest interspecific distance) rather than mean distances which can overestimate it (Meier et al., 2008). If the difference between the furthest intraspecific distance and the smallest interspecific distance results positive, there is a barcoding gap, otherwise it reflects an overlap between the two ranges of distances.

3.6. Delimitation of phylogenetic units To delimit putative species, we applied single-locus species delimitation methods to the gene trees for identifying discrete genetic clusters, following the phylogenetic species concept (PSC; Eldredge & Cracraft, 1980). We used the Generalized Mixed Yule Coalescent (GMYC) model (Pons et al., 2006) implemented in the R package SPLITS (Ezard et al., 2009), which identifies the transition between a Yule-type speciation (inter-species) and a coalescent (intra-species) branching rates using an ultrametric tree by maximizing the likelihood score of the model (Fujisawa & Barraclough, 2013). We ran GMYC single-threshold as the multiple-threshold approach generally yields a lower taxonomic accuracy (Fujisawa & Barraclough, 2013). We also employed the multi-rate Poisson tree processes (mPTP) model (Kapli et al., 2017) implemented in MPTP Web Server (http://mptp.h-its.org/#/tree; Zhang et al., 2013), which relies on the branch lengths to infer putative species boundaries on a given phylogenetic input tree. Contrary to the previous version, the single-rate Poisson tree processes (PTP) model, it assumes different evolution rates for different species (Kapli et al., 2017; Zhang et al., 2013a). These methods were applied in single gene trees obtained for each marker from datasets II. We selected the following settings (otherwise by default): COI tree site model TN93+I+G and RAG1 tree site model K2P+G; strict clock model; Yule tree prior; ingroup enforced monophyly; MCMC length of 1000 000 000 steps and logging FCUP 21 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units parameters every 10 000 000 step, combining three independent runs and a 10% burn- in.

3.7. Spatial ecological modelling Using the distinct taxa identified in the precious sections, we applied spatial ecological modelling to identify environmental correlates of amphibian richness distribution, predict its patterns within the study area and finally map the predictions by implementing the model in a GIS environment. We obtained grid-based datasets for current environmental conditions in the study area comprising topographical variables, land cover types and water indexes. The datasets were downsized to include 12 uncorrelated (R < 7; Appendix: Supplementary Table 3) environmental variables (Table 2; Appendix: Supplementary Figure 2). We applied the terrain function in R package raster v.2.5-8 in a DEM (Hijmans, 2016) to a Digital Elevation Model (DEM) to derive the Terrain Ruggedness Index (Riley et al., 1999) with a 90 m resolution. We obtained eight types of land cover with a 30 m resolution from a remote sensing-derived land cover map (Campos et. al. unpublished data): std_d_lc01, std_d_lc02, std_d_lc04, std_d_lc05, std_d_lc06, std_d_lc08, std_d_lc11, std_d_lc12. and std_d_guel. A last land cover variable with 90 m resolution, std_d_guel was obtained from GPS (WGS84 datum) recorded localities and areas during field missions. The water indexes std_MNDWI and std_NDWI1 were derived from remote sensing and had a resolution of 30 m (Campos et al., 2012). We chose 12 environmental variables according to the spatial auto-correlation coefficients, the spatial scale, the present-day knowledge on the Mauritanian amphibians’ natural history traits and on the regional climate and habitats to avoid overfitting caused by an excess of model parameters. As such, some correlations above 0.7 were allowed, given the likely importance for the distribution of amphibians (Appendix: Supplementary Table 3). To avoid problems caused by negative values and by the differential weight arising from different units of measure, we applied a Min-Max standardization technique to all variables, scaling the data between zero and one: zi =

th th (xi - minx) / (maxx - minx); where zi denotes the i normalized data, xi the i observation, and maxx and minx the maximum and minimum value respectively. All variables were upscaled to 90 m and projected in the WGS84 datum. These analyses were performed in ArcMap v.10.1 (ESRI, 2012).

FCUP 22 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Table 2. Environmental variables used for developing ecological models of amphibian richness distribution in Mauritania.

Standardized Code Type Description Variables

Terrain Ruggedness Measure of variation in adjacent cells of a digital altitude std_TRI Topographic Index grid.

Distance to yellow Frequently fixed dunes composed by yellow sand. Sparse std_d_lc01 Land cover dunes shrubs with isolated Acacia sp. or with no vegetation cover.

Distance to white Mobile dunes composed by white sand (e.g. barchan std_d_lc02 Land cover dunes dunes). No vegetation cover.

Distance to compact Large flat areas composed by consolidated sandy soils. std_d_lc04 Land cover sand Presence of shrubs and sparse trees (e.g. Acacia sp.).

Usually found in large floodplains, in transitions between Distance to gravel + std_d_lc05 Land cover rocky and sandy soils. Soil composed by similar amounts of sand floodplains gravel and sand and with sparse or no vegetation cover.

Distance to gravel Large floodplains covered by gravel (locally known as Reg) std_d_lc06 Land cover floodplains and usually with no vegetation cover.

Non-flat areas with soils composed by stones, silt and/or Distance to std_d_lc08 Land cover clay. Usually associated to mountain regions and with rocky soil sparse or no vegetation cover.

Flat flooding areas covered by grasses (e.g. Cenchrus Distance to grasslands std_d_lc11 Land cover biflorus) and mostly associated to Sahelian regions.

High vegetation cover (grasses, shrubs and trees) and Distance to savannah std_d_lc12 Land cover mostly found in the West Sudanian Savannah.

Distance to gueltas std_d_guel Land cover Mountain rock pools.

Modified normalised std_MNDWI Water index Detects mostly permanent water-bodies. difference water index

Normalised difference std_NDWI1 Water index Detects mostly seasonal water-bodies. water index

We built a generalized linear model (GLM) with a gaussian distribution in R and obtained a first global model for species richness. For this purpose, we generated random absences after applying a 200 km buffer around the sampling stations. We then fitted all possible simple models using the dredge function from the MuMIn v.1.15.6 (Barton, 2016) R package and ranked them following the Akaike information criterion with the correction for small sample sizes (AICc) and selected the model with lower AICc. We performed cross-validation to study model stability and predictive performance (Efron & Gong, 1983; Snee, 1977; Zhang, 1997). We randomly partitioned the dataset into a training subset consisting approximately of 4/5 of the samples, used to build the model following the same procedure as before and to estimate the coefficients; and a validation subset with the remaining samples, used to run the model built in the previous step and to evaluate its predictive ability. Each subset contained equal number of presences and absences. We replicated this process 10 times. Selection criteria like AICc and cross-validation techniques such as data splitting FCUP 23 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units approaches also help avoiding overfitting as they tend not to favour such models (Burnham & Anderson, 2004; Olden & Jackson, 2000). We then searched for influential observations and repeated the model building and model validation processes without those to obtain the final best fitted model for 61 sampling localities and 61 random absences (Figure 7). The chosen GLM was imported into ArcMap to predict amphibian richness distribution in the whole study area.

Figure 7. Distribution of random absences relative to the sampling stations across the study area.

FCUP 24 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

4. Results 4.1. Phylogenetic reconstruction All genera were recovered as monophyletic clades. The 14 formally described taxa formed highly supported monophyletic clades in all trees (Figure 8 to Figure 10). COI produced conflicting topologies with both RAG1 and COI+RAG1 partitioned dataset, but all species clades in COI gene tree are confirmed by the two other reconstructions. All topologies divide most of the sequences in two major clades (Figure 8 to Figure 10). Contrary to the mitochondrial reconstruction, the combined COI+RAG1 and RAG1 topologies returned Leptopelis bufonides as sister clade of genus Kassina, and Tomopterna and Pyxicephalus, from the family Pyxicephalidae, as a strongly supported clade (Figure 8 to Figure 10). The COI+RAG1 and mitochondrial marker split Hoplobatrachus occipitalis in two well supported clades (Figure 8 to Figure 10), confirmed by RAG1 but with lower support values (PP = 0.92 for both groups; Figure 9).

FCUP 25 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 8. Phylogenetic reconstruction inferred from COI extended dataset I including sequences from GenBank (family names in grey). Outgroup is not shown. Support values are provided as Bayesian posterior probabilities above branches (PP: * ≥ 0.99; ° 0.90 – 0.99). FCUP 26 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 9. Phylogenetic reconstruction inferred from RAG1 extended dataset II including sequences from GenBank (family names in grey). Outgroup is not shown. Support values are provided as Bayesian posterior probabilities above branches (PP: * ≥ 0.99; ° 0.90 – 0.99).

FCUP 27 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 10. Phylogenetic reconstruction inferred from COI and RAG extended datasets I including sequences from GenBank (family names in grey). Outgroup is not shown. Support values are provided as Bayesian posterior probabilities above branches (PP: * ≥ 0.99; ° 0.90 – 0.99). FCUP 28 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

4.2. Barcoding accuracy, genetic distances and barcoding gap Total intraspecific genetic distances ranged from zero to 18.10% for COI and from zero to 1.65% for RAG1. Interspecific genetic distances ranged between 13.53% and 22.60% for COI sequences, and between 1.79% and 13.76% for RAG1. Genetic distances varied greatly depending on the taxa, comparatively high levels of intraspecific distances can be observed for COI in Hoplobatrachus occipitalis (Table 3). Further splitting of H. occipitalis in two clusters according to the gene trees resulted in a decrease of intraspecific distances, ranging between zero and 1.25% in COI-based cluster 1, zero and 0.84% in COI-based cluster 2; zero and 0.12% in RAG1-based cluster 1, zero and 0.35% in RAG1-based cluster 2. Minimum interspecific genetic distance falls to 12.44% in COI and 0.12% in RAG1 after considering two H. occipitalis lineages.

Table 3. Summary statistics of uncorrected pairwise distances within species and genera in COI and RAG1: minimum (Min.), maximum (Max.), mean and standard deviation (std). * Species represented by ≤ 2 sequences in one or both genes.

P-distances % COI P-distances % RAG1 Genus/Species Min. Mean ± std Max. Min. Mean ± std Max.

Amnirana galamensis* ------

Hoplobatrachus occipitalis 0.00 9.03 ± 8.48 18.10 0.00 0.21 ± 0.16 0.60

Kassina fusca* 0.00 0.35 ± 0.23 0.71 - - -

Kassina senegalensis 0.00 0.58 ± 0.24 1.09 0.00 0.04 ± 0.06 0.12

Leptopelis bufonides* ------

Phrynobatrachus sp. 0.00 1.26 ± 0.93 2.87 0.00 0.57 ± 0.34 1.65

Ptychadena bibroni 0.00 0.48 ± 0.24 0.83 0.24 0.59 ± 0.27 0.83

Ptychadena schillukorum 0.00 0.33 ± 0.13 0.66 0.00 0.11 ± 0.11 0.47

Ptychadena trinodis 0.00 0.29 ± 0.14 0.54 0.00 0.24 ± 0.17 0.71

Pyxicephalus edulis* ------

Sclerophrys pentoni 0.00 0.34 ± 0.16 0.55 0.00 0.23 ± 0.18 0.82

Sclerophrys regularis* - - - - -

Sclerophrys xeros 0.00 0.43 ± 0.15 0.71 0.00 0.10 ± 0.11 0.47

Tomopterna milletihorsini 0.00 0.38 ± 0.19 0. 89 0.00 0.09 ± 0.08 0.24

An assessment of the barcoding gap in COI did not reveal any overlap between intra- and interspecific distances. H. occipitalis displayed comparatively small barcoding gaps, ranging between 1.93% and 2.65% (Figure 11); which increased significantly after splitting the individuals again in two lineages according to the previous FCUP 29 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units phylogenetic reconstructions, falling now between 16.01% and 17.06% (Figure 12). For the remaining species represented by more than one individual, barcoding gaps ranged as follows: A. galamensis [20.35%], K. fusca [13.50%, 14.44%]; K. senegalensis [12.44%, 14.31%]; Phrynobatrachus sp. [17.26%, 18.30%]; P. bibroni [17.82%, 18.46%]; P. schillukorum [18.98%, 19.53%]; P. trinodis [18.12%, 18.82%]; S. pentoni [16.76%, 17.12%]; S. xeros [16.43%, 16.96%]; T. milletihorsini [21.06%, 21.89%] (Figure 11).

Figure 11. Line-plot of the barcoding gap for the 163 COI haplotypes. No overlap is observed and all specimens display a barcoding gap. The furthest intraspeciﬁc distance is the bottom of line value, whereas the smallest interspecific distance is top of line value.

Figure 12. Line-plot of the barcoding gap for the 163 COI haplotypes after splitting H. occipitalis in two clusters. No overlap is observed and all specimens display a barcoding gap. The furthest intraspeciﬁc distance is the bottom of line value, whereas the smallest interspecific distance is top of line value.

For the species represented by more than one individual, barcoding gap tested with RAG1 revealed gaps ranging as follows: H. occipitalis [8.02%, 9.16%]; K. fusca [1.67%, 2.00%]; Phrynobatrachus sp. [8.48%, 9.33%]; P. bibroni [4.76%, 5.53%]; P. schillukorum [1.77%, 1.90%]; P. trinodis [1.53%, 1.89%]; S. pentoni [1.64%, 2.11%]; S. FCUP 30 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units xeros [2.12%, 2.00%]; T. milletihorsini [12.22%, 12.38%] (Figure 13). Splitting H. occipitalis in two lineages resulted in a partial overlap [0.00%, 0.12%] (

Figure 14).

Figure 13. Line-plot of the barcoding gap for the 127 RAG1 haplotypes. No overlap is observed and all specimens display a barcoding gap. The furthest intraspeciﬁc distance is the bottom of line value, whereas the smallest interspecific distance is top of line value.

Figure 14. Line-plot of the barcoding gap for the 127 RAG1 haplotypes after splitting H. occipitalis in two clusters. For each specimen displaying a barcoding gap, the furthest intraspeciﬁc distance is the bottom of line value, whereas the smallest interspecific distance is top of line value. The lines representing an overlap between the two ranges of distances show where this relationship is reversed and are represented in grey.

Both query identification analyses and threshold values produced the exact same results when applied to the COI (Figure 15). Apart from the singletons, all individuals were successfully identified. Sequences belonging to H. occipitalis were always correctly identified independently of being split in two clusters or merged under the same name. The remaining specimens did not show discordances with formally described species.

FCUP 31 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

4.3. Delimitation of phylogenetic units GMYC and mPTP returned the same clusters when applied to COI gene tree (Figure 15), and suggested 15 groups plus one singleton, partitioning H. occipitalis and Phrynobatrachus sp.. GMYC and mPTP results based on RAG1 gene tree also agreed and suggested 9 groups and four singletons. All GMYC clusters are highly supported (> 0.90) except Phrynobatrachus sp. cluster derived from the nuclear marker (see Appendix: Supplementary Table 4).

FCUP 32 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 15. Phylogenetic reconstruction inferred from COI and RAG1 sequences for Mauritanian amphibians. Outgroup is not shown. Support values are provided as Bayesian posterior probabilities above branches (PP: * ≥ 0.99; ° 0.90 – 0.99). Species delimitations from distance-based query identification analyses, single threshold GMYC and mPTP. Discordances with formally described species are highlighted in orange. Singletons represented by . FCUP 33 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

4.4. Spatial ecological modelling For the purpose of modelling amphibian richness, we considered the following taxa: A. galamensis, H. occipitalis 1, H. occipitalis 2, K. fusca, K. senegalensis, L. bufonides, P. edulis, Phrynobatrachus sp., P. bibroni, P. schillukorum, P. trinodis, S. pentoni, S. regularis, S. xeros and T. milletihorsini. Observed richness ranged between one and eight taxa, with most of the sampling stations registering two or four taxa (Figure 16). The most northern stations located in Adrar Atar as well as northern Tagant registered maximum two taxa also detected in Sahel stations: H. occipitalis 1 and S. xeros. Stations registering three or more entities are located south of the Tagant (Figure 16).

Figure 16. Study area depicting species richness recorded in all sampling stations.

Data dredging originated a model with six environmental variables that better explained the distribution of amphibian richness in Mauritania (Table 4; Figure 17). Higher distances to yellow dunes (std_d_lc01) and to gravel and sand floodplains (std_d_lc05) relate to higher amphibian richness whereas higher distances to savannah (std_d_lc12) and gravel floodplains (std_d_lc06) relate to lower amphibian richness. MNDWI relates negatively with amphibian richness as opposed to NDWI1. All variables displayed non-linear relationship with the observed richness (Figure 17).

FCUP 34 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 17. Predictors from the final model originated by data dredging and its relationship with the observed amphibian richness. R2 is shown.

With regards to model stability, the exact combination of predictors obtained in the final model was returned in three replicates during the cross-validation process (Table 4). Distance to savannah (std_d_lc12) was returned as significant in all replicates, being the most consistent predictor variable. The two water indices, MNDWI and NDWI, distance to gravel floodplains (std_d_lc06), and distance to gravel + sand floodplains (std_d_lc05) were picked by cross-validation replicates nine out of ten times, although std_d_lc05 was significant in five replicates. Distance to yellow dunes (std_d_lc01) appeared in four replicates. Moreover, distance to gravel floodplains is recovered as a significant predictor in only one replicate (Table 4). Simple regression slopes and R2 values were similar within each replication pair, meaning the analyses performed with the training subsets were validated by respective validation subsets (Figure 18).

FCUP 35 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Table 4. Coefficient values for all the predictors returned by the final model and cross-validation process relating amphibian richness with environmental predictors in Mauritania. Significance: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Final Predictors Rep 1 Rep 2 Rep 3 Rep 4 Rep 5 Rep 6 Rep 7 Rep 8 Rep 9 Rep 10 model std_TRI std_d_lc01 2.62*** 2.17*** 2.13 2.58** 2.35** std_d_lc02 1.75** 2.17*** 1.93** 2.05** 1.48 1.87*** 2.62** std_d_lc04 std_d_lc05 2.44* 2.39 2.60* 2.47* 2.42 2.79* 1.89 3.34* 2.19 2.64* std_d_lc06 -3.96*** -3.73** -4.19*** -4.04** -3.77** -4.25*** -3.04** -4.65** -3.78** -4.12*** std_d_lc08 std_d_lc11 -1.49** std_d_lc12 -4.18*** -3.68*** -4.66*** -3.60*** -3.57*** -4.25*** -4.26*** -2.86*** -3.68*** -4.64*** -4.25*** std_d_guel -1.16* std_MNDWI -2.85** -4.51** -10.43* -15.23** -11.43* -11.10* -14.46** -4.53** -10.14* -11.81* std_NDWI1 19.18** 25.06** 18.51** 24.74*** 16.94* 16.11* 23.56*** 23.21** 15.96* 18.32**

FCUP 36 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 18. Predictive accuracy of the chosen model assessed by cross-validation. Relationship between predicted and observed amphibian richness returned by training and validation subsets in ten replicates. Lines fitted by simple regression and respective R2 values are shown. Models built with the training subset are represented in blue and models ran for validation are represented in grey. FCUP 37 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

According to the final model, richness increases southwards following a latitudinal gradient ranging from zero to 13 taxa, but most areas tend to accumulate up to seven taxa. Half of the study area is not expected to harbour any amphibian (Figure 19). Predictions showed some consistency with observed values (R2 = 0.71; Figure 19). Most of the southern half is expected to harbour one to three taxa (Figure 19), with localized small clusters of five or more taxa, as for instance the areas near the Senegal river basin (e.g. Figure 19 (A)). The Adrar Atar and northern Tagant also tend to exhibit this clustering pattern, but reaching maximum three and four/five taxa respectively (Figure 19). Two main diversity hotspots are predicted in the southern border of the country (Figure 19 (B) (C)). FCUP 38 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Figure 19. Amphibian richness map of Mauritania and inset maps showing in detail: (A) Ngouye Classified Forest where the highest observed richness (8 species). (B) Detail of the Kolimbiné river floodplains with high predicted richness (C) Detail of Nioût river floodplains with high predicted richness. (D) Relationship between observed values and model predictions. R2 is shown. FCUP 39 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

5. Discussion

This last segment is structured in two major sections corresponding to the two methodological approaches here employed aimed at challenging the Linnaean and the Wallacean shortfalls in the amphibians of Mauritania. Both sections pinpoint unsolved problems and questions worthy of further research. In section 5.1, we confirm species identification of all formally described taxa. We continue by exploring the complementarity of the information drawn from the two independent markers by analysing the phylogenetic robustness and the outcomes of different barcoding methods tested in our datasets. We examine in more detail a specific case of potential cryptic diversity. In section 5.2, we focus on the biological and ecological interpretation of the main environmental correlates of amphibian diversity distribution in Mauritania. Then it follows an analysis of the distribution patterns predicted for the whole study area highlighting the main biodiversity hotspots and a brief overview on the main threats and conservation status. Discussion ends with general concluding remarks on this preliminary assessment on the conservation biogeography of amphibians of Mauritania.

5.1. Systematics and DNA barcoding of amphibians of Mauritania Barcoding analyses of animal organisms are usually performed on the single mitochondrial marker COI. Selecting a marker with a high mutation rate and small effective population size like COI (Moore, 1995) can help circumventing problems posed by large speciation rates and population sizes to the barcoding gap approach and to species delimitation methods. It has proved to be very useful in flagging cryptic diversity and revising species delimitation in several taxa (Dincă et al., 2015; Vasconcelos et al., 2016) including amphibians (Crawford et al., 2013; Fouquet et al., 2007; Jennings et al., 2015; Vences et al., 2005a). However, incongruences between gene and species trees may occur, and since most of the methods employed in barcoding studies are also applicable in diploid/nuclear datasets, independent lines of evidence should be used for validating mitochondrial results and help clarifying species boundaries. In this study, we assess whether mitochondrial and nuclear DNA patterns of variation conform to the clustering of sequences according to pre-defined amphibian species. We report two new genera, Amnirana and Leptopelis (A. galamensis and L. bufonides), FCUP 40 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units and two new species, T. milletihorsini and K. fusca, for Mauritania. T. milletihorsini is listed as Data Deficient in the IUCN Red List and its ecology and life history traits are largely unknown. The remaining three species are considered of Least Concern (IUCN, 2017), and are commonly associated to savannah (Nago et al., 2006). We confirm the occurrence of nine taxa previously recorded in Mauritania: H. occipitalis, K. senegalensis, P. edulis, Phrynobatrachus sp., P. bibroni, P. schillukorum, P. trinodis, S. pentoni, S. regularis and S. xeros. Data collected from 13 years of field expeditions does not confirm the occurrence of 6 species previously reported or hypothesized to occur in Mauritania: Hyperolius nitidulus, Phrynomantis microps, Ptychadena mascareniensis, Ptychadena pumilio and Tomopterna cryptotis (Table 1).

5.1.1. Systematics All the 14 formally defined taxa were confirmed by both markers. Phrynobatrachus sp. remains identified only until the genus level in the present study, as the small size of the specimens posed substantial difficulties to the morphological identification. Diagnostic characters narrowed down to two possibilities already hypothesized to occur in the region P. francisci and P. natalensis (Table 1). T. milletihorsini is the second case of uncertainty in our dataset. Although previous Mauritanian records of Tomopterna have been assigned to T. cryptotis (Table 1), we considered T. milletihorsini more likely to occur in the region. First, the locality type of T. milletihorsini is in the neighbouring Mali (Angel, 1922), whereas that of T. cryptotis is in Angola (Largen & Borkin, 2000). Second a recent phylogeny of the genus Tomopterna placed a Mauritanian sample in a distinct clade (Zimkus & Larson, 2011), apart from T. cryptotis. T. milletihorsini was, described based on one juvenile, a specimen that is now lost, which poses challenges to the identification of diagnostic characters (Ohler & Frétey, 2008). Both our markers were successful in identifying two divergent lineages of H. occipitalis, for which we were unable to detect differences in the morphology. Although is not the aim of a barcoding study to do phylogenetic inference, we chose RAG1 to validate the mitochondrial clusters because of its different taxonomic breadth and depth (Frost et al., 2006; Hoegg et al., 2004; Irisarri et al., 2012; Portik & Blackburn, 2016; Pyron & Wiens, 2011; Roelants et al., 2007; Diego San Mauro et al., 2004; van der Meijden et al., 2004; Wiens, 2007) and, in fact, it performed better than COI at deepest nodes. In general, topologies from the combined markers and RAG1 alone returned a topologies more similar to Pyron & Wiens, 2011 (Appendix: FCUP 41 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Supplementary Figure 1). For instance, Afrobatrachian individuals, represented in our dataset by the genera Kassina and Leptopelis and clustered within Ranoidea, are also returned as monophyletic (Pyron & Wiens, 2011; Portik & Blackburn, 2016). Tomopterna and Pyxicephalus, members of the family Pyxicepahlidae are also clustered together by combined markers and RAG1.

5.1.2. Barcoding performance COI-distance methods were successful in matching the samples with pre-defined species and the barcode returned a barcoding gap for all the individuals in the dataset. We went further on testing it in the two H. occipitalis lineages identified in the phylogenetic reconstructions and found it to decrease, but still existent. Maximum intraspecific uncorrected pairwise distances found in COI (18.10%) were unusually high for amphibians, which changed after considering two H. occipitalis lineages (2.87%). Examples of other studies on anurans reported maximum intraspecific distances in COI of 9% (Rockney et al., 2015) and 14.5% (Grosjean et al., 2015). Divergences between species were kept at similar levels after splitting H. occipitalis (from 13.53% to 12.44%). Other studies report equal or lower levels: 12.1% (Hofmann et al., 2017), 9% (Rockney et al., 2015) and 5.4% (Grosjean et al., 2015). To the extent of our knowledge, uncorrected pairwise distances are not usually calculated for the nuclear RAG1 rendering comparisons across studies difficult. Nonetheless, it is interesting to note that this marker returns a barcoding gap for the great majority of the individuals. Although the concept of barcoding gap works well when applied to COI due to its intrinsic characteristics (Hebert, et al., 2003; Hebert et al., 2003; Meyer & Paulay, 2005), the fact that the nuclear marker largely confirms it, increases the level of confidence in our results. H. occipitalis appears again with an interesting pattern; recognition of two lineages leads to a marginal situation with small overlaps and gaps across the samples. Species delimitation methods are commonly used in barcoding studies. Both methods employed here use a coalescent approach that do not require reciprocal monophyly of alleles or fixed differences (Fujita et al., 2012). GMYC is thought to oversplit species when undersampling affects maximum intraspecific distance (Talavera et al., 2013) and to underperform with big or uneven population sizes (Fujisawa & Barraclough, 2013). The mPTP by contrast has been suggested to outperform other delimitation methods in the presence of uneven sampling or large differences in effective population sizes of species (Blair & Bryson, 2017). It is interesting to note though, that both methods FCUP 42 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units agreed when applied either to COI or RAG1 datasets, and are mostly in concordance with the pre-defined species. The congruence between methods in splitting Hoplobatrachus occipitalis in two clusters was expected as there is evidence that two separate species exist under the same name, with diploid and tetraploid populations (Bogart & Tandy, 1976). Despite the fact that we did not observe an overlap between the intra- and interspecific distance ranges like in other amphibian species (Smith et al., 2008; Vences et al., 2005a) further support for this splitting is given by the sharp increase of the barcoding gap length after analysing the individuals according to the gene trees. Species delimitation methods recognize only one taxa when applied to the nuclear marker, but all the topologies agreed on the two reciprocal monophyletic lineages within H. occipitalis. Importantly, two major parapatric lineages have been detected by one mitochondrial (16S) and two nuclear markers (NTF3 and POMC), in the same study area (Gonçalves, 2017). Our data confirms the occurrence of one lineage in Tagant, Assaba and Afollé massifs, and another one in Adrar Atar and southern Mauritania, along Senegal river basin. Polyploidy would need to be confirmed through karyotyping (Bogart & Tandy, 1976) or microsatellites (Crespo-López et al., 2007). Data such as morphology and bioacoustics traits should be gathered for complementing the study of potential cryptic diversity and delimiting species (Funk et al., 2012a), subjecting H. occipitalis to a final integrated taxonomic approach. Phrynobatrachus sp. holds the second case of disagreement between markers. The distinct COI lineage identified with GYMC and mPTP was not confirmed in RAG1- based topology and the two COI lineages occur in sympatry along the Senegal river basin. We therefore considered all one single species. A blast in GenBank did not report satisfying matching results to help improving taxonomic resolution of this taxa in Mauritania.

5.2. Amphibian diversity and conservation in Mauritania 5.2.1. Environmental correlates of amphibian diversity Amphibians are water dependent organisms and desert evaders meaning they typically inhabit the edges of deserts where they find microhabitat features for preventing overheating and water loss (Ward, 2016). The harsher the environment, the scarcer the ecological refuges are. When habitat heterogeneity begins to increase at the end of an environmental severity gradient, a positive response in richness is expected (Yang et al., 2015). And so, it makes perfect sense that proximity to the savannah was FCUP 43 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units consistently returned as an important predictor of amphibian diversity in Mauritania. A suitable terrestrial habitat also provides a foraging ground for adult individuals (Semlitsch & Bodie, 2003) and is a key positive factor for desert anurans (Boeing et al., 2014). This pattern is in concordance with the biogeographical affinities of the species sampled since they are not desert specialists, but mainly Afro-tropic (Stuart et al., 2008) in the northern limit of their distribution (Padial et al., 2013). In fact, we only registered two species in the Sahara, away from the savannah influence, H. occipitalis (lineage 1) and S. xeros in the Adrar Atar, confirming Padial et al., 2013. Two more species occur in northern Tagant, in the transition zone between both Palaearctic and Afro-tropic realms, T. milletihorsini and H. occipitalis lineage 2. No more species have been previously registered in Saharan Mauritania (Table 1). Water indexes and proximity to gravel floodplains were next in explaining the observed richness pattern. As proximity to temporary water-bodies increases, so does amphibian diversity. But the exact opposite is seen with permanent waters. These results need to be interpreted with caution. It is important to note that some random absences, mainly in the north, fell in pixels that exhibited in average higher values of the permanent water index (MNDW1) when compared to the sampling stations. Some of these point localities actually correspond to salt pans and brackish waters or artificial structures like dams, which are completely unsuitable for amphibians. Therefore, sources of suitable aquatic habitat for water-dependent species in arid systems like this, are likely to be secured, mainly by an ephemeral hydrographic network (Boeing et al., 2014; Suhling et al., 2004). It is thought to provide seasonal connectivity during the summer rainstorms, which has been hypothesized for explaining the metapopulation system Crocodylus suchus in Mauritania (Velo-Antón et al., 2014). Such a dependency on ephemeral waters coupled with a low dispersal ability renders desert amphibians extremely vulnerable to environmental change and stochastic events in an area that already suffered from a prolonged drought that led relict fish populations in the Adrar to extinction (Trape, 2009). Associated to the seasonal hydrographic network in Mauritania are most of the tâmoûrts. Floodplains have been identified as important landscape features by habitat suitability modelling of amphibians in the Chihuahuan desert (Dayton & Fitzgerald, 2006). However, this particular type of floodplain, locally known as Reg, consists of large features covered by gravel and frequently coated with desert varnish. These desert pavements are seasonally supplied by the wadis. At first glance, they do not seem a suitable habitat for amphibians (Ward, 2016) but a careful examination of the study area reveals its deep association with most water-bodies identified in Mauritania. Gravel floodplains are known to shape vegetation patterns in FCUP 44 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units hyper-arid landscapes. A study conducted in the Sonoran and Mojave deserts revealed an ephemeral plant species capable of colonizing these features with seed dispersal possibly mediated by the passage of the seasonal waters (Martínez-Berdeja et al., 2013). Gravel floodplains might not be completely bare throughout the whole country and, most importantly, associated to a temporary hydrographic network, could prove relevant for reproduction and dispersal of some species. In fact, hundreds of active amphibians have been observed in the tâmoûrts during the rainy season, contrasting with its near absence immediately after the season is over (Brito et al., 2011). On the contrary, floodplains composed of equal proportions of gravel and sand seem to be somewhat avoided by amphibians. The two tâmoûrts differ essentially on the sediment deposits which, in turn, depend on the surrounding landscape and the velocity of the currents that created the floodplain (Gregory et al., 1991). This type of floodplain might be more associated to sandy regions which seem to be avoided by amphibians (Dayton & Fitzgerald, 2006) as we can see by the negative effect that yellow dunes have in the observed richness. They are also more prevalent in the far north of the country where it occupies extensive areas followed by yellow dunes in the centre. Gueltas were not identified as a relevant predictor by our model perhaps because they might be more important in the northern limits of amphibians’ distribution, where the ecosystem becomes increasingly arid and bare, providing the only source of water and associated habitats. Moreover, gueltas are very small and represent just 0.00004% of the country’s surface area (Vale et al., 2015), and therefore, its relevance might not be captured by our model or is partially considered by the water indices.

5.2.2. Predicting amphibian distribution Our modelling approach falls within the category “assemble first, predict later” by Ferrier & Guisan (2006). By modelling richness directly, we are assuming that differences in species traits can be ignored and therefore, we neglect the specific environmental associations of each taxa and does not reveal information about composition (Gaston & Blackburn, 1999). Moreover, wide-ranging species can have a disproportional weight on this type of modelling (Jetz, 2002). For these reasons, approaches based on “predict first, assemble later” as stacked species distribution models (S-SDMs) are arguably (see Distler et al., 2015) more suitable when modelling richness (Ko et al., 2016). Nonetheless, S-SDMs are claimed to overestimate species FCUP 45 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units richness (Guisan & Rahbek, 2011). In any case, our dataset did not include sufficient number of samples for all species to be modelled individually. Increasing accuracy and precision of the model would possibly imply considering variable(s) that take into account the rainfall regimes which are known to vary along the Sahel (Nicholson, 2013) and can potential interact with other environmental variables. For instance, hydroperiods of water-bodies are known to impact amphibian assemblages (Babbitt et al., 2003; Ferreira & Beja, 2013; Snodgrass et al., 2000). However, climatic variables are not available at such a small scale and it would imply modelling at a coarser resolution improper for holding biological relevance as often happens, although alternatives are being developed (Nadeau et al., 2017).

Our model predicts an increase in amphibian diversity that follows a latitudinal gradient as we move away from the drier sandy Sahara into the Sahel savannah, with more localized small clusters of diversity in the north and centre, confirming the general pattern of biodiversity distribution in this transition zone (Brito et al., 2014). A big sampling effort was allocated over the years in southern areas, as well as, along Afollé and Assaba massifs, all previously cited as unexplored (Padial et al., 2013). Padial et al., 2013 suggested 27 important localities for anurans mainly in Adrar Atar, Tagant and Assaba. Approximately half were sampled in this study with values ranging from two to five species, the vast majority holding two species. Our model identified more potentially suitable areas for instance in Katchi hydrographic basin and Senegal river basin. Two major hotspots of diversity stand out in two river basins along the south- eastern border with Mali. This used to be a region dominated by savannah and grassland that have been giving way to agriculture over the last decades (EROS Center; https://eros.usgs.gov/, accessed September 2017). Intensive farming practices surrounding floodplains and temporary ponds are known to have detrimental impacts in amphibian assemblages and correlate with decreased richness in semi-arid landscapes (Beja & Alcazar, 2003; Ferreira & Beja, 2013; Venne et al., 2012). A topic that deserves additional research in Mauritania as extensive areas in the Sahel have been fragmented by soil conversion (Brito et al., 2014). No national parks are implemented nor planned to be in these two hotspots (IUCN & UNEP-WCMC, 2017), yet 244 wetlands have been counted in south-eastern Mauritania during an inventory (Shine, 2003). Only Lake Mahmoudé, the largest wetland in eastern Mauritania and partially sampled during this study, is under discussion. Additional protected areas under discussion include tâmoûrt Bougâri, a wetland identified by our model as an area with moderate to high richness. Another area under FCUP 46 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units discussion is Lac de Mâl, classified as Important Bird Area (IBA), which is a permanent lake fed by seasonal water courses and used by livestock, and predicted to hold no more than four species. Location of priority areas for the conservation of amphibian diversity in Mauritania would profit from a combined approach including spatial information on location of hotspots and priority corridors. Next step should focus on identifying putative corridors in the hydrographic network to be conserved to assure connectivity between the amphibian hotspots.

5.3. Concluding remarks Our study provides a clear example of how combining nuclear and mitochondrial lines of evidence in DNA barcoding can accelerate a taxonomic workflow (Hajibabaei et al., 2007; Hebert et al., 2003a), as well as provide useful information for population-level studies (Hajibabaei et al., 2007), offering an opportunity to test barcoding methods in well-studied groups, like amphibians. Lack of both accurate taxonomic resolution and data on distribution have been pointed out as two main knowledge gaps hampering further conservation research aimed at Mauritanian amphibians (Padial et al., 2013). This present study was able of addressing this issue by combining DNA barcoding techniques with expert morphological identification of specimens. Some species need, however, further efforts for clarifying their taxonomic status. To the extent of our knowledge, this is the first study attempting to model amphibian distribution in the entire Sahara-Sahel. Our analyses show that ecological spatial modelling can be a valuable source of information for fostering our understanding of patterns of biodiversity, particularly in large remote areas difficult to sample.

FCUP 47 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

6. References

Adamowicz, S. J. (2015). International Barcode of Life: evolution of a global research community. Genome, 58(5), 151–162. https://doi.org/10.1139/gen-2015-0094 Aliabadian, M., Kaboli, M., Nijman, V., & Vences, M. (2009). Molecular identification of birds: performance of distance-based DNA barcoding in three genes to delimit parapatric species. PLoS ONE, 4(1), e4119. https://doi.org/10.1371/journal.pone.0004119 Angel, F. (1922). Sur une collection de reptiles et de batraciens, recueillis au Soudan français par la mission du Dr. Millet Horsin. Bulletin Du Muséum d’Histoire Naturelle (Paris), 28, 39–41. Armstrong, K. F., & Ball, S. L. (2005). DNA barcodes for biosecurity: invasive species identification. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 360(1462), 1813–1823. https://doi.org/10.1098/rstb.2005.1713 Austerlitz, F., David, O., Schaeffer, B., Bleakley, K., Olteanu, M., Leblois, R., Veuille, M., & Laredo, C. (2009). DNA barcode analysis: a comparison of phylogenetic and statistical classification methods. BMC Bioinformatics, 10(Suppl 14), S10. https://doi.org/10.1186/1471-2105-10-S14-S10 Babbitt, K. J., Baber, M. J., & Tarr, T. L. (2003). Patterns of larval amphibian distribution along a wetland hydroperiod gradient. Canadian Journal of Zoology, 81(9), 1539–1552. https://doi.org/10.1139/z03-131 Baker, C. C. M., Bittleston, L. S., Sanders, J. G., & Pierce, N. E. (2016). Dissecting host-associated communities with DNA barcodes. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1702), 20150328. https://doi.org/10.1098/rstb.2015.0328 Barej, M. F., Schmitz, A., Günther, R., Loader, S. P., Mahlow, K., & Rödel, M.-O. (2014). The first endemic West African vertebrate family – a new anuran family highlighting the uniqueness of the Upper Guinean biodiversity hotspot. Frontiers in Zoology, 11(1), 8. https://doi.org/10.1186/1742-9994-11-8 Barton, K. (2016). MuMIn: Multi-model inference. R package version 1.15.6. Available athttp://cran.r-project.org/package=MuMIn. https://doi.org/citeulike:11961261 Beja, P., & Alcazar, R. (2003). Conservation of Mediterranean temporary ponds under agricultural intensification: an evaluation using amphibians. Biological Conservation, 114(3), 317–326. https://doi.org/10.1016/S0006-3207(03)00051-X Bensasson, D., Zhang, D.-X., Hartl, D. L., & Hewitt, G. M. (2001). Mitochondrial FCUP 48 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

pseudogenes: evolution’s misplaced witnesses. Trends in Ecology & Evolution, 16(6), 314–321. https://doi.org/10.1016/S0169-5347(01)02151-6 Biju, S., Garg, S., Gururaja, K., Shouche, Y., & Walujkar, S. a. (2014). DNA barcoding reveals unprecedented diversity in dancing frogs of India (Micrixalidae, Micrixalus): a taxonomic revision with description of 14 new species. Ceylon Journal of Science (Biological Sciences), 43(1), 1–87. https://doi.org/10.4038/cjsbs.v43i1.6850 Blair, C., & Bryson, R. W. (2017). Cryptic diversity and discordance in single-locus species delimitation methods within horned lizards (Phrynosomatidae: Phrynosoma). Molecular Ecology Resources. https://doi.org/10.1111/1755- 0998.12658 Blotto, B. L., Nuñez, J. J., Basso, N. G., Úbeda, C. A., Wheeler, W. C., & Faivovich, J. (2013). Phylogenetic relationships of a Patagonian frog radiation, the Alsodes + Eupsophus clade (Anura: Alsodidae), with comments on the supposed paraphyly of Eupsophus. Cladistics, 29(2), 113–131. https://doi.org/10.1111/j.1096- 0031.2012.00417.x Boeing, W. J., Griffis-Kyle, K. L., & Jungels, J. M. (2014). Anuran habitat associations in the northern Chihuahuan desert, USA. Journal of Herpetology, 48(1), 103–110. https://doi.org/10.1670/12-184 Bogart, J., & Tandy, M. (1976). Polyploid amphibians: three more diploid-tetraploid cryptic species of frogs. Science, 193(4250), 334–335. https://doi.org/10.1126/science.935871 Bouckaert, R., Heled, J., Kühnert, D., Vaughan, T., Wu, C.-H., Xie, D., Suchard, M. A., Rambaut, A., & Drummond, A. J. (2014). BEAST 2: a software platform for Bayesian evolutionary analysis. PLoS Computational Biology, 10(4), e1003537. https://doi.org/10.1371/journal.pcbi.1003537 Braunisch, V., Patthey, P., & Arlettaz, R. (2011). Spatially explicit modeling of conflict zones between wildlife and snow sports: prioritizing areas for winter refuges. Ecological Applications : A Publication of the Ecological Society of America, 21(3), 955–67. https://doi.org/10.1890/09-2167.1 Brito, J. C., Godinho, R., Martínez-Freiría, F., Pleguezuelos, J. M., Rebelo, H., Santos, X., Vale, C. G., Velo-Antón, G., Boratyński, Z., Carvalho, S. B., Ferreira, S., Gonçalves, D. V., Silva, T. L., Tarroso, P., Campos, J. C., Leite, J. V., Nogueira, J., Alvares, F., Sillero, N., Sow, A. S., Fahd, S., Crochet, P.-A., & Carranza, S. (2014). Unravelling biodiversity, evolution and threats to conservation in the Sahara-Sahel. Biological Reviews of the Cambridge Philosophical Society, 89(1), FCUP 49 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

215–31. https://doi.org/10.1111/brv.12049 Brito, J. C., Martínez-Freiría, F., Sierra, P., Sillero, N., & Tarroso, P. (2011). Crocodiles in the Sahara desert: an update of distribution, habitats and population status for conservation planning in Mauritania. PloS One, 6(2), e14734. https://doi.org/10.1371/journal.pone.0014734 Brito, J. C., Tarroso, P., Vale, C. G., Martínez-Freiría, F., Boratyński, Z., Campos, J. C., Ferreira, S., Godinho, R., Gonçalves, D. V., Leite, J. V., Lima, V. O., Pereira, P., Santos, X., da Silva, M. J. F., Silva, T. L., Velo-Antón, G., Veríssimo, J., Crochet, P.-A., Pleguezuelos, J. M., & Carvalho, S. B. (2016). Conservation Biogeography of the Sahara-Sahel: additional protected areas are needed to secure unique biodiversity. Diversity and Distributions, 22(4), 371–384. https://doi.org/10.1111/ddi.12416 Brown, J. H., & Lomolino, M. V. (1998). Biogeography. SpringerReference (Vol. 2nd). https://doi.org/10.1007/SpringerReference_29469 Brown, S. D. J., Collins, R. A., Boyer, S., Lefort, M. C., Malumbres-Olarte, J., Vink, C. J., & Cruickshank, R. H. (2012). Spider: An R package for the analysis of species identity and evolution, with particular reference to DNA barcoding. Molecular Ecology Resources, 12(3), 562–565. https://doi.org/10.1111/j.1755- 0998.2011.03108.x Buchadas, A., Vaz, A. S., Honrado, J. P., Alagador, D., Bastos, R., Cabral, J. A., Santos, M., & Vicente, J. R. (2017). Dynamic models in research and management of biological invasions. Journal of Environmental Management, 196, 594–606. https://doi.org/10.1016/j.jenvman.2017.03.060 Burnham, K. P., & Anderson, D. R. (2004). Multimodel inference: understanding AIC and BIC in model selection. Sociological Methods & Research, 33(2), 261–304. https://doi.org/10.1177/0049124104268644 Bwong, B. A., Chira, R., Schick, S., Veith, M., & Lötters, S. (2009). Diversity of ridged frogs (Ptychadenidae: Ptychadena) in the easternmost remnant of the Guineo- Congolian rain forest: an analysis using morphology, bioacoustics and molecular genetics. Salamandra, 45(3), 129–146. Campos, J. C., Sillero, N., & Brito, J. C. (2012). Normalized difference water indexes have dissimilar performances in detecting seasonal and permanent water in the Sahara–Sahel transition zone. Journal of Hydrology, 464–465, 438–446. https://doi.org/10.1016/j.jhydrol.2012.07.042 Che, J., Chen, H.-M., Yang, J.-X., Jin, J.-Q., Jiang, K., Yuan, Z.-Y., Murphy, R. W., & Zhang, Y.-P. (2012). Universal COI primers for DNA barcoding amphibians. FCUP 50 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Molecular Ecology Resources, 12(2), 247–58. https://doi.org/10.1111/j.1755- 0998.2011.03090.x Chen, S., Pang, X., Song, J., Shi, L., Yao, H., Han, J., & Leon, C. (2014). A renaissance in herbal medicine identification: from morphology to DNA. Biotechnology Advances, 32(7), 1237–44. https://doi.org/10.1016/j.biotechadv.2014.07.004 Clemente-Carvalho, R. B. G., Perez, S. I., Tonhatti, C. H., Condez, T. H., Sawaya, R. J., Haddad, C. F. B., & Reis, S. F. (2016). Boundaries of morphological and molecular variation and the distribution of a miniaturized froglet, Brachycephalus nodoterga (Anura: Brachycephalidae). Journal of Herpetology, 50(1), 169–178. https://doi.org/10.1670/14-119 Cooper, A., Shine, T., McCann, T., & Tidane, D. A. (2006). An ecological basis for sustainable land use of Eastern Mauritanian wetlands. Journal of Arid Environments, 67(1), 116–141. https://doi.org/10.1016/j.jaridenv.2006.02.003 Crawford, A. J., Cruz, C., Griffith, E., Ross, H., Ibáñez, R., Lips, K. R., Driskell, A. C., Bermingham, E., & Crump, P. (2013). DNA barcoding applied to ex situ tropical amphibian conservation programme reveals cryptic diversity in captive populations. Molecular Ecology Resources, 13(6), 1005–18. https://doi.org/10.1111/1755-0998.12054 Crespo-López, M. E., Pala, I., Duarte, T. L., Dowling, T. E., & Coelho, M. M. (2007). Genetic structure of the diploid–polyploid fish Squalius alburnoides in southern Iberian basins Tejo and Guadiana, based on microsatellites. Journal of Fish Biology, 71(sc), 423–436. https://doi.org/10.1111/j.1095-8649.2007.01688.x DasGupta, B., Konwar, K. M., Mandoiu, I. I., & Shvartsman, A. A. (2005). DNA-BAR: distinguisher selection for DNA barcoding. Bioinformatics (Oxford, England), 21(16), 3424–6. https://doi.org/10.1093/bioinformatics/bti547 Dawson, M. N., Axmacher, J. C., Beierkuhnlein, C., Blois, J. L., Bradley, B. A., Cord, A. F., Dengler, J., He, K. S., Heaney, L. R., Jansson, R., Mahecha, M. D., Myers, C., Nogués-Bravo, D., Papadopoulou, A., Reu, B., Rodríguez-Sánchez, F., Steinbauer, M. J., Stigall, A., Tuanmu, M.-N., & Gavin, D. G. (2016). A second horizon scan of biogeography: golden ages, Midas touches, and the Red Queen. Frontiers of Biogeography, 8(4), 217–220. https://doi.org/10.21425/F58429770 Dayton, G. H., & Fitzgerald, L. A. (2006). Habitat suitability models for desert amphibians. Biological Conservation, 132(1), 40–49. https://doi.org/10.1016/j.biocon.2006.03.012 Degnan, J. H., & Rosenberg, N. A. (2009). Gene tree discordance, phylogenetic FCUP 51 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

inference and the multispecies coalescent. Trends in Ecology & Evolution, 24(6), 332–340. https://doi.org/10.1016/j.tree.2009.01.009 Dincă, V., Lukhtanov, V. a, Talavera, G., & Vila, R. (2011). Unexpected layers of cryptic diversity in wood white Leptidea butterflies. Nature Communications, 2, 324. https://doi.org/10.1038/ncomms1329 Dincă, V., Montagud, S., Talavera, G., Hernández-Roldán, J., Munguira, M. L., García- Barros, E., Hebert, P. D. N., & Vila, R. (2015). DNA barcode reference library for Iberian butterflies enables a continental-scale preview of potential cryptic diversity. Scientific Reports, 5(1), 12395. https://doi.org/10.1038/srep12395 Dinerstein, E., Olson, D., Joshi, A., Vynne, C., Burgess, N. D., Wikramanayake, E., Hahn, N., Palminteri, S., Hedao, P., Noss, R., Hansen, M., Locke, H., Ellis, E. C., Jones, B., Barber, C. V., Hayes, R., Kormos, C., Martin, V., Crist, E., Sechrest, W., Price, L., Baillie, J. E. M., Weeden, D., Suckling, K., Davis, C., Sizer, N., Moore, R., Thau, D., Birch, T., Potapov, P., Turubanova, S., Tyukavina, A., de Souza, N., Pintea, L., Brito, J. C., Llewellyn, O. A., Miller, A. G., Patzelt, A., Ghazanfar, S. A., Timberlake, J., Klöser, H., Shennan-Farpón, Y., Kindt, R., Lillesø, J.-P. B., van Breugel, P., Graudal, L., Voge, M., Al-Shammari, K. F., & Saleem, M. (2017). An ecoregion-based approach to protecting half the terrestrial realm. BioScience, 67(6), 534–545. https://doi.org/10.1093/biosci/bix014 Dirzo, R., Young, H. S., Galetti, M., Ceballos, G., Isaac, N. J. B., & Collen, B. (2014). Defaunation in the Anthropocene. Science, 345(6195), 401–406. https://doi.org/10.1126/science.1251817 Distler, T., Schuetz, J. G., Velásquez-Tibatá, J., & Langham, G. M. (2015). Stacked species distribution models and macroecological models provide congruent projections of avian species richness under climate change. Journal of Biogeography, 42(5), 976–988. https://doi.org/10.1111/jbi.12479 Dupont, L., Porco, D., Symondson, W. O. C., & Roy, V. (2016). Hybridization relics complicate barcode-based identification of species in earthworms. Molecular Ecology Resources, 16(4), 883–894. https://doi.org/10.1111/1755-0998.12517 Durant, S. M., Pettorelli, N., Bashir, S., Woodroffe, R., Wacher, T., De Ornellas, P., Ransom, C., Abaigar, T., Abdelgadir, M., El Alqamy, H., Beddiaf, M., Belbachir, F., Belbachir-Bazi, A., Berbash, a. a., Beudels-Jamar, R., Boitani, L., Breitenmoser, C., Cano, M., Chardonnet, P., Collen, B., Cornforth, W. a., Cuzin, F., Gerngross, P., Haddane, B., Hadjeloum, M., Jacobson, A., Jebali, A., Lamarque, F., Mallon, D., Minkowski, K., Monfort, S., Ndoassal, B., Newby, J., Ngakoutou, B. E., Niagate, B., Purchase, G., Samaila, S., Samna, a. K., Sillero-Zubiri, C., Soultan, FCUP 52 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

a. E., Stanley Price, M. R., & Baillie, J. E. M. (2012). Forgotten biodiversity in desert ecosystems. Science, 336(6087), 1379–1380. https://doi.org/10.1126/science.336.6087.1379 Efron, B., & Gong, G. (1983). A leisurely look at the bootstrap, the jackknife, and cross- Validation. The American Statistician, 37(1), 36–48. https://doi.org/10.1080/00031305.1983.10483087 Eldredge, N., & Cracraft, J. (1980). Phylogenetic patterns and the evolutionary process. Method and theory in comparative biology. New York: Columbia Univ. Press ESRI. (2012). ArcMap 10.1. ESRI. Ezard, T., Fujisawa, T., & Barraclough, T. G. (2009). Splits: species’ limits by threshold statistics. R Package Version, 1. Faivovich, J., Nicoli, L., Blotto, B. L., Pereyra, M. O., Baldo, D., Barrionuevo, J. S., Fabrezi, M., Wild, E. R., & Haddad, C. F. B. (2014). Big, bad, and beautiful: phylogenetic relationships of the horned frogs (Anura: Ceratophryidae). South American Journal of Herpetology, 9(3), 207–227. https://doi.org/10.2994/SAJH-D- 14-00032.1 Ferreira, M., & Beja, P. (2013). Mediterranean amphibians and the loss of temporary ponds: are there alternative breeding habitats? Biological Conservation, 165, 179– 186. https://doi.org/10.1016/j.biocon.2013.05.029 Ferrier, S., & Guisan, A. (2006). Spatial modelling of biodiversity at the community level. Journal of Applied Ecology, 43(3), 393–404. https://doi.org/10.1111/j.1365- 2664.2006.01149.x Fischer, M. M. (2006). Computational neural networks - tools for spatial data analysis. In Spatial Analysis and GeoComputation: Selected Essays (pp. 79–102). Berlin, Heidelberg: Springer Berlin Heidelberg: Available athttps://doi.org/10.1007/3-540- 35730-0_6. https://doi.org/10.1007/3-540-35730-0_6 Fleischack, P. C., & Small, C. P. (1978). The vocalizations and breeding behaviour of Kassina senegalensis (Anura, Rhacophoridae) in summer breeding aggregations. Koedoe, 21(1), 91–99. Fouquet, A., Blotto, B. L., Maronna, M. M., Verdade, V. K., Juncá, F. A., de Sá, R., & Rodrigues, M. T. (2013a). Unexpected phylogenetic positions of the genera Rupirana and Crossodactylodes reveal insights into the biogeography and reproductive evolution of leptodactylid frogs. Molecular Phylogenetics and Evolution, 67(2), 445–457. https://doi.org/10.1016/j.ympev.2013.02.009 Fouquet, A., Gilles, A., Vences, M., Marty, C., Blanc, M., & Gemmell, N. J. (2007). Underestimation of species richness in neotropical frogs revealed by mtDNA FCUP 53 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

analyses. PLoS ONE, 2(10), e1109. https://doi.org/10.1371/journal.pone.0001109 Fouquet, A., Loebmann, D., Castroviejo-Fisher, S., Padial, J. M., Orrico, V. G. D., Lyra, M. L., Roberto, I. J., Kok, P. J. R., Haddad, C. F. B., & Rodrigues, M. T. (2012). From Amazonia to the Atlantic forest: molecular phylogeny of Phyzelaphryninae frogs reveals unexpected diversity and a striking biogeographic pattern emphasizing conservation challenges. Molecular Phylogenetics and Evolution, 65(2), 547–561. https://doi.org/10.1016/j.ympev.2012.07.012 Fouquet, A., Pineau, K., Rodrigues, M. T., Mailles, J., Schneider, J.-B., Ernst, R., & Dewynter, M. (2013b). Endemic or exotic: the phylogenetic position of the Martinique Volcano frog Allobates chalcopis (Anura: Dendrobatidae) sheds light on its origin and challenges current conservation strategies. Systematics and Biodiversity, 11(1), 87–101. https://doi.org/10.1080/14772000.2013.764944 Fouquet, A., Santana Cassini, C., Fernando Baptista Haddad, C., Pech, N., & Trefaut Rodrigues, M. (2014). Species delimitation, patterns of diversification and historical biogeography of the Neotropical frog genus Adenomera (Anura, Leptodactylidae). Journal of Biogeography, 41(5), 855–870. https://doi.org/10.1111/jbi.12250 Friedman, J. H. (2001). Greedy function approximation: a gradient boosting machine. Annals of Statistics, 29(5), 1189–1232. https://doi.org/DOI 10.1214/aos/1013203451 Frost, D. R. (2017). Amphibian species of the world: an online reference. Version 6.0. Retrieved from http://research.amnh.org/herpetology/amphibia/index.html. Frost, D. R., Grant, T., Faivovich, J., Bain, R. H., Hass, A., Haddad, C. F. B., De Sá, R. O., Channing, A., Wilkinson, M., Donnellan, S. C., Raxworthy, C. J., Campbell, J. A., Blotto, B. L., Moler, P., Drewes, R. C., Nussbaum, R. A., Lynch, J. D., Green, D. M., & Wheeler, W. C. (2006). the Amphibian Tree of Life. Bulletin of the American Museum of Natural History, 297, 1–291. https://doi.org/10.1206/0003- 0090(2006)297[0001:TATOL]2.0.CO;2 Fujisawa, T., & Barraclough, T. G. (2013). Delimiting species using single-locus data and the generalized mixed Yule coalescent approach: a revised method and evaluation on simulated data sets. Systematic Biology, 62(5), 707–724. https://doi.org/10.1093/sysbio/syt033 Fujita, M. K., Leaché, A. D., Burbrink, F. T., McGuire, J. A., & Moritz, C. (2012). Coalescent-based species delimitation in an integrative taxonomy. Trends in Ecology & Evolution, 27(9), 480–488. https://doi.org/10.1016/j.tree.2012.04.012 Funk, D. J., & Omland, K. E. (2003). Species-level paraphyly and polyphyly: frequency, FCUP 54 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

causes, and consequences, with insights from animal mitochondrial DNA. Annual Review of Ecology, Evolution, and Systematics, 34(1), 397–423. https://doi.org/10.1146/annurev.ecolsys.34.011802.132421 Funk, W. C., Caminer, M., & Ron, S. R. (2012a). High levels of cryptic species diversity uncovered in Amazonian frogs. Proceedings of the Royal Society B: Biological Sciences, 279(1734), 1806–1814. https://doi.org/10.1098/rspb.2011.1653 Funk, W. C., McKay, J. K., Hohenlohe, P. A., & Allendorf, F. W. (2012b). Harnessing genomics for delineating conservation units. Trends in Ecology & Evolution, 27(9), 489–496. https://doi.org/10.1016/j.tree.2012.05.012 Gaston, K. J., & Blackburn, T. M. (1999). A critique for macroecology. Oikos, 84(3), 353. https://doi.org/10.2307/3546417 Geary, J., Camicioli, E., Bubela, T., & Steinke, D. (2016). DNA barcoding in the media: does coverage of cool science reflect its social context? Genome, 59(9), 738–750. https://doi.org/10.1139/gen-2015-0210 Gonçalves, D. V. (2017). Integrative inference of evolutionary patterns of desert biodiversity: a spatial and temporal multiscale approach using herpetofauna from North-Africa. (Unpublished doctoral thesis, University of Porto). Grant, T., Frost, D. R., Caldwell, J. P., Gagliardo, R., Haddad, C. F. B., Kok, P. J. R., Means, D. B., Noonan, B. P., Schargel, W. E., & Wheeler, W. C. (2006). Phylogenetic systematics of dart-poison frogs and their relatives (Amphibia, Athesphatanura, Dendrobatidae). Bulletin of the American Museum of Natural History, 299, 1–262. https://doi.org/10.1206/0003- 0090(2006)299[1:PSODFA]2.0.CO;2 Gregory, S. V., Swanson, F. J., McKee, W. A., & Cummins, K. W. (1991). An ecosystem perspective of riparian zones. BioScience, 41(8), 540–551. https://doi.org/10.2307/1311607 Grosjean, S., Ohler, A., Chuaynkern, Y., Cruaud, C., & Hassanin, A. (2015). Improving biodiversity assessment of anuran amphibians using DNA barcoding of tadpoles. Case studies from Southeast Asia. Comptes Rendus Biologies, 338(5), 351–361. https://doi.org/10.1016/j.crvi.2015.03.015 Guisan, A., Edwards, T. C., & Hastie, T. (2002). Generalized linear and generalized additive models in studies of species distributions: setting the scene. Ecological Modelling, 157(2–3), 89–100. https://doi.org/10.1016/S0304-3800(02)00204-1 Guisan, A., & Rahbek, C. (2011). SESAM - a new framework integrating macroecological and species distribution models for predicting spatio-temporal patterns of species assemblages. Journal of Biogeography, 38(8), 1433–1444. FCUP 55 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

https://doi.org/10.1111/j.1365-2699.2011.02550.x Guisan, A., & Zimmermann, N. E. (2000). Predictive habitat distribution models in ecology. Ecological Modelling, 135(2–3), 147–186. https://doi.org/10.1016/S0304- 3800(00)00354-9 Hajibabaei, M., Singer, G. A. C., Hebert, P. D. N., & Hickey, D. A. (2007). DNA barcoding: how it complements taxonomy, molecular phylogenetics and population genetics. Trends in Genetics, 23(4), 167–172. https://doi.org/10.1016/j.tig.2007.02.001 Harfoot, M. B. J., Newbold, T., Tittensor, D. P., Emmott, S., Hutton, J., Lyutsarev, V., Smith, M. J., Scharlemann, J. P. W., & Purves, D. W. (2014). Emergent global patterns of ecosystem structure and function from a mechanistic general ecosystem model. PLoS Biology, 12(4), e1001841. https://doi.org/10.1371/journal.pbio.1001841 Hebert, P. D. N., Cywinska, A., Ball, S. L., & DeWaard, J. R. (2003a). Biological identifications through DNA barcodes. Proceedings of the Royal Society B: Biological Sciences, 270(1512), 313–321. https://doi.org/10.1098/rspb.2002.2218 Hebert, P. D. N., Hollingsworth, P. M., Hajibabaei, M., & Hebert, P. D. N. (2016). From writing to reading the encyclopedia of life. Philosophical Transactions of the Royal Society of London B: Biological Sciences`, 371(1702), 1–9. https://doi.org/10.1098/rstb.2015.0321 Hebert, P. D. N., Ratnasingham, S., & de Waard, J. R. (2003b). Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proceedings of the Royal Society B: Biological Sciences, 270(Suppl_1), S96–S99. https://doi.org/10.1098/rsbl.2003.0025 Hebert, P. D. N., Stoeckle, M. Y., Zemlak, T. S., & Francis, C. M. (2004). Identification of birds through DNA barcodes. PLoS Biology, 2(10), e312. https://doi.org/10.1371/journal.pbio.0020312 Helyar, S. J., Lloyd, H. A. D., de Bruyn, M., Leake, J., Bennett, N., & Carvalho, G. R. (2014). Fish product mislabelling: failings of traceability in the production chain and implications for illegal, unreported and unregulated (IUU) fishing. PLoS ONE, 9(6), e98691. https://doi.org/10.1371/journal.pone.0098691 Henter, H. J., Imondi, R., James, K., Spencer, D., & Steinke, D. (2016). DNA barcoding in diverse educational settings: five case studies. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1702), 20150340. https://doi.org/10.1098/rstb.2015.0340 Hodgetts, J., Ostojá-Starzewski, J. C., Prior, T., Lawson, R., Hall, J., Boonham, N., & FCUP 56 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Wilson, J.-J. (2016). DNA barcoding for biosecurity: case studies from the UK plant protection program. Genome, 59(11), 1033–1048. https://doi.org/10.1139/gen-2016-0010 Hoegg, S. (2004). Phylogeny and comparative substitution rates of frogs inferred from sequences of three nuclear genes. Molecular Biology and Evolution, 21(7), 1188– 1200. https://doi.org/10.1093/molbev/msh081 Hofmann, S., Stöck, M., Zheng, Y., Ficetola, F. G., Li, J., Scheidt, U., & Schmidt, J. (2017). Molecular phylogenies indicate a Paleo-Tibetan origin of Himalayan lazy toads (Scutiger). Scientific Reports, 7(1), 3308. https://doi.org/10.1038/s41598- 017-03395-4 Hortal, J., de Bello, F., Diniz-Filho, J. A. F., Lewinsohn, T. M., Lobo, J. M., & Ladle, R. J. (2015). Seven shortfalls that beset large-scale knowledge of biodiversity. Annual Review of Ecology, Evolution, and Systematics, 46(1), 523–549. https://doi.org/10.1146/annurev-ecolsys-112414-054400 Hosein, F. N., Austin, N., Maharaj, S., Johnson, W., Rostant, L., Ramdass, A. C., & Rampersad, S. N. (2017). Utility of DNA barcoding to identify rare endemic vascular plant species in Trinidad. Ecology and Evolution, 7(18), 7311–7333. https://doi.org/10.1002/ece3.3220 Hubert, N., & Hanner, R. (2015). DNA Barcoding, species delineation and taxonomy: a historical perspective. DNA Barcodes, 3(1), 44–58. https://doi.org/10.1515/dna- 2015-0006 IEP. (2016). Global Peace Index 2016: Ten Years of Measuring Peace. Institute for Economics and Peace. Available at https://reliefweb.int/report/world/global-peace- index-2016. Accessed 20.08.2017 IPCC. (2014). Climate Change 2014: Synthesis Report. [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, 151 pp: Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Available at http://www.ipcc.ch/report/ar5/syr/. Accessed 19.08.2017 Irisarri, I., Mauro, D. S., Abascal, F., Ohler, A., Vences, M., & Zardoya, R. (2012). The origin of modern frogs (Neobatrachia) was accompanied by acceleration in mitochondrial and nuclear substitution rates. BMC Genomics, 13(1), 626. https://doi.org/10.1186/1471-2164-13-626 Irisarri, I., Vences, M., San Mauro, D., Glaw, F., & Zardoya, R. (2011). Reversal to air- driven sound production revealed by a molecular phylogeny of tongueless frogs, family Pipidae. BMC Evolutionary Biology, 11(1), 114. FCUP 57 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

https://doi.org/10.1186/1471-2148-11-114 IUCN. (2015). The IUCN Red List of Threatened Species. Version 2015-3. Gland, Switzerland: International Union for Conservation of Nature and Natural Resources. Available at http://www.iucnredlist.org/ IUCN. (2017). The IUCN Red List of Threatened Species. Version 2017.2. Available at www.iucnredlist.org. Accessed 03.12.2017 IUCN, & UNEP-WCMC. (2017). The World Database on Protected Areas (WDPA). Available at https://doi.org/www.protectedplanet.net. Accessed 19.08.2017 Jennings, W. B., Wogel, H., Bilate, M., Salles, R. D. O. L., & Buckup, P. A. (2015). DNA barcoding reveals species level divergence between populations of the microhylid frog genus Arcovomer (Anura: Microhylidae) in the Atlantic Rainforest of southeastern Brazil. Mitochondrial DNA, 0(0), 1–8. https://doi.org/10.3109/19401736.2015.1022731 Jetz, W. (2002). Geographic range size and determinants of avian species richness. Science, 297(5586), 1548–1551. https://doi.org/10.1126/science.1072779 Jetz, W., McPherson, J. M., & Guralnick, R. P. (2012). Integrating biodiversity distribution knowledge: toward a global map of life. Trends in Ecology & Evolution, 27(3), 151–159. https://doi.org/10.1016/j.tree.2011.09.007 Jørgensen, S. E., & Bendoricchio, G. (2001). Fundamentals of Ecological Modelling. Amsterdam: Elsevier. Jørgensen, S. E., & Fath, B. D. (2011). Fundamentals of Ecological Modelling: Applications in Environmental Management and Research. Amsterdam: Elsevier. Kalyaanamoorthy, S., Minh, B. Q., Wong, T. K. F., von Haeseler, A., & Jermiin, L. S. (2017). ModelFinder: fast model selection for accurate phylogenetic estimates. Nature Methods, 14(6), 587–591. https://doi.org/10.1038/nmeth.4285 Kapli, P., Lutteropp, S., Zhang, J., Kobert, K., Pavlidis, P., Stamatakis, A., & Flouri, T. (2017). Multi-rate Poisson tree processes for single-locus species delimitation under maximum mikelihood and Markov chain Monte Carlo. Bioinformatics, 33(11), btx025. https://doi.org/10.1093/bioinformatics/btx025 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(12), 1647–1649. https://doi.org/10.1093/bioinformatics/bts199 Ko, C.-Y., Schmitz, O. J., & Jetz, W. (2016). The limits of direct community modeling approaches for broad-scale predictions of ecological assemblage structure. FCUP 58 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Biological Conservation, 201, 396–404. https://doi.org/10.1016/j.biocon.2016.07.026 Kurabayashi, A., Matsui, M., Belabut, D. M., Yong, H., Ahmad, N., Sudin, A., Kuramoto, M., Hamidy, A., & Sumida, M. (2011). From Antarctica or Asia? New colonization scenario for Australian-New Guinean narrow mouth toads suggested from the findings on a mysterious genus Gastrophrynoides. BMC Evolutionary Biology, 11(1), 175. https://doi.org/10.1186/1471-2148-11-175 Largen, M. J., & Borkin, L. J. (2000). Rana cryptotis Boulenger, 1907 (currently Tomopterna cryptotis; Amphibia, Anura): proposed precedence of the specific name over that of Chiromantis kachowskii Nikolsky, 1900. Bulletin of Zoological Nomenclature, 57(1), 32–35. Le Houérou, H. N. (1997). Climate, flora and fauna changes in the Sahara over the past 500 million years. Journal of Arid Environments, 37(4), 619–647. https://doi.org/10.1006/jare.1997.0315 Leliaert, F., Verbruggen, H., Vanormelingen, P., Steen, F., López-Bautista, J. M., Zuccarello, G. C., & De Clerck, O. (2014). DNA-based species delimitation in algae. European Journal of Phycology, 49(2), 179–196. https://doi.org/10.1080/09670262.2014.904524 Loarie, S. R., Duffy, P. B., Hamilton, H., Asner, G. P., Field, C. B., & Ackerly, D. D. (2009). The velocity of climate change. Nature, 462(7276), 1052–1055. https://doi.org/10.1038/nature08649 Lomolino, M. V. (2004). Conservation Biogeography. In M. V Lomolino & L. R. Heaney (Eds.), Frontiers of Biogeography: new directions in the geography of nature (pp. 293–296). Sinauer Associates, Inc., Sunderland, Massachussets Lyra, M. L., Haddad, C. F. B., & de Azeredo-Espin, A. M. L. (2017). Meeting the challenge of DNA barcoding Neotropical amphibians: polymerase chain reaction optimization and new COI primers. Molecular Ecology Resources. https://doi.org/10.1111/1755-0998.12648 Mallo, D., & Posada, D. (2016). Multilocus inference of species trees and DNA barcoding. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1702), 20150335. https://doi.org/10.1098/rstb.2015.0335 Martínez-Berdeja, A., Pietrasiak, N., Tamase, A., Ezcurra, E., & Allen, E. B. (2013). Living where others dare not: microhabitat distribution in Chorizanthe rigida, a serotinous desert annual. Journal of Arid Environments, 97, 120–126. https://doi.org/10.1016/j.jaridenv.2013.05.010 Mauro, D. S., Vences, M., Alcobendas, M., Zardoya, R., & Meyer, A. (2005). Initial FCUP 59 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

diversification of living amphibians predated the breakup of Pangaea. The American Naturalist, 165(3), 590–599. https://doi.org/https://doi.org/10.1086/429523 McLean, A. H. C., Parker, B. J., Hrček, J., Henry, L. M., & Godfray, H. C. J. (2016). Insect symbionts in food webs. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1702), 20150325. https://doi.org/10.1098/rstb.2015.0325 Meier, R., Shiyang, K., Vaidya, G., Ng, P. K. L., & Hedin, M. (2006). DNA barcoding and taxonomy in Diptera: a tale of high intraspecific variability and low identification success. Systematic Biology, 55(5), 715–728. https://doi.org/10.1080/10635150600969864 Meier, R., Zhang, G., Ali, F., & Zamudio, K. (2008). The use of mean instead of smallest interspecific distances exaggerates the size of the “barcoding gap” and leads to misidentification. Systematic Biology, 57(5), 809–813. https://doi.org/10.1080/10635150802406343 Melo, M. A., Lyra, M. L., Brischi, A. M., Geraldi, V. C., & Haddad, C. F. B. (2014). First record of the invasive frog Eleutherodactylus johnstonei (Anura: Eleutherodactylidae) in São Paulo, Brazil. Salamandra, 50(3), 177–180. Meyer, C. P., & Paulay, G. (2005). DNA barcoding: error rates based on comprehensive sampling. PLoS Biology, 3(12), e422. https://doi.org/10.1371/journal.pbio.0030422 Miller, M. A., Pfeiffer, W., & Schwartz, T. (2010). Creating the CIPRES science gateway for inference of large phylogenetic trees. In 2010 Gateway Computing Environments Workshop, GCE 2010. https://doi.org/10.1109/GCE.2010.5676129 Mittanck, C. M., Rogers, P. C., Ramsey, R. D., Bartos, D. L., & Ryel, R. J. (2014). Exploring succession within aspen communities using a habitat-based modeling approach. Ecological Modelling, 288, 203–212. https://doi.org/10.1016/j.ecolmodel.2014.06.010 Moore, W. S. (1995). Inferring phylogenies from mtDNA variation: mitochondrial-gene trees versus nuclear-gene trees. Evolution, 49(4), 718–726. https://doi.org/10.1111/j.1558-5646.1995.tb02308.x Murphy, R. W., Crawford, A. J., Bauer, A. M., Che, J., Donnellan, S. C., Fritz, U., Haddad, C. F. B., Nagy, Z. T., Poyarkov, N. A., Vences, M., Wang, W.-Z., & Zhang, Y.-P. (2013). Cold Code: the global initiative to DNA barcode amphibians and nonavian reptiles. Molecular Ecology Resources, 13(2), 161–167. https://doi.org/10.1111/1755-0998.12050 Mutanen, M., Kivelä, S. M., Vos, R. A., Doorenweerd, C., Ratnasingham, S., FCUP 60 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Hausmann, A., Huemer, P., Dincă, V., van Nieukerken, E. J., Lopez-Vaamonde, C., Vila, R., Aarvik, L., Decaëns, T., Efetov, K. A., Hebert, P. D. N., Johnsen, A., Karsholt, O., Pentinsaari, M., Rougerie, R., Segerer, A., Tarmann, G., Zahiri, R., & Godfray, H. C. J. (2016). Species-level para- and polyphyly in DNA barcode gene trees: strong operational bias in European Lepidoptera. Systematic Biology, 65(6), 1024–1040. https://doi.org/10.1093/sysbio/syw044 Nadeau, C. P., Urban, M. C., & Bridle, J. R. (2017). Coarse climate change projections for species living in a fine-scaled world. Global Change Biology, 23(1), 12–24. https://doi.org/10.1111/gcb.13475 Nago, S. G. A., Grell, O., Sinsin, B., & Rödel, M.-O. (2006). The amphibian fauna of Pendjari National Park and surroundings, northern Benin. Salamandra, 42(2–3), 93–108. Nei, M., & Kumar, S. (2000). Molecular Evolution and Phylogenetics. Oxford University Press. Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, Dr., Chan, K. M., Daily, G. C., Goldstein, J., Kareiva, P. M., Lonsdorf, E., Naidoo, R., Ricketts, T. H., & Shaw, Mr. (2009). Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Frontiers in Ecology and the Environment, 7(1), 4–11. https://doi.org/10.1890/080023 Nicholson, S. E. (2013). The West African Sahel: a review of recent studies on the rainfall regime and its interannual variability. ISRN Meteorology, 2013(453521), 1– 32. https://doi.org/10.1155/2013/453521 Nielsen, R., Matz, M., & Lewis, P. (2006). Statistical approaches for DNA barcoding. Systematic Biology, 55(1), 162–169. https://doi.org/10.1080/10635150500431239 Ohler, A., & Frétey, T. (2008). Status of the name Arthroleptis milletihorsini Angel, 1922 (Amphibia , Anura). Alytes, 25, 173–175. Olden, J. D., & Jackson, D. A. (2000). Torturing data for the sake of generality: how valid are our regression models? Écoscience, 7(4), 501–510. https://doi.org/10.1080/11956860.2000.11682622 Padial, J. M., Crochet, P.-A., Geniez, P., & Brito, J. C. (2013). Amphibian conservation in Mauritania. Basic and Applied Herpetology, 27(January), 11–22. https://doi.org/10.11160/bah.13002 Peloso, P. L. V., Frost, D. R., Richards, S. J., Rodrigues, M. T., Donnellan, S., Matsui, M., Raxworthy, C. J., Biju, S. D., Lemmon, E. M., Lemmon, A. R., & Wheeler, W. C. (2016). The impact of anchored phylogenomics and taxon sampling on phylogenetic inference in narrow-mouthed frogs (Anura, Microhylidae). Cladistics, FCUP 61 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

32(2), 113–140. https://doi.org/10.1111/cla.12118 Penado, A., Rebelo, H., & Goulson, D. (2016). Spatial distribution modelling reveals climatically suitable areas for bumblebees in undersampled parts of the Iberian Peninsula. Insect Conservation and Diversity, 9(5), 391–401. https://doi.org/10.1111/icad.12190 Phillips, S. J., Anderson, R. P., & Schapire, R. E. (2006). Maximum entropy modeling of species geographic distributions. Ecological Modelling, 190(3–4), 231–259. https://doi.org/10.1016/j.ecolmodel.2005.03.026 Pons, J., Barraclough, T. G., Gomez-Zurita, J., Cardoso, A., Duran, D. P., Hazell, S., Kamoun, S., Sumlin, W. D., Vogler, A. P., & Hedin, M. (2006). Sequence-based species delimitation for the DNA taxonomy of undescribed insects. Systematic Biology, 55(4), 595–609. https://doi.org/10.1080/10635150600852011 Portik, D. M., & Blackburn, D. C. (2016). The evolution of reproductive diversity in Afrobatrachia: a phylogenetic comparative analysis of an extensive radiation of African frogs. Evolution, 70(9), 2017–2032. https://doi.org/10.1111/evo.12997 Pyron, R. A., & Wiens, J. J. (2011). A large-scale phylogeny of Amphibia including over 2800 species, and a revised classification of extant frogs, salamanders, and caecilians. Molecular Phylogenetics and Evolution, 61(2), 543–583. https://doi.org/10.1016/j.ympev.2011.06.012 R Core Team. (2017). R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available athttp://www.r- project.org/. https://doi.org/http://www.R-project.org/ Rambaut, A., Drummond, A. J., & Suchard, M. (2003). Tracer v1.6: MCMC trace analysis package. Institute of Evolutionary Biology, Department of Computer Science. Available at http://tree.bio.ed.ac.uk/software/tracer/. Ratnasingham, S., & Hebert, P. D. N. (2007). BOLD: The Barcode of Life Data System (http://www.barcodinglife.org). Molecular Ecology Notes, 7(3), 355–364. https://doi.org/10.1111/j.1471-8286.2007.01678.x Ratnasingham, S., & Hebert, P. D. N. (2013). A DNA-based registry for all animal species: the Barcode Index Number (BIN) system. PLoS ONE, 8(7), e66213. https://doi.org/10.1371/journal.pone.0066213 Raupach, M. J., Barco, A., Steinke, D., Beermann, J., Laakmann, S., Mohrbeck, I., Neumann, H., Kihara, T. C., Pointner, K., Radulovici, A., Segelken-Voigt, A., Wesse, C., & Knebelsberger, T. (2015). The application of DNA barcodes for the identification of marine crustaceans from the North Sea and adjacent regions. PLOS ONE, 10(9), e0139421. https://doi.org/10.1371/journal.pone.0139421 FCUP 62 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Richardson, D. M. (2012). Conservation biogeography: what’s hot and what’s not? Diversity and Distributions, 18(4), 319–322. https://doi.org/10.1111/j.1472- 4642.2012.00910.x Riley, S. J., DeGloria, S. D., & Elliot, R. (1999). A terrain ruggedness index that quantifies topographic heterogeneity. Intermountain Journal of Sciences. https://doi.org/citeulike-article-id:8858430 Rockney, H. J., Ofori-Boateng, C., Porcino, N., & Leaché, A. D. (2015). A comparison of DNA barcoding markers in West African frogs. African Journal of Herpetology, 64(2), 135–147. https://doi.org/10.1080/21564574.2015.1114530 Rödel, M.-O. (2000). Herpetofauna of West Africa: Amphibians of the West African Savanna (Vol. I). Frankfurt am Main: Edition Chimaira. Roelants, K., Gower, D. J., Wilkinson, M., Loader, S. P., Biju, S. D., Guillaume, K., Moriau, L., & Bossuyt, F. (2007). Global patterns of diversification in the history of modern amphibians. Proceedings of the National Academy of Sciences, 104(3), 887–892. https://doi.org/10.1073/pnas.0608378104 Rosauer, D. F., Catullo, R. A., VanDerWal, J., Moussalli, A., & Moritz, C. (2015). Lineage range estimation method reveals fine-scale endemism linked to Pleistocene stability in Australian rainforest herpetofauna. PLOS ONE, 10(5), e0126274. https://doi.org/10.1371/journal.pone.0126274 Roscioni, F., Rebelo, H., Russo, D., Carranza, M. L., Di Febbraro, M., & Loy, A. (2014). A modelling approach to infer the effects of wind farms on landscape connectivity for bats. Landscape Ecology, 29(5), 891–903. https://doi.org/10.1007/s10980-014- 0030-2 Rosenberg, N. A. (2003). The shapes of neutral gene genealogies in two species: probabilities of monophyly, paraphyly, and polyphyly in a coalescent model. Evolution, 57(7), 1465–1477. https://doi.org/10.1111/j.0014-3820.2003.tb00355.x Roslin, T., Majaneva, S., & Clare, E. (2016). The use of DNA barcodes in food web construction—terrestrial and aquatic ecologists unite! Genome, 59(9), 603–628. https://doi.org/10.1139/gen-2015-0229 Sá, F. P. De, Canedo, C., Lyra, M. L., & Haddad, C. F. B. (2015). A new species of Hylodes (Anura, Hylodidae) and its secretive underwater breeding behavior. Herpetologica, 71(1), 58–71. https://doi.org/10.1655 San Mauro, D., García-París, M., & Zardoya, R. (2004a). Phylogenetic relationships of discoglossid frogs (Amphibia:Anura:Discoglossidae) based on complete mitochondrial genomes and nuclear genes. Gene, 343(2), 357–366. https://doi.org/10.1016/j.gene.2004.10.001 FCUP 63 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

San Mauro, D., Gower, D. J., Oommen, O. V., Wilkinson, M., & Zardoya, R. (2004b). Phylogeny of caecilian amphibians (Gymnophiona) based on complete mitochondrial genomes and nuclear RAG1. Molecular Phylogenetics and Evolution, 33(2), 413–427. https://doi.org/10.1016/j.ympev.2004.05.014 Sánchez-Vialas, A., Calvo-Revuelta, M., & Márquez, R. (2017). Ptychadena in Mauritania and the first record of Ptychadena schillukorum. ZooKeys, 673, 125– 133. https://doi.org/10.3897/zookeys.673.10265 Sano, N., Kurabayashi, A., Fujii, T., Yonekawa, H., & Sumida, M. (2004). Complete nucleotide sequence and gene rearrangement of the mitochondrial genome of the bell-ring frog, Buergeria buergeri (family Rhacophoridae). Genes & Genetic Systems, 79(3), 151–163. https://doi.org/10.1266/ggs.79.151 Sayre, R., Comer, P., Hak, J., Josse, C., Bow, J., Warner, H., Larwanou, M., Kelbessa, E., Bekele, T., Kehl, H., Amena, R., Andriamasimanana, R., Ba, T., Benson, L., Boucher, T., Brown, M., Cress, J., Dassering, O., Friesen, B., Gachathi, F., Houcine, S., Keita, M., Khamala, E., Marangu, D., Mokua, F., Morou, B., Mucina, L., Mugisha, S., Mwavu, E., Rutherford, M., Sanou, P., Syampungani, S., Tomor, B., Vall, A., Weghe, J. Vande, Wangui, E., & Waruingi, L. (2013). A new map of standardized terrestrial ecosystems of Africa. Washington, DC: Association of American Geographers: Available at http://rmgsc.cr.usgs.gov/ecosystems/docs/. Accessed 24.09.2017 Schwalm, C. R., Anderegg, W. R. L., Michalak, A. M., Fisher, J. B., Biondi, F., Koch, G., Litvak, M., Ogle, K., Shaw, J. D., Wolf, A., Huntzinger, D. N., Schaefer, K., Cook, R., Wei, Y., Fang, Y., Hayes, D., Huang, M., Jain, A., & Tian, H. (2017). Global patterns of drought recovery. Nature, 548(7666), 202–205. https://doi.org/10.1038/nature23021 Semlitsch, R. D., & Bodie, J. R. (2003). Biological criteria for buffer zones around wetlands and riparian habitats for amphibians and reptiles. Conservation Biology, 17(5), 1219–1228. https://doi.org/10.1046/j.1523-1739.2003.02177.x Shine, T. (2003). The conservation status of eastern Mauritania’s ephemeral wetlands and their role in the migration and wintering of black storks (Ciconia nigra). Aves, 40, 228–240. Simeone, M. C., Piredda, R., Papini, A., Vessella, F., & Schirone, B. (2013). Application of plastid and nuclear markers to DNA barcoding of Euro-Mediterranean oaks (Quercus, Fagaceae): problems, prospects and phylogenetic implications. Botanical Journal of the Linnean Society, 172(4), 478–499. https://doi.org/10.1111/boj.12059 FCUP 64 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Smith, M. A., Eveleigh, E. S., McCann, K. S., Merilo, M. T., McCarthy, P. C., & Van Rooyen, K. I. (2011). Barcoding a quantified food web: crypsis, concepts, ecology and hypotheses. PLoS ONE, 6(7), e14424. https://doi.org/10.1371/journal.pone.0014424 Smith, M. A., Poyarkov, N. A., & Hebert, P. D. N. (2008). CO1 DNA barcoding amphibians: take the chance, meet the challenge. Molecular Ecology Resources, 8(2), 235–246. https://doi.org/10.1111/j.1471-8286.2007.01964.x Snee, R. D. (1977). Validation of regression models: methods and examples. Technometrics, 19(4), 415. https://doi.org/10.2307/1267881 Snodgrass, J. W., Komoroski, M. J., Bryan, A. L., & Burger, J. (2000). Relationships among isolated wetland size, hydroperiod, and amphibian species richness: implications for wetland regulations. Conservation Biology, 14(2), 414–419. https://doi.org/10.1046/j.1523-1739.2000.99161.x Spector, S. (2002). Biogeographic crossroads as priority areas for biodiversity conservation. Conservation Biology, 16(6), 1480–1487. https://doi.org/10.1046/j.1523-1739.2002.00573.x Stuart, S. N. (2004). Status and trends of amphibian declines and extinctions worldwide. Science, 306(5702), 1783–1786. https://doi.org/10.1126/science.1103538 Stuart, S. N., Hoffmann, M., Chanson, J. S., Cox, N. A., Berridge, R. J., Ramani, P., & Young, B. E. (2008). Threatened amphibians of the world. Barcelona: Lynx Edicions; Gland: IUCN; Arlington: Conservation International. Suhling, F., Schenk, K., Padeffke, T., & Martens, A. (2004). A field study of larval development in a dragonfly assemblage in African desert ponds (Odonata). Hydrobiologia, 528(1–3), 75–85. https://doi.org/10.1007/s10750-004-3047-8 Sutherland, W. J., Bardsley, S., Clout, M., Depledge, M. H., Dicks, L. V., Fellman, L., Fleishman, E., Gibbons, D. W., Keim, B., Lickorish, F., Margerison, C., Monk, K. A., Norris, K., Peck, L. S., Prior, S. V., Scharlemann, J. P. W., Spalding, M. D., & Watkinson, A. R. (2013). A horizon scan of global conservation issues for 2013. Trends in Ecology & Evolution, 28(1), 16–22. https://doi.org/10.1016/j.tree.2012.10.022 Taberlet, P., Coissac, E., Pompanon, F., Brochmann, C., & Willerslev, E. (2012). Towards next-generation biodiversity assessment using DNA metabarcoding. Molecular Ecology, 21(8), 2045–50. https://doi.org/10.1111/j.1365- 294X.2012.05470.x Talavera, G., Dincă, V., & Vila, R. (2013). Factors affecting species delimitations with FCUP 65 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

the GMYC model: insights from a butterfly survey. Methods in Ecology and Evolution, 4(12), 1101–1110. https://doi.org/10.1111/2041-210X.12107 Trape, S. (2009). Impact of climate change on the relict tropical fish fauna of central Sahara: threat for the survival of Adrar mountains fishes, Mauritania. PLoS ONE, 4(2), e4400. https://doi.org/10.1371/journal.pone.0004400 Trifinopoulos, J., Nguyen, L.-T., von Haeseler, A., & Minh, B. Q. (2016). W-IQ-TREE: a fast online phylogenetic tool for maximum likelihood analysis. Nucleic Acids Research, 44(18), 1–4. https://doi.org/10.1093/nar/gkw256 UNDP. (2016). Human Development Report 2016. United Nations Development Programme. Available at http://hdr.undp.org/en/2016-report. Accessed 20.08.2017 Vale, C. G., & Brito, J. C. (2015). Desert-adapted species are vulnerable to climate change: insights from the warmest region on Earth. Global Ecology and Conservation, 4, 369–379. https://doi.org/10.1016/j.gecco.2015.07.012 Vale, C. G., Pimm, S. L., & Brito, J. C. (2015). Overlooked mountain rock pools in deserts are critical local hotspots of biodiversity. PLOS ONE, 10(2), e0118367. https://doi.org/10.1371/journal.pone.0118367 van der Meijden, A., Vences, M., Hoegg, S., Boistel, R., Channing, A., & Meyer, A. (2007). Nuclear gene phylogeny of narrow-mouthed toads (Family: Microhylidae) and a discussion of competing hypotheses concerning their biogeographical origins. Molecular Phylogenetics and Evolution, 44(3), 1017–1030. https://doi.org/10.1016/j.ympev.2007.02.008 van der Meijden, A., Vences, M., Hoegg, S., & Meyer, A. (2005). A previously unrecognized radiation of ranid frogs in Southern Africa revealed by nuclear and mitochondrial DNA sequences. Molecular Phylogenetics and Evolution, 37(3), 674–685. https://doi.org/10.1016/j.ympev.2005.05.001 van der Meijden, A., Vences, M., & Meyer, A. (2004). Novel phylogenetic relationships of the enigmatic brevicipitine and scaphiophrynine toads as revealed by sequences from the nuclear Rag-1 gene. Proceedings of the Royal Society B: Biological Sciences, 271(Suppl_5), S378–S381. https://doi.org/10.1098/rsbl.2004.0196 Vasconcelos, R., Montero-Mendieta, S., Simó-Riudalbas, M., Sindaco, R., Santos, X., Fasola, M., Llorente, G., Razzetti, E., & Carranza, S. (2016). Unexpectedly high levels of cryptic diversity uncovered by a complete DNA barcoding of reptiles of the Socotra Archipelago. PLOS ONE, 11(3), e0149985. https://doi.org/10.1371/journal.pone.0149985 Velo-Antón, G., Godinho, R., Campos, J. C., & Brito, J. C. (2014). Should I stay or FCUP 66 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

should I go? Dispersal and population structure in small, isolated desert populations of west African crocodiles. PLoS ONE, 9(4), e94626. https://doi.org/10.1371/journal.pone.0094626 Vences, M., Thomas, M., Bonett, R. M., & Vieites, D. R. (2005a). Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philosophical Transactions of the Royal Society B: Biological Sciences, 360(1462), 1859–1868. https://doi.org/10.1098/rstb.2005.1717 Vences, M., Thomas, M., van der Meijden, A., Chiari, Y., & Vieites, D. R. (2005b). Comparative performance of the 16S rRNA gene in DNA barcoding of amphibians. Frontiers in Zoology, 2, 5. https://doi.org/10.1186/1742-9994-2-5 Venne, L. S., Tsai, J.-S., Cox, S. B., Smith, L. M., & McMurry, S. T. (2012). Amphibian community richness in cropland and grassland playas in the Southern High Plains, USA. Wetlands, 32(4), 619–629. https://doi.org/10.1007/s13157-012-0305-9 Villesen, P. (2007). FaBox: an online toolbox for FASTA sequences. Molecular Ecology Notes, 7(6), 965–968. https://doi.org/10.1111/j.1471-8286.2007.01821.x Ward, D. (2016). The Biology of Deserts. Oxford: Oxford University Press Wesener, T., & Conrad, C. (2016). Local hotspots of endemism or artifacts of incorrect taxonomy? The status of microendemic pill millipede species of the genus Glomeris in Northern Italy (Diplopoda, Glomerida). PLOS ONE, 11(9), e0162284. https://doi.org/10.1371/journal.pone.0162284 Whittaker, R. J., Araújo, M. B., Jepson, P., Ladle, R. J., Watson, J. E. M., & Willis, K. J. (2005). Conservation biogeography: assessment and prospect. Diversity and Distributions, 11(1), 3–23. https://doi.org/10.1111/j.1366-9516.2005.00143.x Wiens, J. J. (2007). Global patterns of diversification and species richness in amphibians. The American Naturalist, 170(S2), S86–S106. https://doi.org/10.1086/519396 Williamson, J., Maurin, O., Shiba, S. N. S., van der Bank, H., Pfab, M., Pilusa, M., Kabongo, R. M., van der Bank, M., & Adamowicz, S. (2016). Exposing the illegal trade in cycad species (Cycadophyta: Encephalartos) at two traditional medicine markets in South Africa using DNA barcoding. Genome, 59(9), 771–781. https://doi.org/10.1139/gen-2016-0032 Wollenberg, K. C., Vieites, D. R., Glaw, F., & Vences, M. (2011). Speciation in little: the role of range and body size in the diversification of Malagasy mantellid frogs. BMC Evolutionary Biology, 11(1), 217. https://doi.org/10.1186/1471-2148-11-217 Yahr, R., Schoch, C. L., & Dentinger, B. T. M. (2016). Scaling up discovery of hidden diversity in fungi: impacts of barcoding approaches. Philosophical Transactions of FCUP 67 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

the Royal Society B: Biological Sciences, 371(1702), 20150336. https://doi.org/10.1098/rstb.2015.0336 Yang, Z., Liu, X., Zhou, M., Ai, D., Wang, G., Wang, Y., Chu, C., & Lundholm, J. T. (2015). The effect of environmental heterogeneity on species richness depends on community position along the environmental gradient. Scientific Reports, 5(1), 15723. https://doi.org/10.1038/srep15723 Zhang, J., Kapli, P., Pavlidis, P., & Stamatakis, A. (2013a). A general species delimitation method with applications to phylogenetic placements. Bioinformatics, 29(22), 2869–2876. https://doi.org/10.1093/bioinformatics/btt499 Zhang, L. (1997). Cross-validation of non-linear growth functions for modelling tree height–diameter relationships. Annals of Botany, 79(3), 251–257. https://doi.org/10.1006/anbo.1996.0334 Zhang, P., Liang, D., Mao, R.-L., Hillis, D. M., Wake, D. B., & Cannatella, D. C. (2013b). Efficient sequencing of anuran mtDNAs and a mitogenomic exploration of the phylogeny and evolution of frogs. Molecular Biology and Evolution, 30(8), 1899–1915. https://doi.org/10.1093/molbev/mst091 Zhang, P., Zhou, H., Chen, Y.-Q., Liu, Y.-F., Qu, L.-H., & Kjer, K. (2005). Mitogenomic perspectives on the origin and phylogeny of living amphibians. Systematic Biology, 54(3), 391–400. https://doi.org/10.1080/10635150590945278 Zimkus, B. M., & Larson, J. G. (2011). Examination of the molecular relationships of sand frogs (Anura: Pyxicephalidae: Tomopterna) and resurrection of two species from the Horn of Africa. Zootaxa, 45(2933), 27–45.

FCUP 68 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Appendix

Supplementary Table 1. Amphibian samples amplified for at least one marker.

Latitude Code Species/Clade Country Province Local Markers Longitude 14.900035 7584 A. galamensis Mauritania Guidimaka Gouraye COI; RAG1 -12.451357 16.487187 10319 A. galamensis Senegal Saint-Louis Terjît, oasis COI -16.227608 15.490148 10815 A. galamensis Mauritania Gorgol Nsafenni barrage COI -12.940880 14.833805 10877 A. galamensis Mauritania Guidimaka Guelta El Gleitât, 8km N of COI -12.346535 Gorgol el Abiod and Gorgol 16.142217 1315 H. occipitalis 1 Mauritania Gorgol COI el Akhdar river junction -13.476883 15.995950 1324 H. occipitalis 1 Mauritania Gorgol Gueltet Thor COI -12.722933 20.323193 1644 H. occipitalis 1 Mauritania Adrar Guidimoni COI -13.142101 20.580946 1714 H. occipitalis 1 Mauritania Adrar Zinder, 50km NE of COI -13.136361 20.252804 1737 H. occipitalis 1 Mauritania Adrar Dibé COI; RAG1 -13.088188 19.756953 1849 H. occipitalis 1 Mauritania Adrar Papara river COI -13.044403 16.763912 2153 H. occipitalis 1 Mauritania Hodh El Gharbi Megta es Sfeira barrage COI; RAG1 -9.770217 19.999475 2918 H. occipitalis 1 Mauritania Adrar Guelta Leouel, valley S of COI; RAG1 -13.288731 16.599938 4523 H. occipitalis 1 Mauritania Trarza Oued Soufa COI -15.764785 16.174852 4614 H. occipitalis 1 Mauritania Gorgol El Housseînîya COI; RAG1 -12.974898 15.506530 4656 H. occipitalis 1 Mauritania Gorgol Guelta Goumbel COI -12.969508 15.288627 4685 H. occipitalis 1 Mauritania Guidimaka Guelta Tin Waadine COI -12.535657 15.144240 4702 H. occipitalis 1 Mauritania Guidimaka Guelta Galoûla COI -12.009562 15.591367 4726 H. occipitalis 1 Mauritania Guidimaka Soufa, 2km E of COI -11.880490 15.576308 4733 H. occipitalis 1 Mauritania Guidimaka Guelta El Barda COI -11.943937 15.682200 4757 H. occipitalis 1 Mauritania Guidimaka Ayoûn en Na'aj COI -12.163043 17.069876 4939 H. occipitalis 1 Mauritania Assaba Aghneizir COI; RAG1 -12.688625 17.391257 4974 H. occipitalis 1 Mauritania Brakna Guelta El Barda COI -13.455600 16.470393 6121 H. occipitalis 1 Mauritania Assaba Oued el 'Adam 2 COI -12.484832 15.692380 6148 H. occipitalis 1 Mauritania Gorgol Oued el 'Adam 2 COI -12.471420 15.484232 6169 H. occipitalis 1 Mauritania Guidimaka Oued el 'Adam 2 COI -12.271150 Passes de Diégoum, 2km S 13.699577 6631 H. occipitalis 1 Niger Zinder COI; RAG1 of 9.530377 13.966488 6784 H. occipitalis 1 Niger Zinder Bednam COI; RAG1 9.280777 13.966488 6785 H. occipitalis 1 Niger Zinder El-Khom Sânîyé COI 9.280777 16.233942 7393 H. occipitalis 1 Mauritania Trarza Pass Soufa, E exit COI -16.429970 15.321832 7486 H. occipitalis 1 Mauritania Gorgol Oued Soufa COI -12.842127 15.046893 7546 H. occipitalis 1 Mauritania Guidimaka Dèsili, 6km E of COI -12.449695 14.852988 7586 H. occipitalis 1 Mauritania Guidimaka Dèsili, 6km E of COI -12.396247 FCUP 69 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude 14.846083 7613 H. occipitalis 1 Mauritania Guidimaka Garli COI -12.181675 14.995570 7618 H. occipitalis 1 Mauritania Guidimaka Tâmoûrt Goungel COI -12.184423 15.355967 7628 H. occipitalis 1 Mauritania Guidimaka Oued Garfa COI -12.210093 16.143328 7674 H. occipitalis 1 Mauritania Assaba Oued el 'Adam 3 COI -12.065630 15.376918 9399 H. occipitalis 1 Mali Kayes Koundi COI -11.598272 Gorgol el Abiod and Gorgol 14.666305 9430 H. occipitalis 1 Mali Kayes COI; RAG1 el Akhdar river junction -11.462032 14.420018 9465 H. occipitalis 1 Mali Kayes Ngouye COI -11.414922 14.917935 9595 H. occipitalis 1 Mali Kayes Kalinioro, 5km N of COI; RAG1 -9.364348 16.577108 11308 H. occipitalis 1 Mauritania Hodh El Gharbi Samba Ngoma COI -9.983632 Sagné, 3km NW of (Senegal 15.468042 11746 H. occipitalis 1 Mauritania Gorgol COI river basin) -12.882955 Ngouye Classified Forest 15.468042 11747 H. occipitalis 1 Mauritania Gorgol COI (Senegal river basin) -12.882955 15.468042 11748 H. occipitalis 1 Mauritania Gorgol Ngouye COI -12.882955 15.468042 11749 H. occipitalis 1 Mauritania Gorgol Wali, 5km SE of COI -12.882955 15.468042 11751 H. occipitalis 1 Mauritania Gorgol Sagné, 8km SE of COI -12.882955 17.126067 12765 H. occipitalis 1 Mauritania Assaba Sagné, 8km SE of COI -10.990067 16.640367 1385 H. occipitalis 2 Mauritania Assaba Wompou, 3km NW of COI; RAG1 -11.056183 17.267250 1399 H. occipitalis 2 Mauritania Tagant Wompou, 3km NW of COI; RAG1 -12.198667 17.834850 2022 H. occipitalis 2 Mauritania Tagant Wompou, 3km NW of COI -11.557833 17.249855 2069 H. occipitalis 2 Mauritania Hodh El Gharbi Guelta Galoûla COI -10.667613 17.635303 2083 H. occipitalis 2 Mauritania Assaba Papara river COI -11.323567 17.261378 2091 H. occipitalis 2 Mauritania Hodh El Gharbi Papara river COI -10.690148 17.032268 2111 H. occipitalis 2 Mauritania Hodh El Gharbi Fanguéné COI -10.245293 16.515562 2231 H. occipitalis 2 Mauritania Hodh El Gharbi Borotossou, NE of COI -10.452908 16.538033 2343 H. occipitalis 2 Mauritania Assaba Madalea COI -10.741550 16.579150 2376 H. occipitalis 2 Mauritania Assaba Hangouné Massassi COI -10.704550 15.932785 2456 H. occipitalis 2 Mauritania Guidimaka Hangouné Massassi COI; RAG1 -12.010887 16.547455 2554 H. occipitalis 2 Mauritania Assaba Hangouné Massassi COI -12.009590 16.888725 2586 H. occipitalis 2 Mauritania Assaba Koungo COI -12.184868 Ngouye Classified Forest 17.401433 2609 H. occipitalis 2 Mauritania Tagant COI (Senegal river basin) -12.364150 17.737962 2669 H. occipitalis 2 Mauritania Tagant Baédiam, S of COI; RAG1 -12.245253 17.737962 2670 H. occipitalis 2 Mauritania Tagant Baédiam, N of (Karakoro) COI -12.245253 17.887298 3129 H. occipitalis 2 Mauritania Tagant Chelkha Dakhna COI -12.110844 16.756482 3331 H. occipitalis 2 Mauritania Assaba Dolol Sivré, SE of COI -11.997233 17.070297 3358 H. occipitalis 2 Mauritania Assaba Dolol Sivré, SE of COI -12.207848 3394 H. occipitalis 2 Mauritania Assaba Dolol Sivré, SE of 17.152482 COI FCUP 70 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude -12.199115 17.188344 3409 H. occipitalis 2 Mauritania Assaba Kankossa COI -12.248097 15.957078 4779 H. occipitalis 2 Mauritania Guidimaka Kankossa COI; RAG1 -12.009859 18.053485 6036 H. occipitalis 2 Mauritania Tagant Kankossa COI; RAG1 -11.942891 16.540090 6086 H. occipitalis 2 Mauritania Assaba Kankossa COI -10.801490 16.338801 7694 H. occipitalis 2 Mauritania Assaba Ngouye COI; RAG1 -11.978097 16.297145 7718 H. occipitalis 2 Mauritania Assaba Papara river COI -12.004692 15.934512 7724 H. occipitalis 2 Mauritania Guidimaka Fanguéné COI; RAG1 -11.999650 16.002553 7745 H. occipitalis 2 Mauritania Assaba Moulessimou COI -11.871748 15.944687 7772 H. occipitalis 2 Mauritania Guidimaka Moulessimou COI; RAG1 -11.929082 15.949060 7783 H. occipitalis 2 Mauritania Guidimaka Sagné, 4km SE of COI -11.682150 16.763097 7867 H. occipitalis 2 Mauritania Assaba Gadianguer COI -11.222535 16.489910 9368 H. occipitalis 2 Mauritania Assaba Rkiz lake, W margin COI; RAG1 -11.052508 18.150443 9897 H. occipitalis 2 Mauritania Tagant Rkiz lake, W margin COI; RAG1 -12.065303 Yama Ndiaye Classified 15.712375 11055 H. occipitalis 2 Mauritania Assaba COI Forest (Senegal river basin) -11.264820 15.621197 11112 H. occipitalis 2 Mauritania Assaba Sagné, 8km SE of COI -11.289797 15.543132 11138 H. occipitalis 2 Mauritania Assaba Sagné, 8km SE of COI -11.086700 15.785913 11177 H. occipitalis 2 Mauritania Assaba Wompou, 3km NW of COI -10.795858 16.548062 11358 H. occipitalis 2 Mauritania Assaba Nouma COI -11.163975 15.944687 11881 H. occipitalis 2 Mauritania Assaba Sagné, 4km SE of COI; RAG1 -11.929082 15.903972 11897 H. occipitalis 2 Mauritania Guidimaka Testai, SE of COI; RAG1 -11.936185 15.903972 11903 H. occipitalis 2 Mauritania Guidimaka El Ghaira, 5km S of COI; RAG1 -11.936185 15.903972 11904 H. occipitalis 2 Mauritania Guidimaka El Ghaira, 5km S of COI; RAG1 -11.936185 15.985027 12536 H. occipitalis 2 Mauritania Assaba El Ghaira, 5km S of COI; RAG1 -10.877833 15.970850 12555 H. occipitalis 2 Mauritania Assaba COI; RAG1 -10.901560 16.153808 12576 H. occipitalis 2 Mauritania Assaba Testai, SE of COI; RAG1 -10.752278 16.923777 12723 H. occipitalis 2 Mauritania Hodh El Gharbi Oued Niordé COI -10.593842 14.874380 10914 K. fusca Mauritania Guidimaka Oued Niordé COI -11.902012 14.874380 10916 K. fusca Mauritania Guidimaka Oued Niordé COI -11.902012 15.903972 11905 K. fusca Mauritania Guidimaka Mbout, 18km E of COI -11.936185 15.900948 11906 K. fusca Mauritania Guidimaka Mbout, 18km E of COI; RAG1 -11.834388 15.932785 2468 K. senegalensis Mauritania Guidimaka Mbout, 18km E of COI; RAG1 -12.010887 15.330253 4678 K. senegalensis Mauritania Guidimaka Mbout, 18km E of COI; RAG1 -12.646468 15.330253 4679 K. senegalensis Mauritania Guidimaka Oudelemguil, 2km NE of COI; RAG1 -12.646468 15.330253 4680 K. senegalensis Mauritania Guidimaka Oudelemguil, 2km NE of COI -12.646468 FCUP 71 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude 15.625853 6250 K. senegalensis Mauritania Gorgol Maré de Guénétir COI; RAG1 -13.146973 16.403148 9618 K. senegalensis Mauritania Hodh El Gharbi Maré de Guénétir COI; RAG1 -9.559860 14.874380 10913 K. senegalensis Mauritania Guidimaka El Barda, N of COI -11.902012 14.874380 10915 K. senegalensis Mauritania Guidimaka Oued el 'Adam 1 COI -11.902012 15.468042 11750 K. senegalensis Mauritania Gorgol Oued el 'Adam 1 COI -12.882955 15.468042 11752 K. senegalensis Mauritania Gorgol Oued el 'Adam 1 COI -12.882955 15.634883 11825 K. senegalensis Mauritania Guidimaka Oued el 'Adam 2 COI; RAG1 -12.433415 15.944687 11880 K. senegalensis Mauritania Assaba Dolol Sivré, SE of COI -11.929082 15.908323 11869 L. bufonides Mauritania Guidimaka Sagné, 4km SE of COI; RAG1 -11.921210 Phrynobatrachus 16.679137 4568 Mauritania Trarza Sagné, 4km SE of COI sp. -15.098995 Phrynobatrachus 16.679137 4569 Mauritania Trarza Harr, 3km S of COI; RAG1 sp. -15.098995 Phrynobatrachus 16.650117 4570 Mauritania Brakna Harr, 3km S of COI sp. -14.411867 Phrynobatrachus 16.209480 4609 Mauritania Gorgol Harr, 3km E of COI; RAG1 sp. -12.974970 Phrynobatrachus 16.209480 4610 Mauritania Gorgol Harr, 3km E of COI sp. -12.974970 Phrynobatrachus 15.506530 4655 Mauritania Gorgol Harr, 3km E of COI; RAG1 sp. -12.969508 Phrynobatrachus 15.407227 4709 Mauritania Guidimaka Harr, 3km E of COI; RAG1 sp. -11.749625 Phrynobatrachus 15.576308 4748 Mauritania Guidimaka Oued Niordé COI; RAG1 sp. -11.943937 Phrynobatrachus 15.576308 4749 Mauritania Guidimaka Artémou mountain COI sp. -11.943937 Phrynobatrachus 15.222317 6204 Mauritania Gorgol Soufa, 8km W of COI sp. -12.824233 Phrynobatrachus 15.222317 6205 Mauritania Gorgol Soufa, 8km W of COI; RAG1 sp. -12.824233 Phrynobatrachus 15.490148 6225 Mauritania Gorgol Maré de Guénétir COI sp. -12.940880 Phrynobatrachus 15.490503 6229 Mauritania Gorgol Oued el 'Adam 3 COI sp. -12.942663 Phrynobatrachus 15.490503 6230 Mauritania Gorgol El Barda, N of COI; RAG1 sp. -12.942663 Phrynobatrachus 15.598700 6239 Mauritania Gorgol El Barda, N of COI sp. -13.111170 Phrynobatrachus 15.476092 7460 Mauritania Gorgol Guelta El Barda COI; RAG1 sp. -12.933600 Phrynobatrachus 15.476092 7461 Mauritania Gorgol Oued el 'Adam 1 COI sp. -12.933600 Phrynobatrachus 15.476092 7462 Mauritania Gorgol Oued el 'Adam 1 COI sp. -12.933600 Phrynobatrachus 15.476092 7463 Mauritania Gorgol Oued el 'Adam 1 COI sp. -12.933600 Phrynobatrachus 15.321832 7490 Mauritania Gorgol Oued el 'Adam 2 COI sp. -12.842127 Phrynobatrachus 15.321832 7491 Mauritania Gorgol Oued el 'Adam 2 COI sp. -12.842127 Phrynobatrachus 15.321832 7492 Mauritania Gorgol Oued el 'Adam 2 COI; RAG1 sp. -12.842127 Phrynobatrachus 15.321832 7493 Mauritania Gorgol Pass Soufa, E exit COI sp. -12.842127 Phrynobatrachus 15.159485 7504 Mauritania Gorgol Laout, 4km S of COI; RAG1 sp. -12.761298 Phrynobatrachus 15.159485 7505 Mauritania Gorgol El Ghaira, 5km S of COI; RAG1 sp. -12.761298 7506 Phrynobatrachus Mauritania Gorgol El Ghaira, 5km S of 15.159485 COI FCUP 72 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude sp. -12.761298 Phrynobatrachus 15.147870 7509 Mauritania Gorgol Toueichît, 5km NE of COI; RAG1 sp. -12.734438 Phrynobatrachus 15.147870 7510 Mauritania Gorgol Toueichît, 5km NE of COI; RAG1 sp. -12.734438 Phrynobatrachus 15.147870 7511 Mauritania Gorgol Gânçai source, 4km S of COI sp. -12.734438 Phrynobatrachus 15.147870 7512 Mauritania Gorgol Gânçai source, 4km S of COI; RAG1 sp. -12.734438 Phrynobatrachus 15.046893 7549 Mauritania Guidimaka Gânçai source, 4km S of COI sp. -12.449695 Phrynobatrachus 15.046893 7550 Mauritania Guidimaka Oued Achram COI sp. -12.449695 Phrynobatrachus 15.046893 7551 Mauritania Guidimaka Fada N'Gourna, 100km S of COI sp. -12.449695 Phrynobatrachus Fada N'Gourma: Hotel 15.046893 7552 Mauritania Guidimaka COI sp. Panache -12.449695 Phrynobatrachus 14.799003 7605 Mauritania Guidimaka Kayes, 20km W of COI sp. -12.221147 Phrynobatrachus Niamey; Hotel Les Rouniers 14.799003 7606 Mauritania Guidimaka COI sp. Doum -12.221147 Phrynobatrachus 14.799003 7607 Mauritania Guidimaka Marigot SE COI sp. -12.221147 Phrynobatrachus 16.338801 7695 Mauritania Assaba Ndjamena hotel near airport COI sp. -11.978097 Phrynobatrachus 16.338801 7696 Mauritania Assaba Aleg, 65km NE of COI sp. -11.978097 Phrynobatrachus 16.338801 7697 Mauritania Assaba Air: Guelta of Timia COI; RAG1 sp. -11.978097 Phrynobatrachus 16.002553 7735 Mauritania Assaba Kayes, 120km E of COI sp. -11.871748 Phrynobatrachus 14.420018 9458 Mali Kayes Oued Toueirga COI sp. -11.414922 Phrynobatrachus 14.420018 9459 Mali Kayes Guelb Samba COI sp. -11.414922 Phrynobatrachus 14.420018 9460 Mali Kayes Rosso COI; RAG1 sp. -11.414922 Phrynobatrachus Gorgol el Abiod and Gorgol 14.420018 9461 Mali Kayes COI; RAG1 sp. el Akhdar river junction -11.414922 Phrynobatrachus 14.247392 9487 Mali Kayes Nelbi, Ferlo COI; RAG1 sp. -11.256788 Phrynobatrachus 14.247392 9488 Mali Kayes Zinder COI sp. -11.256788 Phrynobatrachus 14.247392 9489 Mali Kayes Ngouye COI sp. -11.256788 Phrynobatrachus 14.108748 9508 Mali Kayes Sagné, 8km SE of COI sp. -11.022875 Phrynobatrachus 14.108748 9509 Mali Kayes Diogountourou, 8km N of COI sp. -11.022875 Phrynobatrachus 14.108748 9510 Mali Kayes Guelta Galoûla COI; RAG1 sp. -11.022875 Phrynobatrachus 14.223133 9535 Mali Kayes Gobernie COI; RAG1 sp. -10.657137 Phrynobatrachus 14.274715 9537 Mali Kayes Baédiam, N of (Karakoro) COI sp. -10.527250 Phrynobatrachus 14.274715 9538 Mali Kayes Bovél, S of COI sp. -10.527250 Phrynobatrachus 14.274715 9539 Mali Kayes Bovél, S of COI sp. -10.527250 Phrynobatrachus 14.612575 9565 Mali Kayes Nima COI; RAG1 sp. -9.215873 Phrynobatrachus 14.612575 9566 Mali Kayes Nouma COI; RAG1 sp. -9.215873 Phrynobatrachus 14.612575 9567 Mali Kayes Nouma COI; RAG1 sp. -9.215873 Phrynobatrachus 14.918820 9573 Mali Kayes Toulel, 5km S of COI; RAG1 sp. -9.367417 Phrynobatrachus 14.918820 9576 Mali Kayes Sagné, 4km SE of COI sp. -9.367417 FCUP 73 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude Phrynobatrachus 14.918820 9577 Mali Kayes Sagné, 4km SE of COI sp. -9.367417 Phrynobatrachus 15.490148 10806 Mauritania Gorgol Oudelemguil, 2km NE of COI sp. -12.940880 Phrynobatrachus Passes de Diégoum, 2km S 15.490148 10807 Mauritania Gorgol COI; RAG1 sp. of -12.940880 Phrynobatrachus 15.490148 10808 Mauritania Gorgol Tîntâne, 3km E of COI sp. -12.940880 Phrynobatrachus 15.490148 10809 Mauritania Gorgol Irad well, 5km S of COI sp. -12.940880 Phrynobatrachus 15.490148 10810 Mauritania Gorgol Irad well, 5km S of COI sp. -12.940880 Phrynobatrachus 15.490148 10811 Mauritania Gorgol Irad well, 5km S of COI sp. -12.940880 Phrynobatrachus 15.490148 10812 Mauritania Gorgol Irad well, 5km S of COI sp. -12.940880 Phrynobatrachus 15.490148 10813 Mauritania Gorgol Oumm el Khezz, 5km E of COI sp. -12.940880 Phrynobatrachus 15.357202 10823 Mauritania Gorgol Tâmoûrt Ayyer COI sp. -12.838435 Phrynobatrachus 15.357202 10824 Mauritania Gorgol Tâmoûrt Ayyer COI sp. -12.838435 Phrynobatrachus 15.357202 10825 Mauritania Gorgol Pass Soufa, E exit COI sp. -12.838435 Phrynobatrachus 15.110205 10833 Mauritania Guidimaka Guelta Laout COI sp. -12.661363 Phrynobatrachus 15.110717 10835 Mauritania Guidimaka El Ghaira, 5km S of COI sp. -12.653207 Phrynobatrachus 15.110717 10836 Mauritania Guidimaka El Ghaira, 5km S of COI sp. -12.653207 Phrynobatrachus 14.833805 10868 Mauritania Guidimaka Oued Achram COI sp. -12.346535 Phrynobatrachus 14.833805 10869 Mauritania Guidimaka Oued Achram COI sp. -12.346535 Phrynobatrachus 14.833805 10870 Mauritania Guidimaka Moulessimou COI sp. -12.346535 Phrynobatrachus 14.832068 10911 Mauritania Guidimaka Gouraye COI sp. -11.933057 Phrynobatrachus 14.874930 10928 Mauritania Guidimaka Terjît, oasis COI sp. -11.889280 Phrynobatrachus 14.874930 10929 Mauritania Guidimaka Nsafenni barrage COI sp. -11.889280 Phrynobatrachus 14.874930 10930 Mauritania Guidimaka Guelta El Gleitât, 8km N of COI sp. -11.889280 Phrynobatrachus Gorgol el Abiod and Gorgol 15.027553 10943 Mauritania Guidimaka COI sp. el Akhdar river junction -11.848473 Phrynobatrachus 15.027553 10944 Mauritania Guidimaka Gueltet Thor COI sp. -11.848473 Phrynobatrachus 15.027553 10945 Mauritania Guidimaka Guidimoni COI; RAG1 sp. -11.848473 Phrynobatrachus 15.092013 10956 Mauritania Guidimaka Zinder, 50km NE of COI sp. -11.849033 Phrynobatrachus 15.092013 10957 Mauritania Guidimaka Dibé COI; RAG1 sp. -11.849033 Phrynobatrachus 15.092013 10958 Mauritania Guidimaka Papara river COI sp. -11.849033 Phrynobatrachus 15.092013 10962 Mauritania Guidimaka Megta es Sfeira barrage COI sp. -11.849033 Phrynobatrachus 15.386320 10971 Mauritania Guidimaka Guelta Leouel, valley S of COI sp. -11.759625 Phrynobatrachus 15.472977 10974 Mauritania Guidimaka Oued Soufa COI; RAG1 sp. -11.733532 Phrynobatrachus 15.472977 10975 Mauritania Guidimaka El Housseînîya COI sp. -11.733532 Phrynobatrachus 15.472977 10976 Mauritania Guidimaka Guelta Goumbel COI sp. -11.733532 Phrynobatrachus 15.694747 11071 Mauritania Assaba Guelta Tin Waadine COI sp. -11.237802 FCUP 74 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude Phrynobatrachus 15.694747 11072 Mauritania Assaba Guelta Galoûla COI sp. -11.237802 Phrynobatrachus 15.694747 11073 Mauritania Assaba Soufa, 2km E of COI sp. -11.237802 Phrynobatrachus 15.621197 11119 Mauritania Assaba Guelta El Barda COI sp. -11.289797 Phrynobatrachus 15.621197 11120 Mauritania Assaba Ayoûn en Na'aj COI sp. -11.289797 Phrynobatrachus 15.621197 11121 Mauritania Assaba Aghneizir COI sp. -11.289797 Phrynobatrachus 15.547173 11133 Mauritania Assaba Guelta El Barda COI sp. -11.087957 Phrynobatrachus 15.547173 11134 Mauritania Assaba Oued el 'Adam 2 COI sp. -11.087957 Phrynobatrachus 15.547173 11135 Mauritania Assaba Oued el 'Adam 2 COI sp. -11.087957 Phrynobatrachus 15.554275 11739 Mauritania Gorgol Oued el 'Adam 2 COI; RAG1 sp. -13.048047 Phrynobatrachus Passes de Diégoum, 2km S 15.554275 11740 Mauritania Gorgol COI; RAG1 sp. of -13.048047 Phrynobatrachus 15.554275 11741 Mauritania Gorgol Bednam COI; RAG1 sp. -13.048047 Phrynobatrachus 15.922192 12416 Mauritania Assaba El-Khom Sânîyé COI; RAG1 sp. -11.544487 Phrynobatrachus 15.922192 12417 Mauritania Assaba Pass Soufa, E exit COI; RAG1 sp. -11.544487 Phrynobatrachus 15.922192 12418 Mauritania Assaba Oued Soufa COI; RAG1 sp. -11.544487 Phrynobatrachus 15.922192 12419 Mauritania Assaba Dèsili, 6km E of COI; RAG1 sp. -11.544487 15.598700 6240 P. bibroni Mauritania Gorgol Dèsili, 6km E of COI -13.111170 15.476092 7448 P. bibroni Mauritania Gorgol Garli COI; RAG1 -12.933600 14.420018 9457 P. bibroni Mali Kayes Tâmoûrt Goungel COI; RAG1 -11.414922 14.247392 9483 P. bibroni Mali Kayes Oued Garfa COI; RAG1 -11.256788 14.247392 9486 P. bibroni Mali Kayes Oued el 'Adam 3 COI -11.256788 15.110205 10832 P. bibroni Mauritania Guidimaka Koundi COI -12.661363 Gorgol el Abiod and Gorgol 14.833805 10871 P. bibroni Mauritania Guidimaka COI el Akhdar river junction -12.346535 14.833805 10872 P. bibroni Mauritania Guidimaka Ngouye COI; RAG1 -12.346535 14.833805 10873 P. bibroni Mauritania Guidimaka Kalinioro, 5km N of COI -12.346535 14.833805 10874 P. bibroni Mauritania Guidimaka Samba Ngoma COI -12.346535 Sagné, 3km NW of (Senegal 14.833805 10876 P. bibroni Mauritania Guidimaka COI; RAG1 river basin) -12.346535 Ngouye Classified Forest 15.176985 11777 P. bibroni Mauritania Gorgol COI; RAG1 (Senegal river basin) -12.774027 16.403148 2214 P. schillukorum Mauritania Hodh El Gharbi Ngouye COI -9.559860 16.378000 4498 P. schillukorum Mauritania Trarza Wali, 5km SE of COI; RAG1 -16.386700 16.810068 4534 P. schillukorum Mauritania Trarza Sagné, 8km SE of COI; RAG1 -15.416052 16.810068 4535 P. schillukorum Mauritania Trarza Sagné, 8km SE of COI; RAG1 -15.416052 16.810068 4537 P. schillukorum Mauritania Trarza Wompou, 3km NW of COI -15.416052 16.810068 4538 P. schillukorum Mauritania Trarza Wompou, 3km NW of COI -15.416052 16.810068 4539 P. schillukorum Mauritania Trarza Wompou, 3km NW of COI -15.416052 FCUP 75 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude 15.598700 6244 P. schillukorum Mauritania Gorgol Guelta Galoûla COI; RAG1 -13.111170 15.598700 6245 P. schillukorum Mauritania Gorgol Papara river COI -13.111170 15.598700 6246 P. schillukorum Mauritania Gorgol Papara river COI -13.111170 15.159485 7501 P. schillukorum Mauritania Gorgol Fanguéné COI -12.761298 15.159485 7502 P. schillukorum Mauritania Gorgol Borotossou, NE of COI; RAG1 -12.761298 15.159485 7503 P. schillukorum Mauritania Gorgol Madalea COI; RAG1 -12.761298 15.147870 7913 P. schillukorum Mauritania Gorgol Hangouné Massassi COI; RAG1 -12.734438 15.490148 10814 P. schillukorum Mauritania Gorgol Hangouné Massassi COI -12.940880 16.551782 11303 P. schillukorum Mauritania Hodh El Gharbi Hangouné Massassi COI -9.956650 16.551782 11304 P. schillukorum Mauritania Hodh El Gharbi Koungo COI -9.956650 Ngouye Classified Forest 16.551782 11305 P. schillukorum Mauritania Hodh El Gharbi COI (Senegal river basin) -9.956650 16.596373 11313 P. schillukorum Mauritania Hodh El Gharbi Baédiam, S of COI -9.979062 16.548062 11356 P. schillukorum Mauritania Assaba Baédiam, N of (Karakoro) COI -11.163975 16.548062 11357 P. schillukorum Mauritania Assaba Chelkha Dakhna COI -11.163975 15.369968 11755 P. schillukorum Mauritania Gorgol Dolol Sivré, SE of COI; RAG1 -12.848953 15.176985 11779 P. schillukorum Mauritania Gorgol Dolol Sivré, SE of COI; RAG1 -12.774027 15.244178 11804 P. schillukorum Mauritania Guidimaka Dolol Sivré, SE of COI; RAG1 -12.423562 17.152018 11927 P. schillukorum Mauritania Assaba Kankossa COI; RAG1 -12.336408 17.152018 11928 P. schillukorum Mauritania Assaba Kankossa COI; RAG1 -12.336408 17.152018 11929 P. schillukorum Mauritania Assaba Kankossa COI; RAG1 -12.336408 15.878200 2472 P. trinodis Mauritania Guidimaka Kankossa COI -12.039233 16.002553 7748 P. trinodis Mauritania Assaba Ngouye COI; RAG1 -11.871748 15.244178 11803 P. trinodis Mauritania Guidimaka Papara river COI; RAG1 -12.423562 15.296210 11814 P. trinodis Mauritania Guidimaka Fanguéné COI; RAG1 -12.316357 15.296210 11815 P. trinodis Mauritania Guidimaka Moulessimou COI; RAG1 -12.316357 15.296210 11816 P. trinodis Mauritania Guidimaka Moulessimou COI; RAG1 -12.316357 15.976007 11835 P. trinodis Mauritania Gorgol Sagné, 4km SE of COI; RAG1 -12.405287 15.976007 11836 P. trinodis Mauritania Gorgol Gadianguer COI; RAG1 -12.405287 15.976007 11837 P. trinodis Mauritania Gorgol Rkiz lake, W margin COI; RAG1 -12.405287 15.976007 11838 P. trinodis Mauritania Gorgol Rkiz lake, W margin COI; RAG1 -12.405287 Yama Ndiaye Classified 15.944573 11854 P. trinodis Mauritania Gorgol COI; RAG1 Forest (Senegal river basin) -12.189270 15.944573 11855 P. trinodis Mauritania Gorgol Sagné, 8km SE of COI -12.189270 15.944573 11856 P. trinodis Mauritania Gorgol Sagné, 8km SE of COI; RAG1 -12.189270 15.934090 11866 P. trinodis Mauritania Guidimaka Wompou, 3km NW of COI; RAG1 -12.046318 FCUP 76 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude 15.934090 11867 P. trinodis Mauritania Guidimaka Nouma COI; RAG1 -12.046318 15.952422 11874 P. trinodis Mauritania Guidimaka Sagné, 4km SE of COI; RAG1 -11.926680 15.899318 11891 P. trinodis Mauritania Guidimaka Testai, SE of COI; RAG1 -11.951478 15.899318 11892 P. trinodis Mauritania Guidimaka El Ghaira, 5km S of COI; RAG1 -11.951478 15.899318 11893 P. trinodis Mauritania Guidimaka El Ghaira, 5km S of COI; RAG1 -11.951478 15.903972 11898 P. trinodis Mauritania Guidimaka El Ghaira, 5km S of COI; RAG1 -11.936185 15.554275 11742 P. edulis Mauritania Gorgol Dolol Sivré, SE of COI; RAG1 -13.048047 15.244178 11805 P. edulis Mauritania Guidimaka Testai, SE of COI -12.423562 16.308000 11644 S. pentoni Mauritania Trarza Oued Niordé COI -16.412000 15.939805 11720 S. pentoni Mauritania Gorgol Oued Niordé COI -13.320462 15.939805 11721 S. pentoni Mauritania Gorgol Oued Niordé COI -13.320462 15.939805 11723 S. pentoni Mauritania Gorgol Mbout, 18km E of COI -13.320462 15.939805 11725 S. pentoni Mauritania Gorgol Mbout, 18km E of COI -13.320462 15.176985 11774 S. pentoni Mauritania Gorgol Mbout, 18km E of COI; RAG1 -12.774027 15.176985 11775 S. pentoni Mauritania Gorgol Mbout, 18km E of COI; RAG1 -12.774027 15.249180 11785 S. pentoni Mauritania Guidimaka Oudelemguil, 2km NE of COI; RAG1 -12.548060 15.249180 11788 S. pentoni Mauritania Guidimaka Oudelemguil, 2km NE of COI; RAG1 -12.548060 15.259737 11791 S. pentoni Mauritania Guidimaka Maré de Guénétir COI; RAG1 -12.483460 15.259737 11792 S. pentoni Mauritania Guidimaka Maré de Guénétir COI; RAG1 -12.483460 15.259737 11793 S. pentoni Mauritania Guidimaka El Barda, N of COI; RAG1 -12.483460 15.259737 11795 S. pentoni Mauritania Guidimaka Oued el 'Adam 1 COI; RAG1 -12.483460 15.296210 11812 S. pentoni Mauritania Guidimaka Oued el 'Adam 1 COI; RAG1 -12.316357 15.480707 11817 S. pentoni Mauritania Guidimaka Oued el 'Adam 1 COI; RAG1 -12.305692 15.480707 11820 S. pentoni Mauritania Guidimaka Oued el 'Adam 2 COI -12.305692 15.957050 11859 S. pentoni Mauritania Gorgol Dolol Sivré, SE of COI; RAG1 -12.107358 15.957050 11862 S. pentoni Mauritania Gorgol Sagné, 4km SE of COI; RAG1 -12.107358 15.934090 11864 S. pentoni Mauritania Guidimaka Sagné, 4km SE of COI; RAG1 -12.046318 15.908323 11870 S. pentoni Mauritania Guidimaka Harr, 3km S of COI; RAG1 -11.921210 15.952422 11871 S. pentoni Mauritania Guidimaka Harr, 3km S of COI; RAG1 -11.926680 15.952422 11873 S. pentoni Mauritania Guidimaka Harr, 3km E of COI; RAG1 -11.926680 15.944687 11878 S. pentoni Mauritania Assaba Harr, 3km E of COI; RAG1 -11.929082 15.899318 11887 S. pentoni Mauritania Guidimaka Harr, 3km E of COI; RAG1 -11.951478 15.899318 11888 S. pentoni Mauritania Guidimaka Harr, 3km E of COI; RAG1 -11.951478 15.899318 11890 S. pentoni Mauritania Guidimaka Oued Niordé COI; RAG1 -11.951478 11895 S. pentoni Mauritania Guidimaka Artémou mountain 15.903972 COI; RAG1 FCUP 77 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude -11.936185 15.903972 11896 S. pentoni Mauritania Guidimaka Soufa, 8km W of COI; RAG1 -11.936185 15.903972 11900 S. pentoni Mauritania Guidimaka Soufa, 8km W of COI; RAG1 -11.936185 15.900948 11908 S. pentoni Mauritania Guidimaka Maré de Guénétir COI; RAG1 -11.834388 17.240833 11913 S. pentoni Mauritania Tagant Oued el 'Adam 3 COI -12.101667 17.208523 11918 S. pentoni Mauritania Tagant El Barda, N of COI; RAG1 -12.097223 17.152018 11924 S. pentoni Mauritania Assaba El Barda, N of COI; RAG1 -12.336408 17.152018 11926 S. pentoni Mauritania Assaba Guelta El Barda COI; RAG1 -12.336408 16.722908 11931 S. pentoni Mauritania Assaba Oued el 'Adam 1 COI; RAG1 -12.288958 16.722908 11932 S. pentoni Mauritania Assaba Oued el 'Adam 1 COI; RAG1 -12.288958 17.067572 11939 S. pentoni Mauritania Assaba Oued el 'Adam 1 COI; RAG1 -12.260290 17.067572 11942 S. pentoni Mauritania Assaba Oued el 'Adam 2 COI; RAG1 -12.260290 17.067572 11945 S. pentoni Mauritania Assaba Oued el 'Adam 2 COI; RAG1 -12.260290 17.389812 11952 S. pentoni Mauritania Assaba Oued el 'Adam 2 COI; RAG1 -12.427763 12.474800 411 S. regularis Niger Niamey Pass Soufa, E exit COI 2.427600 11.962333 417 S. regularis Burkina Faso Gourma Laout, 4km S of COI; RAG1 0.392783 12.060333 423 S. regularis Burkina Faso Gourma El Ghaira, 5km S of COI; RAG1 0.369333 12.060333 424 S. regularis Burkina Faso Gourma El Ghaira, 5km S of COI 0.369333 14.504000 460 S. regularis Mali Kayes Toueichît, 5km NE of COI; RAG1 -11.090983 16.369022 4505 S. regularis Mauritania Trarza Toueichît, 5km NE of COI -16.393192 16.158185 4597 S. regularis Mauritania Gorgol Gânçai source, 4km S of COI -13.606255 13.528658 6606 S. regularis Niger Niamey Gânçai source, 4km S of COI; RAG1 2.052208 13.528658 6607 S. regularis Niger Niamey Gânçai source, 4km S of COI 2.052208 16.303340 7403 S. regularis Mauritania Trarza Oued Achram COI -16.401357 16.516551 10317 S. regularis Senegal Saint-Louis Fada N'Gourna, 100km S of COI -16.247864 Fada N'Gourma: Hotel 16.483122 10318 S. regularis Senegal Saint-Louis COI; RAG1 Panache -16.202280 12.125292 12932 S. regularis Chad Ndjamena Kayes, 20km W of COI; RAG1 15.033573 Borkou-Ennedi- Niamey; Hotel Les Rouniers 19.058798 12944 S. regularis Chad COI Tibesti Doum 20.521307 Borkou-Ennedi- 19.058798 12945 S. regularis Chad Marigot SE COI Tibesti 20.521307 16.611667 57 S. xeros Mauritania Hodh Ech Chargui Ndjamena hotel near airport COI -7.269583 17.392517 118 S. xeros Mauritania Brakna Aleg, 65km NE of COI; RAG1 -13.452850 18.094950 368 S. xeros Niger Agadez Air: Guelta of Timia COI; RAG1 8.761267 14.505667 445 S. xeros Mali Kayes Kayes, 120km E of COI -9.633000 14.511500 452 S. xeros Mali Kayes Oued Toueirga COI -9.702700 455 S. xeros Mali Kayes Guelb Samba 14.683100 COI; RAG1 FCUP 78 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude -10.384917 14.545417 462 S. xeros Mali Kayes Rosso COI -11.943033 Gorgol el Abiod and Gorgol 16.178167 473 S. xeros Senegal Saint-Louis COI el Akhdar river junction -13.997233 13.593750 994 S. xeros Mali Kayes Nelbi, Ferlo COI -10.381817 20.580946 1699 S. xeros Mauritania Adrar Zinder COI -13.136361 19.756953 1857 S. xeros Mauritania Adrar Ngouye COI; RAG1 -13.044403 18.818115 1952 S. xeros Mauritania Tagant Sagné, 8km SE of COI -11.777500 17.249855 2074 S. xeros Mauritania Hodh El Gharbi Diogountourou, 8km N of COI -10.667613 16.654987 2176 S. xeros Mauritania Hodh El Gharbi Guelta Galoûla COI; RAG1 -9.707835 17.240833 3196 S. xeros Mauritania Tagant Gobernie COI -12.101667 16.511328 4515 S. xeros Mauritania Trarza Baédiam, N of (Karakoro) COI; RAG1 -15.790067 16.810068 4531 S. xeros Mauritania Trarza Bovél, S of COI -15.416052 16.650117 4572 S. xeros Mauritania Brakna Bovél, S of COI -14.411867 16.257035 4592 S. xeros Mauritania Brakna Nima COI -13.658015 16.161588 4617 S. xeros Mauritania Gorgol Nouma COI; RAG1 -12.981702 15.338977 4662 S. xeros Mauritania Gorgol Nouma COI -12.736763 15.330253 4668 S. xeros Mauritania Guidimaka Toulel, 5km S of COI -12.646468 15.545348 4715 S. xeros Mauritania Guidimaka Sagné, 4km SE of COI -11.709218 17.069876 4941 S. xeros Mauritania Assaba Sagné, 4km SE of COI -12.688625 15.296650 5698 S. xeros Senegal Saint-Louis Oudelemguil, 2km NE of COI; RAG1 -13.964417 Passes de Diégoum, 2km S 15.824783 6140 S. xeros Mauritania Gorgol COI of -12.530327 15.598700 6237 S. xeros Mauritania Gorgol Tîntâne, 3km E of COI -13.111170 13.803345 6617 S. xeros Niger Zinder Irad well, 5km S of COI; RAG1 8.978597 15.476092 7444 S. xeros Mauritania Gorgol Irad well, 5km S of COI; RAG1 -12.933600 15.159485 7499 S. xeros Mauritania Gorgol Irad well, 5km S of COI; RAG1 -12.761298 14.799003 7604 S. xeros Mauritania Guidimaka Irad well, 5km S of COI; RAG1 -12.221147 15.355967 7633 S. xeros Mauritania Guidimaka Oumm el Khezz, 5km E of COI -12.210093 16.338801 7703 S. xeros Mauritania Assaba Tâmoûrt Ayyer COI; RAG1 -11.978097 15.816492 9602 S. xeros Mauritania Hodh El Gharbi Tâmoûrt Ayyer COI; RAG1 -9.414263 15.092013 10961 S. xeros Mauritania Guidimaka Pass Soufa, E exit COI; RAG1 -11.849033 15.341160 10969 S. xeros Mauritania Guidimaka Guelta Laout COI -11.775303 15.546645 11136 S. xeros Mauritania Assaba El Ghaira, 5km S of COI -11.087988 15.939805 11722 S. xeros Mauritania Gorgol El Ghaira, 5km S of COI -13.320462 15.939805 11724 S. xeros Mauritania Gorgol Oued Achram COI -13.320462 15.866288 11728 S. xeros Mauritania Gorgol Oued Achram COI -13.295188 FCUP 79 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude 15.866288 11729 S. xeros Mauritania Gorgol Bovél, S of COI; RAG1 -13.295188 15.866288 11730 S. xeros Mauritania Gorgol Gouraye COI; RAG1 -13.295188 15.802653 11732 S. xeros Mauritania Gorgol Terjît, oasis COI; RAG1 -13.270592 15.369968 11756 S. xeros Mauritania Gorgol Nsafenni barrage COI; RAG1 -12.848953 15.369968 11757 S. xeros Mauritania Gorgol Guelta El Gleitât, 8km N of COI; RAG1 -12.848953 Gorgol el Abiod and Gorgol 15.369968 11758 S. xeros Mauritania Gorgol COI el Akhdar river junction -12.848953 15.290632 11765 S. xeros Mauritania Gorgol Gueltet Thor COI; RAG1 -12.808530 15.290632 11766 S. xeros Mauritania Gorgol Guidimoni COI -12.808530 15.176985 11773 S. xeros Mauritania Gorgol Zinder, 50km NE of COI; RAG1 -12.774027 15.176985 11776 S. xeros Mauritania Gorgol Dibé COI; RAG1 -12.774027 15.146177 11781 S. xeros Mauritania Guidimaka Papara river COI -12.651052 15.296210 11813 S. xeros Mauritania Guidimaka Megta es Sfeira barrage COI -12.316357 15.944573 11857 S. xeros Mauritania Gorgol Guelta Leouel, valley S of COI; RAG1 -12.189270 15.985027 12533 S. xeros Mauritania Assaba Oued Soufa COI; RAG1 -10.877833 16.399033 73 T. milletihorsini Mauritania Hodh El Gharbi El Housseînîya COI; RAG1 -10.146367 16.399033 148 T. milletihorsini Mauritania Hodh El Gharbi Guelta Goumbel COI -10.146367 18.874512 1951 T. milletihorsini Mauritania Tagant Guelta Tin Waadine COI -11.817575 18.818115 1955 T. milletihorsini Mauritania Tagant Guelta Galoûla COI -11.777500 18.223743 2006 T. milletihorsini Mauritania Tagant Soufa, 2km E of COI -11.536628 18.223743 2010 T. milletihorsini Mauritania Tagant Guelta El Barda COI -11.536628 18.223743 2012 T. milletihorsini Mauritania Tagant Ayoûn en Na'aj COI -11.536628 18.223743 2013 T. milletihorsini Mauritania Tagant Aghneizir COI; RAG1 -11.536628 18.223743 2014 T. milletihorsini Mauritania Tagant Guelta El Barda COI -11.536628 18.223743 2015 T. milletihorsini Mauritania Tagant Oued el 'Adam 2 COI; RAG1 -11.536628 18.223743 2016 T. milletihorsini Mauritania Tagant Oued el 'Adam 2 COI -11.536628 18.223743 2017 T. milletihorsini Mauritania Tagant Oued el 'Adam 2 COI; RAG1 -11.536628 Passes de Diégoum, 2km S 18.223743 2018 T. milletihorsini Mauritania Tagant COI of -11.536628 18.223743 2019 T. milletihorsini Mauritania Tagant Bednam COI; RAG1 -11.536628 17.102153 2066 T. milletihorsini Mauritania Hodh El Gharbi El-Khom Sânîyé COI; RAG1 -10.955758 16.909592 2138 T. milletihorsini Mauritania Hodh El Gharbi Pass Soufa, E exit COI -10.143335 16.909592 2140 T. milletihorsini Mauritania Hodh El Gharbi Oued Soufa COI; RAG1 -10.143335 16.909592 2143 T. milletihorsini Mauritania Hodh El Gharbi Dèsili, 6km E of COI -10.143335 16.909592 2146 T. milletihorsini Mauritania Hodh El Gharbi Dèsili, 6km E of COI -10.143335 16.909592 2148 T. milletihorsini Mauritania Hodh El Gharbi Garli COI -10.143335 2416 T. milletihorsini Mauritania Assaba Tâmoûrt Goungel 16.540090 COI; RAG1 FCUP 80 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Latitude Code Species/Clade Country Province Local Markers Longitude -10.801490 16.303340 11623 T. milletihorsini Mauritania Trarza Oued Garfa COI -16.401357 16.674617 11673 T. milletihorsini Mauritania Trarza Oued el 'Adam 3 COI -16.394827 15.900948 11909 T. milletihorsini Mauritania Guidimaka Koundi COI; RAG1 -11.834388 Gorgol el Abiod and Gorgol 17.240833 11917 T. milletihorsini Mauritania Tagant COI; RAG1 el Akhdar river junction -12.101667 17.152018 11923 T. milletihorsini Mauritania Assaba Ngouye COI; RAG1 -12.336408 17.152018 11925 T. milletihorsini Mauritania Assaba Kalinioro, 5km N of COI; RAG1 -12.336408 17.389812 11950 T. milletihorsini Mauritania Assaba Samba Ngoma COI; RAG1 -12.427763 Sagné, 3km NW of (Senegal 17.389812 11951 T. milletihorsini Mauritania Assaba COI; RAG1 river basin) -12.427763

FCUP 81 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Sclerophrys

Kassina Leptopelis

Ptychadena Phrynobatrachus

Pyxicephalus Tomopterna

Hoplobatrachus Amnirana

Supplementary Figure 1. Most recent phylogenetic reconstruction for extant amphibians. Maximum likelihood tree built from 2781 species. Bootstrap proportions > 50% are shown. Genera sampled in Mauritania are shown in blue. Adapted from Pyron & Wiens, 2011.

FCUP 82 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Supplementary Table 2. Amphibian samples retrieved from GenBank for both markers.

Species Family Marker Accession Number Reference

COI NC_019998 (Irisarri et al., 2012) Heleophryne regis Heleophrynidae RAG1 AY323764 (Hoegg, 2004) COI NC_020001 (Irisarri et al., 2012) Sooglossus thomasseti Sooglossidae RAG1 AY323778 (Hoegg, 2004) COI JF703230 (Irisarri et al., 2012) Lechriodus melanopyga Myobatrachidae RAG1 AY583341 (Mauro et al., 2005) COI KJ649789 (Clemente-Carvalho et al., 2016) Brachycephalus nodoterga Brachycephalidae RAG1 JX298187 (Fouquet et al., 2012) Eleutherodactylus COI KF981374 (Melo et al., 2014) Eleutherodactylidae johnstonei RAG1 JX298190 (Fouquet et al., 2012) COI AY458593 (Zhang et al., 2005) Hyla chinensis. Hylidae RAG1 HM998976 (Irisarri et al., 2011) COI KU494819 (Lyra et al., 2017) Vitreorana uranoscopa Centrolenidae RAG1 JX298194 (Fouquet et al., 2012) COI KC520689 (Fouquet et al., 2013) Adenomera andreae Leptodactylidae RAG1 KF674176 (Fouquet et al., 2014) COI KP295688 (Faivovich et al., 2014) Ceratophrys cornuta Ceratophryidae RAG1 DQ679269 Unpublished COI KC603982 (Fouquet et al., 2013) Proceratophrys boiei Odontophrynidae RAG1 KC604004 (Fouquet et al., 2013) COI DQ502856 (Grant et al., 2006) Cycloramphus boraceiensis Cycloramphidae RAG1 KJ961590 (Sá et al., 2015) COI JX203930 (Blotto et al., 2013) Eupsophus roseus Alsodidae RAG1 KC604032 (Fouquet et al., 2013) COI KJ961558 (Sá et al., 2015) Hylodes ornatus Hylodidae RAG1 KJ961598 (Sá et al., 2015) COI JF703234 (Irisarri et al., 2012) Telmatobius bolivianus Telmatobiidae RAG1 AY583344 (Mauro et al., 2005) COI JX564855 (Zhang et al., 2013b) Callulina kreffti Brevicipitidae RAG1 EF396077 (van der Meijden et al., 2007) COI JX564868 (Zhang et al., 2013b) Hemisus marmoratus Hemisotidae RAG1 AB612013 (Kurabayashi et al., 2011) COI KM509872 (Peloso et al., 2016) Stumpffia pygmaea Microhylidae RAG1 EF396108 (van der Meijden et al., 2007) COI KJ711416 (Biju et al., 2014) Micrixalus fuscus Micrixalidae RAG1 KF991333 (Barej et al., 2014) COI AY883979 (Vences et al., 2005b) Ceratobatrachus guentheri Ceratobatrachidae RAG1 DQ019496 (van der Meijden et al., 2005) COI AB127977 (Sano et al., 2004) Buergeria buergeri Rhacophoridae RAG1 AB612031 (Kurabayashi et al., 2011) COI JN133081 (Wollenberg et al., 2011) Boophis doulioti Mantellidae RAG1 AY571643 (van der Meijden et al., 2004)

FCUP 83 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Supplementary Table 3. Correlation matrix between the 12 environmental variables. Lat stands for latitude.

d_l d_l d_l d_l d_l d_l d_l d_l d_ TRI MNDWI NDWI1 Lat c01 c02 c04 c05 c06 c08 c11 c12 guel

TRI 1.00 d_lc01 -0.16 1.00 d_lc02 -0.18 0.76 1.00 d_lc04 0.00 0.182 0.08 1.00 d_lc05 -0.02 -0.03 -0.02 0.424 1.00 d_lc06 0.08 -0.24 -0.29 0.47 0.66 1.00 d_lc08 -0.04 0.25 0.18 0.67 0.17 0.32 1.00 d_lc11 0.083 -0.15 -0.22 0.52 0.44 0.76 0.49 1.00 d_lc12 -0.07 0.31 0.21 0.26 -0.15 -0.05 0.53 0.23 1.00 d_guel -0.09 0.12 0.08 0.22 0.20 0.32 0.36 0.46 0.50 1.00

MNDWI 0.02 0.07 -0.08 0.37 -0.17 0.13 0.50 0.27 0.52 0.25 1.00

NDWI1 0.06 -0.11 -0.18 0.14 -0.01 0.19 0.20 0.19 0.06 0.11 0.60 1.00

Lat -0.06 0.30 0.19 0.25 -0.19 -0.09 0.51 0.19 0.99 0.44 0.54 0.05 1.00

FCUP 84 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Supplementary Figure 2. Spatial variation of the environmental variables in the study area

FCUP 85 Conservation Biogeography of Amphibians in Mauritania: integrating DNA barcoding and spatial analyses for the identification of cryptic diversity and phylogenetic units

Supplementary Table 4. Support values for clusters obtained from GMYC single-threshold method applied to COI and RAG1 gene trees. Only values > 0.50 are shown. * Taxa represented by singletons for one or both molecular markers.

Support values GMYC Clusters COI RAG1

Amnirana galamensis 1.00 -

Hoplobatrachus occipitalis - 1.00 Hoplobatrachus occipitalis 1 1.00 - Hoplobatrachus occipitalis 2 1.00 -

Kassina fusca* 1.00 - Kassina senegalensis 1.00 1.00 Leptopelis bufonides* - - Phrynobatrachus sp. - 0.56

Phrynobatrachus sp. 1 1.00 - Phrynobatrachus sp. 2 1.00 - Ptychadena bibroni 1.00 1.00 Ptychadena schillukorum 1.00 1.00

Ptychadena trinodis 1.00 0.90 Pyxicephalus edulis* 1.00 - Sclerophrys pentoni 1.00 1.00 Sclerophrys regularis* 1.00 -

Sclerophrys xeros 1.00 1.00 Tomopterna milletihorsini 1.00 1.00