Zoosyst. Evol. 98 (2) 2022, 435-453 | DO! 10.3897/zse.98.86299 oe a D> PENSOFT. ijt A NATURKUNDE BERLIN Rock island melody remastered: two new species 1n the Afroedura bogerti Loveridge, 1944 group from Angola and Namibia Werner Conradie!*, Andreas Schmitz, Javier Lob6n-Rovira*®®, Francois S. Becker”®, Pedro Vaz Pinto*®?:°, Morgan L. Hauptfleisch*? Port Elizabeth Museum (Bayworld), P.O. Box 13147, Humewood 6013, South Africa 2 Department of Nature Conservation Management, Natural Resource Science and Management Cluster, Faculty of Science, George Campus, Nelson Mandela University, George, South Africa 3 Natural History Museum of Geneva, Route de Malagnou 1, C.P. 6434, 1211 Geneva 6, Switzerland CIBIO, Centro de Investigacao em Biodiversidade e Recursos Genéticos, INBIO Laboratorio Associado, Campus de Vairdo, Universidade do Porto, 4485-661 Vairdo, Portugal 5 Departamento de Biologia, Faculdade de Ciéncias, Universidade do Porto, 4099-002 Porto, Portugal BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairdo, 4485-661 Vairdo, Portugal de Ciéncias da Educagdao da Huila (ISCED-Huila), Rua Sarmento Rodrigues, Lubango, Angola 7 National Museum of Namibia, Windhoek, Namibia 8 School of Animal, Plant and Environmental Sciences, University of the Witwatersrand, Private Bag 3, Wits 2050, Johannesburg, South Africa 9 Fundacgdo Kissama, Luanda, Angola 10 TwinLab CIBIOISCED — Instituto de Ciéncias da Educagdo da Huila, Rua Sarmento Rodrigues 2, C.P. 230, Lubango, Angola 11 Biodiversity Research Centre, Namibia University of Science and Technology, Windhoek, Namibia https://zoobank. org/6EA087B2-3245-455F-AD10-15016E8417D3 Corresponding author: Werner Conradie (werner@bayworld.co.za) Academic editor: Johannes Penner # Received 8 May 2022 # Accepted 11 October 2022 Published 21 November 2022 Abstract Newly collected material from northern Namibia’s Otjihipa Mountains and west-central Angola allowed us to revisit the Afroedura bogerti Loveridge, 1944 group. The employment of additional gene markers, including nuclear markers, allowed us to identify two new species in the group and infer species boundaries and potential speciation events in Afroedura from southwestern Africa. The new Namibian material is recovered as a sister species to A. donveae, from which it differs mostly by the colour of the iris (copper versus black) and dorsal colouration. Material from the first elevational gradient of the escarpment in Benguela Province, Angola was found to be more closely related to A. bogerti than A. wulfhaackei. The differences between these two species are more subtle, although the new species exhibits higher mid-body scale rows (79.5 versus 74.8), different dorsal colouration and supranasal scales always in contact (versus 57% in contact). Key Words endemism, flat geckos, Gekkonidae, Reptilia, speciation Resumo O material recém-colectado nas montanhas Otjihipa do norte da Namibia e no centro-oeste de Angola permitiu-nos revisitar 0 grupo Afro- edura bogerti Loveridge, 1944. O emprego de marcadores genéticos adicionais, incluindo marcadores nucleares, permitiu-nos identificar duas novas espécies no grupo e inferir limites para separar as espécies e potenciais eventos de especia¢gao nos Afroedura do sudoeste Africano. O novo material da Namibia é recuperado como espécie mais proxima de A. donveae, do qual difere sobretudo pela cor da iris (acobreada ao invés de negra) e pela coloragaéo dorsal. Ao passo que o material obtido no primeiro gradiente topografico da escarpa na Copyright Werner Conradie et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. 436 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa provincia de Benguela, Angola, revelou ser mais relacionado com A. bogerti do que com A. wulfhaackei. As diferengas entre estas duas espécies so mais subtis, muito embora as novas espécies axibam maior numero de escamas a meio do corpo (79.5 em vez de 74.8), diferente coloracao dorsal e escamas supranasais sempre em contacto (em vez de apenas em contacto em 57%). Palavras-chave endemismo, especiagaéo, Gekkonidae, Osga-achatada, Reptilia Introduction African flat geckos Afroedura Loveridge, 1944 currently comprise 32 species (Uetz et al. 2022), occurring from west- ern Angola southwards to South Africa and along the eastern escarpment northwards to central Mozambique (Jacobsen et al. 2014; Branch et al. 2017, 2021). In recent years there has been a considerable increase in the number of new Afroedu- ra species described from Angola (Branch et al. 2021), Mo- zambique (Branch et al. 2017) and the northern provinces of South Africa (Jacobsen et al. 2014). Numerous additional candidate new species have been identified and await formal description (Makhubo et al. 2015; Busschau et al. 2019). Until recently, only three species of Afroedura had been recorded from Namibia and Angola, namely A. africana (Boulenger, 1888), A. tirasensis Haacke, 1965 and A. bogerti Loveridge, 1944. However, a recent revision of Angolan material of the A. bogerti-group (Branch et al. 2021), revealed cryptic diversity and increased the total number to five species for the country: A. bogerti Loveridge, 1944, A. wulfhaackei Branch, Schmitz, Lobon-Rovira, Baptista, Antonio, Conradie, 2021, A. donveae Branch, Schmitz, Lobon-Rovira, Baptista, Antonio, Conradie, 2021, A. preadicta Branch, Schmitz, Lobon-Rovira, Baptista, Antonio, Conradie, 2021, and A. vazpintorum Branch, Schmitz, Lobon-Rovira, Baptista, Antonio, Conradie, 2021. These Angolan species were grouped in two genetic clusters: south-western coastal low-lying (A. donveae, A. vazpintorum [with an isolated southern escarpment population], A. praedicta) and inland central high-lying (A. wulfhaackei, A. bogerti). All species are currently regarded as Angolan endemics. In both wide-ranging species, A. vazpintorum and A. wulfhaackei, genetic sub-structuring was documented (Branch et al. 2021). In A. vazpintorum two subclades were identified, which differed by an average 4.1% for the /6S mitochondrial marker. One subclade was found to be widely distributed across much of the semi-arid ‘Pro- Namib’ coastal zone north of Mocamedes, while the other subclade was restricted to the Humpata plateau. What complicated the matter further was that a sample collected syntopic with other Humpata samples was imbedded in the coastal clade and no morphological differences were observed. On the other hand, in A. wulfhaackei four subclades were identified that differed genetically by a similar margin from the previous species (3.3—4.2% 16S), with once again no clear morphological differences. zse.pensoft.net Follow-up work with more material and gene coverage was recommended to resolve this issue. Unfortunately, the status of historical records of A. bogerti from northern Namibia (see Branch 1998; Griffin, 2002) could not be assessed together with the An- golan material given the lack of fresh material suitable for inclusion in a phylogenetic analysis. During recent expe- ditions new material was collected in northern Namibia and Angola, giving the opportunity to revisit this group in more detail. In this study we build on the previous work of Branch et al. (2017, 2021) by adding additional sam- ples and gene markers to infer species boundaries and potential speciation events within the A. bogerti-group in south-western Africa. This allowed us to look a bit deeper into the genetic sub-structuring documented in the two widespread species, A. vazpintorum and A. wulfhaackei. Materials and methods Sampling The material used in Branch et al. (2021) was supplement- ed with sequences from additional genes, as well as with newly-collected samples from Otjihipa Mountains, north- ern Namibia (n = 2) and Angola (n = 17) (see Table 1, Fig. 1). Specimens were collected and processed following the protocols described in Branch et al. (2021). Newly collect- ed voucher specimens were deposited in the natural history collections of the National Museum of Namibia, Windhoek (NMNW) and Fundacao Kissama (FKH), Luanda, Angola. DNA extraction, amplification and sequencing Total genomic DNA for the new samples was extracted from tissue samples using the E.Z.N.A. Tissue DNA Kit (VWR/Omega bio-tek) and the Qiagen DNeasy Tissue Kit, following the manufacturer’s protocols. The following genes were amplified: two partial mitochondrial ribosomal genes (ribosomal ribonucleic acid [/2S and /6S]), two partial mitochondrial genes (cytochrome b [Cyt-b] and NADH-dehydrogenase subunit 2 [VD2]) and two partial nuclear gene (oocyte maturation factor [c-mos]| and recombination activating protein [RAG/]). Respective primers and reference to PCR protocols are given in Table 2. PCR products Zoosyst. Evol. 98 (2) 2022, 435-453 12°S 14°S 16°S 12°E 14°E (Mountains / Inselbergs [_] Plateu >1.600 ma.s.1 CUNENE ANGOLA @ Afroedura bogerti @ Afroedura wulfhaackei © Afroedura pundomoniana sp. nov. @ Afroedura vazpintorum © Afroedura praedicta @ Afrcedura donveae 437 © Afroedura otjihipa sp. nov. 18°S (¢) Molecular data OSHIKOTO Figure 1. All occurrence records (coloured circles) and predicted distribution for the Afroedura bogerti group from southwestern A fri- ca. No predicted distribution could be created for A. othjihipa sp. nov. Angolan provinces and Namibian regions are labelled according- ly. Stars represent the respective type localities and black dots with white borders represent localities used in the phylogenetic analysis. Table 1. Afroedura specimens with generalised localities and GenBank accession numbers of vouchers used in this study. * Addi- tional samples added during this study. ANG/AG — William R. Branch field numbers; CHL — Colec¢ao Herpetologica do Lubango (CHL), Instituto Superior de Ciéncias de Educa¢éo da Huila ISCED-Huila), Angola; FKH — Fundacao Kissama, Luanda, Angola; JLRZC — Javier Lobon-Rovira field numbers; KTH — Krystal Tolley field numbers; NB — Ninda Baptista field numbers; NUNW — National Museum of Namibia, Windhoek; P — Pedro Vaz Pinto field numbers; PEM — Port Elizabeth Museum, South Africa; WC — Werner Conradie field numbers; ZMB — Museum fiir Naturkunde, Berlin, Germany. SC — subclade. Species Locality Sample Number Museum 16S 12S c-mos RAG1 Cyt-b ND2 Number A. praedicta Serra da Neve, Angola NB 853 ZMB 91607 = =MW354010 OP653587 OP686766 OP686640 OP686714 OP686613 A. praedicta Serra da Neve, Angola NB 854. CHL 854 MW354011 OP653588 OP686767 OP686641 OP686715 OP686614 A. praedicta Serra da Neve, Angola NB 855 CHL 855 MW354012 OP653589 OP686768 OP686642 NA OP686615 A. otjihipa sp. nov. Otjihipa, Namibia SMR 11182* NMNWR11253 + O0P653544 NA OP686789 OP686638 NA OP686623 A. otjihipa sp. nov. Otjihipa, Namibia SMR 11183* NMNWR11254 O0P653545 NA OP686790 OP686639 NA OP686624 A. donveae Omauha Lodge, Angola E259.17/KTHO9- PEMR17936 LM993776 OP653553 OP686732 OP686633 NA OP686594. 196 A. donveae Omauha Lodge, Angola E259.18/KTHO9- PEMRI17937 LM993777 OP653554 OP686733 OP686634 NA OP686595 197 A. donveae Omauha Lodge, Angola P9-284 Na MW354008 OP653602 OP686780 OP686635 NA OP686621 A. donveae Omauha Lodge, Angola P9-285 Na MW354009 OP653603 OP686781 OP686636 NA OP686622 A. vazpintorum SC1 52 km north on tar road on road to =E259.12/ANG 311 PEMR21596 MF565461 OP653548 OP686727 OP686644 OP686685 NA Lucira, Angola A. vazpintorum SC1_ 1 km east of Farm Mucungo, Angola E259.13/AG138 PEMR24115 MF565463 OP653549 OP686728 OP686645 OP686686 OP686590 A. vazpintorum SC1__1 km east of Farm Mucungo, Angola E259.14/AG137 PEMR24114 MF565460 OP653550 OP686729 OP686646 OP686687 OP686591 A. vazpintorum SC1_1 km east of Farm Mucungo, Angola E259.15/AG141 PEMR24118 MF565462 OP653551 OP686730 OP686647 OP686688 OP686592 A. vazpintorum SC1 10.4 km south of Rio Mucungo on tar E260.12/samp39 Na MF565459 OP653560 OP686739 OP686650 OP686693 OP686598 road to Bentiaba, Angola A. vazpintorum SC1_ 10.4 km south of Rio Mucungo on tar E260.13/samp57_ PEMR24203 MF565458 OP653561 OP686740 OP686651 OP686694 OP686599 road to Bentiaba, Angola A. vazpintorum SC1_ 10.4 km south of Rio Mucungo on tar E260.14/samp58 PEMR24204 MF565457 OP653562 OP686741 OP686652 OP686695 OP686600 road to Bentiaba, Angola A. vazpintorum SC1 20 km south Bentiaba, Angola E260.15/samp62 PEMR24219 MF565456 OP653563 OP686742 OP686653 NA OP686601 A. vazpintorum SC1 approx. 18 km E Lucira, Angola NB 834 CHL 834 MW354019 OP653585 OP686764 OP686658 OP686712 OP686611 A. vazpintorum SC1 approx. 18 km E Lucira, Angola NB 835 CHL 835 MW354020 OP653586 OP686765 OP686659 OP686713 OP686612 A. vazpintorum SC1 Mariquita, Angola P9-154 Na MW354018 OP653601 OP686779 OP686666 NA NA A. vazpintorum SC1 50 km east Namibe on main tar road E259.16/ANG 289 PEMR21595 MF565454 OP653552 OP686731 OP686648 NA OP686593 to Leba, Angola A. vazpintorum SC1_—_ Bimbe, Estacao Zootecnica, Angola NB 743 CHL 743 MW354017 OP653578 OP686757 OP686654 NA OP686607 A. vazpintorum SC1 Tundavala, Angola P0-103* Na OP653527 OP653590 OP686769 OP686660 NA OP686616 A. vazpintorum SC1 Tundavala, Angola PO-104* OP653528 OP653591 NA OP686661 NA OP686617 FKH-0518 A. vazpintorum SC1 Meva Beach, Angola E259.9/samp30 PEM R22488 MF565455 OP653556 OP686735 OP686649 NA NA zse.pensoft.net 438 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa Species Locality Sample Number Museum 16S 12S c-mos RAG1 Cyt-b ND2 Number A. vazpintorum SC1 Carivo, Angola P8-19 Na MW354015 OP653598 OP686776 OP686664 NA OP686620 A. vazpintorum SC1 Carivo, Angola P8-20 Na MW354016 OP653599 OP686777 OP686665 NA NA A. vazpintorum SC2___ Bimbe, Estacao Zootecnica, Angola NB 744* CHL 744 OP653529 OP653579 OP686758 OP686655 OP686707 OP686608 A. vazpintorum SC2 Bimbe, Estacao Zootecnica, Angola NB 745 CHL 745 MW354013 OP653580 OP686759 OP686656 OP686708 OP686609 A. vazpintorum SC2_—_ Bimbe, Estacao Zootecnica, Angola NB 746 CHL 746 MW354014 0P653581 OP686760 OP686657 OP686709 OP686610 A. vazpintorum SC2 Tchivinguiro, Angola P0-97* FKHO514 OP653530 OP653596 OP686774 OP686662 OP686716 OP686618 A. vazpintorum SC2 Tchivinguiro, Angola P0-98* FKHO515 OP653531 OP653597 OP686775 OP686663 OP686717 OP686619 A. pundomontana Alto Pundo — Bocoio, Angola WC-6524* PEM R24743 OP653543 NA OP686791 OP686643 NA OP686625 Sp. nov. A. pundomontana Alto Pundo — Bocoio, Angola P1-280* FKHO688 OP653532 OP653607 OP686785 NA OP686722 NA Sp. nov. A. pundomontana Alto Pundo — Bocoio, Angola P1-281* FKHO689 OP653533 OP653608 OP686786 NA OP686723 NA Sp. nov. A. pundomontana Alto Pundo — Bocoio, Angola P1-282* FKHO690 OP653534 OP653609 OP686787 NA OP686724 NA Sp. nov. A. bogerti Farm Namba , Angola E260.1/samp23 PEMR24184 MF565467 OP653557 OP686736 OP686626 OP686690 OP686597 A. bogerti Farm Namba, Angola E260.2/samp24 PEMR24185 MF565468 OP653568 OP686747 OP686627 OP686700 OP686602 A. bogerti Farm Namba, Angola E260.3/samp25 PEMR24186 MF565466 OP653569 OP686748 OP686628 OP686701 OP686603 A. bogerti 400 m north of Missionde Namba =-E260.4/samp27 PEM R24187 MF565465 OP653570 OP686749 OP686629 NA OP686604 grounds, Angola A. bogerti 400 m north of Mission de Namba -E260.5/samp28 Na MF565464 OP653571 OP686750 OP686630 OP686702 OP686605 grounds, Angola A. bogerti Namba, Angola JLRZCO015 Na MW354021 P653576 OP686755 OP686631 NA NA A. bogerti Namba, Angola JLRZCO016 Na MW354022 OP653577 OP686756 OP686632 NA OP686606 A. bogerti Namba, Angola P1-286* Na OP653535 OP653610 OP686788 NA NA NA A. wulfhaackei SC1 Farm VictoriaVerdun, 2kmS of Mt. E260.6/samp31 Na MF565470 OP653572 OP686751 OP686675 OP686/03 NA Sandula, Angola A. wulfhaackei SC1 Farm VictoriaVerdun, 2kmS of Mt. E260.7/samp32 Na MF565469 OP653573 OP686752 OP686676 OP6867/04 NA Sandula, Angola A. wulfhaackei SC1 Farm VictoriaVerdun, 2kmS of Mt. E£260.8/samp33 PEM R24191 MF565471 OP653574 OP686753 OP686677 OP686705 NA Sandula, Angola A. wulfhaackei SC1 Farm VictoriaVerdun, 2 km S of Mt. E260.9/samp34 PEM R24192 MF565469 OP653575 OP686/754 OP686678 OP686706 NA Sandula, Angola A. wulfhaackei SC1 Sandula, Angola P9-141 MW354023 OP653600 OP686778 OP686682 OP686718 NA A. wulfhaackei SC1 Moco - Kapa Kuito, Angola P0-49* FKH-0472 OP653536 OP653592 OP686770 NA NA NA A. wulfhaackei SC2 5 km southwest of Lepi, Angola E260.11/samp37 PEMR24201 MF565472 OP653559 OP686738 OP686670 OP686692 NA A. wulfhaackei SC2 Lepi, Angola P1-162* FKH-0593 OP653537 OP653604 OP686782 NA OP686719 NA A. wulfhaackei SC2 Lepi, Angola P1-163* FKH-0594 OP653538 OP653605 OP686783 NA OP686720 NA A. wulfhaackei SC2 Lepi, Angola P1-164* FKH-0595 0653539 +OP653606 OP686784 NA OP686721 NA A. wulfhaackei SC3. Candumbo Rocks Memorial, Angola E259.10/VC-4037 PEMR22490 MF565474 OP653546 OP686725 OP686667 OP686683 NA A. wulfhaackei SC3. Candumbo Rocks Memorial, Angola E259.11AVC-4038 PEMR22491 MF565475 OP653547 OP686726 OP686668 OP686684 NA A. wulfhaackei SC3. Candumbo Rocks Memorial, Angola £260.10/samp35 PEMR24200 MF565473 OP653558 OP686737 OP686669 OP686691 NA A. wulfhaackei SC4. Maka-Mombolo, north-east of E260.16/samp70 PEMR24236 MF565476 OP653564 OP686743 OP686671 OP686696 NA Balombo, Angola A. wulfhaackei SC4_ 5 km west of MakaMombolo, Angola E260.17/samp71 PEMR24232 MF565477 OP653565 OP686744 OP686672 OP686697 NA A. wulfhaackei SC4_ 5 km west of Maka-Mombolo, Angola E260.18/samp72 PEMR24233 MF565478 OP653566 OP686745 OP686673 OP686698 NA A. wulfhaackei SC4_ 5 km west of MakaMombolo, Angola E260.19/samp73 PEMR24234 MF565479 OP653567 OP686746 OP686674 OP686699 NA A. wulfhaackei SC4. Morro do Moco, camp near NB 817 CHL 817 MW354024 OP653582 OP686761 OP686679 OP686710 NA Canjonde, Angola A. wulfhaackei SC4 Morro do Moco, camp near NB 818 CHL 818 MW354025 OP653583 OP686762 OP686680 OP686711 NA Canjonde, Angola A. wulfhaackei SC4 Morro do Moco, camp near NB 819 CHL 819 MW354026 OP653584 OP686763 OP686681 NA NA Canjonde, Angola A. wulfhaackei SC4 Moco - Kapa Kuito, Angola PO-50* FKH-0473 OP653540 OP653593 OP686771 NA NA NA A. wulfhaackei SC4 Moco - Kapa Kuito, Angola P0-51* FKH-0474 OP653541 OP653594 OP686772 NA NA NA A. wulfhaackei SC4 Moco - Kapa Kuito, Angola P0-52* FKH-0475 OP653542 OP653595 OP686773 NA NA NA A. loveridgei Near Moatize, Tete Province, E1123 MF565446 OP653555 OP686734 OP686637 OP686689 OP686596 Mozambique A. karroica Eastern Cape Province, 41km SE PEM FN1112 LM993744 NA JQ945523 KMO073485 NA JX041302 (outgroup) Murraysburg, South Africa were sequenced at Macrogen Corp. (Amsterdam, Netherlands). For quality assurance, both directions of the amplified PCR products were sequenced. For the molecular comparisons, newly-sequenced vouchers (n = 19) were used as well as extending previously used samples (n = 48; used for the previously published /6S sequences [Branch et al. 2017, 2021]) for five additional genes. The final dataset comprised 68 ingroup samples from different localities covering the entire distribution of Angolan and Namibian Afroedura ‘bogerti’ populations and Afroedura karroica as outgroup. Locality data and zse.pensoft.net respective GenBank (https://www.ncbi.nlm.nih.gov/ genbank/; Benson et al. 2013) numbers for each sample are listed in Table 1. Phylogenetic analysis Sequences were checked for reliability using the orig- inal chromatograph data in the program BioEdit v.7.2.5 (Hall 1999), aligned using ClustalX v.1.6 (Thompson et al. 1997), with each alignment then checked manually for Zoosyst. Evol. 98 (2) 2022, 435-453 439 Table 2. The primers and final sequence lengths for the two nuclear genes and four mitochondrial genes used in this study. Gene Sequence length (bp) Primer Sequence (5’ -> 3’) Reference 16S 594 16Sa CGC CTG TTT ATC AAA AAC AT Palumbi et al. 1991 16Sb CCG GTC TGA ACT CAG ATC ACG T Palumbi et al. 1991 C-mos 399 CmosG73 GCG GTA AAG CAG GTG AAG AAA Wiens et al. 2010 CmosG74 TGA GCA TCC AAA GTC TCC AAT C Wiens et al. 2010 Cyt-b 1008 CytBL14910 GAC CTG TGA TMT GAA AAC CAY CGT TGT Burbrink et al. 2000 CytBH16064 CTT TGG TTT ACA AGA ACA ATG CTT TA Burbrink et al. 2000 RAGI 639 RAG1F700 GGA GAC ATG GAC ACA ATC CAT CCT AC Bauer et al. 2007 RAG1R700 TTT GTA CTG AGA TGG ATC TTT TTG CA Bauer et al. 2007 ND2 1014 ND2L4437R AAG CTT TCG GGC CCA TAC C Stanley et al. 2011 ND2H5540F TTT AGG GCT TTG AAG GC Bauer et al. 2010 ND2R102 CAG CCT AGG TGG GCG ATT G Bauer et al. 2010 12S 437 12sf700 AAA CTG GGA TTA GAT ACC CCA CTA T Stanley et al. 2011 12sr600 GAG GGT GAC GGC GGT GTG T Stanley et al. 2011 errors. Protein-coding partitions of mitochondrial (Cyr-b, ND2) and nuclear genes (c-mos, RAG), were translated to amino acids with the program Geneious Prime v.2021.2.2 (https://www.geneious.com) to set codon positions and confirm absence of stop codons. The final alignment of all Six genes, including nuclear and mitochondrial loci, con- sisted of 4091 base pairs. Sequence lengths are detailed in Table 2. Two techniques for phylogenetic estimation were applied: Bayesian Inference (BI) using MrBayes v.3.26 (Ronquist et al. 2012) and Maximum Likelihood (ML) using IQ-Tree v.2.1.2 (Nguyen et al. 2015; Minh et al. 2020) as implemented in PhyloSuit v.1.2.2 (Zhang et al. 2020) using the ultrafast bootstrap option. Both BI and ML used the recognised partition schemes identified with PartitionFinder 2 (Lanfear et al. 2016) as implemented in PhyloSuit v.1.2.2. Partitioning schemes and substitution models are provided in Table 3. Table 3. Partition schemes and models of substitution for the Bayesian (PP) and maximum-likelihood (ML) calculations. Substitution model Included partitions 1 GIR+1+G 12S, 16S, ND2 (P1), Cyt-b (P1) 2 HKY +1 RAGI (P1, P2, P3), c-mos (P1, P2. P3) 3 HKY+1+G ND2 (P2), Cyt-b (P2) 4 TrN+G ND2 (P3), Cyt-b (P3) Support values for the two phylogenetic approaches were calculated. Bootstrap analyses (BS) with 50000 ul- trafast bootstraps evaluated the relative branch support in the ML analysis. As we used the ultrafast bootstrap option, only clades with support > 95% were consid- ered strongly supported. Bayesian analyses were run under partitioned schemes for 50 million generations with four chains sampled every 1000 generations, with a burn-in of 10000 trees. Clades with posterior proba- bilities (PP) => 0.95 were considered strongly supported. Convergence and mixing of the parameters for each run of the Bayes analysis was checked with the Effective Samples Size (ESS) using Tracer v.1.72 (Rambaut et al. 2018). The final trees were visualised with FigTree v.1.4.4 (Rambaut 2014; http://tree.bio.ed.ac.uk/soft- ware/figtree/). Uncorrected p-distances between opera- tional taxonomical units (OTUs) and within OTUs were calculated with MEGA X v.11.0.8 (Kumar et al. 2018) for the partial /6S gene as these values were often used to prove distinctness at the species level in Afroedura (Branch et al. 2017, 2021). Morphology We examined newly collected material in the collections of the National Museum of Namibia (NMNW), Windhoek, Namibia and Funda¢ao Kissama (FKH), Luanda, Angola. The following characters were assessed: 1) presence or ab- sence of internasal granules between the supranasal scales; 2) number of postmental scales; 3) number of scales in contact in a straight line between the anterior corners of eyes across the crown of the head; 4) number of scales be- tween upper edge of earhole and rear margin of eye count- ed along the shortest distance between them; 5) number of scales between nostril and front edge of orbit, including postnasal; 6) number of enlarged supralabials to the angle of the jaw at midorbital position; 7) number of enlarged infralabials to the angle of the jaw at midorbital position; 8) number of midbody scale rows (MSR), counted at the widest part of the trunk; 9) number of scale rows on dorsal surface per tail whorl (counted 3-6 verticils posterior to the cloaca); 10) number of scales rows on ventral surface per tail whorl (counted 3-6 verticils posterior to the cloaca); and 11) number of precloacal pores in males. In Branch et al. (2021) it was erroneously recorded that the nostril is in direct contact with the rostral in all species in the A. bogerti-group. The nostril is actually pierced between the first supralabial and three nasal scales, and is narrowly excluded from the rostral. We here rectify this mistake by updating the diagnosis of the group (see Taxonomy). The following measurements were taken in millime- ters (mm) using a digital calliper (accuracy of 0.01 mm) with the aid of a Nikon SMZ1270 dissecting microscope: 1) snout-vent length (SVL — from the tip of the snout to the cloaca with the gecko flattened on its back), 2) tail length (TL, only original tails were measured); 3) head length (HL — tip of snout to retro-articular process of jaw); 4) head width (HW — widest point of head approx- imately at the level of eyes); 5) snout length (SL — tip of snout to front of orbit); 6) eye diameter (ED — measured in horizontal orientation); 7) ear to eye length (EE — top edge of earhole to back of eye); 8) ear opening (EO — zse.pensoft.net 440 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa 100 100 A. karroica (outgroup) £259.19 A. loveridgei WC-4037 __F wes 4. wulfhaackei SC3 100 E260.10 A, wulfhaackei SC4 A. wulfhaackei SC2 A. wulfhaackei SC1 0.03 substitutions/site Figure 2. Phylogenetic tree topology based on the combined mitochondrial (/2S, 16S, Cyt-b, ND2) and nuclear (c-mos, RAG1) genes, using Afroedura karroica as outgroup. Support values for Bayesian posterior probabilities (above nodes) and maximum likelihood bootstraps (below nodes) are indicated in the tree (shown values: ML: > 70% / PP: > 0.75). greatest length); and 9) internostril distance (IN — shortest distance between nostrils). All head measurements were done on the right side of the head. Predicted species distribution mapping Due to the low number of occurrence records we pro- duce predicted distribution maps for each species in the A. bogerti-group by performing species distribution mod- els (SDM) based on suitable bioclimatic areas using Max- ent (Yang et al. 2013). The model included a buffer of 2 degrees (~250 km) from the most peripheral observations of A. bogerti-group species (Branch et al. 2021). Nineteen bioclimatic variables were obtained from the WorldClim data set (Fick and Hiymans 2017; http:/Awww.worldclim. org/) at a spatial resolution of 30 arc-second (~1 km/?). For those variables, we ran a correlation model to eliminate collinearity between variables in the sampled area and within sample points (Candau and Fleming 2005), and variables with correlation coefficient > 0.7 were select- ed in order to capture all the bioclimatic range over the distribution of the species (Enriquez-Urzelai et al. 2019; Branch et al. 2021). Therefore, the variables included for analysis were: mean diurnal temperature range (Biol); maximum temperature of the warmest month (Bio5), min- imum temperature of the coldest month (Bio6); annual precipitation (Biol2); precipitation seasonality (Bio15), and precipitation of wettest quarter (Biol6). Given the small sample size for some species, we ran a cross-val- idation model, which utilises all the samples except for leaving out one sample in each run (Bittencourt-Silva et al. 2016; Branch et al. 2021), and hinge features only zse.pensoft.net with the regularisation parameter set to 2.5 to produce smoother response curves and reduce overfitting (Briscoe et al. 2016). The final maps were generated selecting ar- eas with more than 90% of climate suitability for each species with the exception of “Afroedura sp. 2” for which the sample size was too low to run a SDM. Results Molecular analyses Molecular analyses concur with previously published results on the Angolan members of the genus Afroedura (Branch et al. 2017, 2021). The ESS values for the com- bined Bayes analysis were well above the recommended 200 threshold, thus indicating that the burn-in was suffi- cient and convergence was achieved. The addition of five new genes strengthened node support on all levels, with the two nuclear genes especially adding support for the deeper nodes form both the BI and ML trees. In addition to the previously identified five genetically supported clades of Angolan Afroedura, analysis of our additional material resulted in the recognition of two further well-supported clades, namely Afroedura sp. 1 from Bocoio, Angola and Afroedura sp. 2 from Otjihipa, Namibia (Fig. 2). In both new clades we observe only very low intraspecific dif- ferences, but comparatively high interspecific differenc- es, compared to both the two widespread species and the other major clades of Angolan Afroedura (Table 4), but this low intraspecific variation can also be due to sam- pling localities of the vouchers being either identical or very close to each other (Table 1). Individual analyses of Zoosyst. Evol. 98 (2) 2022, 435-453 441 Table 4. Summary of intra- and interclade uncorrected pairwise sequence divergences (%) for specimens of Afroedura clades compared to A. Joveridgei for 16S rRNA. Interclade distances below 5% (see the relevant discussion in the main text) are marked in bold. SC — subclades. 5 pa o zi} Intraclade = 5 g 3 distances 3 Rr) Ss s < a < < A. loveridgei - - A. bogerti 0.2 17.2 - A. donveae 0.1 19.6 10.7 - A. praedicta 0.7 17.5 8.8 6.5 - A. pundomontana sp. nov. 0.1 17.8 7.8 12.1 11.0 A. otjihipa sp. nov. 0.0 17.3 11.0 74 79 A. vazpintorum SC1 2.2 18.0 11.0 8.8 8.4 A. vazpintorum SC2 1.0 17.3 9.5 8.9 7.6 A. wulfhaackei SC1 0.1 14.9 6.4 10.0 9.6 A. wulfhaackei SC2 0.1 15.5 6.7 10.2 9.1 A. wulfhaackei SC3 0.2 13.3 7.2 10.4 9.6 A. wulfhaackei SC4 0.4 14.1 5.9 9.4 8.7 the four mitochondrial genes (not shown here) show that the recovered topology is identical for each gene, without any discrepancies in the recovered nodes. The addition of newly sequenced specimens from A. wulfhaackei subclade 2 (Afroedura sp. 6 sensu Branch et al. 2021) indicated the same general pattern as the other lineages, with some moderate genetic distances to the oth- er A. wulfhaackei subclades as well as significantly higher genetic distances to any other Angolan Afroedura species (see Table 4). As the four A. wulfhaackei subclades show much lower inter-taxon distances to one another (< 5% in the partial /6S gene; Table 4) than the average distance within recognised Angolan Afroedura species, this neither supports nor rejects the previous suggestion by Branch et al. (2021) to separate the identified candidate species around the central highlands (Afroedura sp. 5-7 sensu Branch et al. 2021) from the recently described A. wulf- haackei. The same 1s true for the previously discussed split of two identified subclades within the species A. vazpin- torum (see Branch et al. 2021): the three genetically-di- vergent Bimbe specimens (NB 744-6) formed a separate clade with two new vouchers from Tchivinguiro (PO-97— 8), while two new samples from Tundavala (PO-103-4) and one from Bimbe (NB 743) cluster with the lowland coastal subclade. Bimbe, Tchivinguiro and Tundavala are nearby sites situated in the southern highlands, on the edge of the Humpata plateau. Similar to the divergence within A. wulfhaackei subclades, the genetic divergence of the highland subclade from true A. vazpintorum is well below the 5% threshold identified here (Table 4). Nonetheless, since the average of the genetic differences for these can- didate species clades 1s significantly lower than between the currently recognised Afroedura species in Angola, they are herewith pooled into their closest known species (A. wulfhaackei and A vazpintorum, respectively) for the SDM analyses and morphological comparisons, pending further research to verify their taxonomic status. A. pundomontana Sp. nov. A. otjihipa sp. nov. A. vazpintorum SCl A. vazpintorum $c2 A. wulfhaackei SCl A. wulfhaackei $c2 A. wulfhaackei Sc3 A. wulfhaackei SC4 8.4 10.0 10.5 99 4.0 3.4 3.7 - Morphology Results for the morphological comparisons are sum- marised in Table 5, and are discussed in more detail in the species descriptions below. The new material from Bo- coio, Angola (Afroedura sp. 1) was similar to the central highland subgroup (A. wulfhaackei and A. bogerti) with regard to the lower numbers of scale rows on ventral and dorsal surface per tail verticil (4 and 5, respectively) and venter pigmented with fine black specks, but differed in that the supranasals were always in contact (similar to the south-western coastal sub-group comprising A. donveae and A. vazpintorum) versus separated by smaller granules (33% in A. bogerti and 57% in A. wulfhaackei) and higher number of precloacal pores (12 versus 8 in A. bogerti and 9.5 mean in A. wulfhaackei [although this might not be a true reflection as our male sample sizes of all species are limited]). The material from Otjihipa Mountains, northern Namibia (Afroedura sp. 2), conforms to the south-western coastal sub-group in the number of scale rows on ventral and dorsal surface per tail verticil (5 and 6, respectively), immaculate venter, bright dorsal colouration and boldly black-barred tail. It differs only in some minor scalation features and in that the iris is brown or copper-coloured (versus black in A. donveae). Species descriptions Both genetics and morphology, as well as geographical separation, suggest that the Afroedura sp. 1 from Boco- io, Angola and Afroedura sp. 2 from Otjihipa, Namibia populations should both be regarded as new species. We apply the general lineage-based species concept, treating all populations that represent independent historical lin- eages supported by multiple different lines of evidence as separate species (de Queiroz 1998). zse.pensoft.net 442 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa Table 5. Summary of morphological data for Afroedura bogerti, A. wulfhaackei (including members of the morphologically-indis- tinguishable subclades), A. donveae, A. vazpintorum (including isolated escarpment population), A. praedicta, A. pundomontana sp. nov. and A. ofjihipa sp. nov. Values are given as a range with mean in parenthesis for scalation and mean + standard deviation for meristic ratios. M = male, F = female, n = sample size. Character A. bogerti A. wulfhaackei A. donveae A. vazpintorum — A. praedicta A. pundomontana A. otjihipa sp. nov. sp. nov. (n = 9) (n = 35) (n= 17) (n = 48) (n = 5) (n = 7) (n = 2) Snout vent length (maximum) M50 mm M60 mm M 59.6 mm M58 mm M52 mm M58 mm M60 mm F 54 mm F 59 mm F 65 mm F 59mm F 51 mm F 58 mm F 58 mm Head Length/Head Width 1.3 + 0.09 1.4+0.14 1.4 + 0.08 1.3 + 0.13 1.3+0.14 1.3+0.14 1.1+0.12 Snout Length/Eye Distance 1.6 + 0.34 2.0 + 0.20 2.0 + 0.19 1.8 + 0.29 1.7+0.19 2.0 + 0.96 1.7+0.12 Snout Length/Eye-Ear Distance 1.2 + 0.07 1.2 + 0.14 1.3 + 0.30 1.2 +0.17 1.1 + 0.09 1.2 + 0.30 1.1 + 0.06 Precloacal pores (males only) 8(n=1) 9-11 (9.5) (n= 12) 11-12 (11.5) (n= 4) 9-11 (10.2) (n = 12) 8 (8.0) (n = 3) 12 (n= 1) 12 (n= 1) Ventral rows per tail verticil 4 (4.0) 4 (4.0) 5-6 (5.5) 5-7 (5.0) 4 (4.0) 4-5 (4.4) 5 Dorsal rows per tail verticil 5 (5.0) 5-6 (5.0) 6-7 (6.6) 6-7 (6.1) 5 (5.0) 5-6 (5.6) 6 Scales below 4" toe 6-9 (6.9) 6-9 (7.3) 6-8 (7.7) 6-10 (8.0) 9-11 (9.6) 7-9 (7.7) 8 Mid-body scale rows 69-77 (73.5) 73-88 (79.5) 64-78 (72.8) 73-86 (80.3) 73-78 (74.8) 78-82 (79.5) 65-67 Scales between eyes 11-14 (12.4) 11-16 (13.7) 11-14 (11.0) 11-15 (13.1) 12-15 (13.5) 13-15 (13.9) 14 Scales: nostril to eye 8-12 (9.9) 7-10 (8.3) 8-11 (9.3) 7-11 (9.1) 9-10 (10.2) 10-13 (10.9) 10-11 Scales: ear to eye 14-16 (15.4) 12-18 (15.90) 11-14 (11.9) 13-17 (15.6) 13-16 (14.8) 16-19 (16.9) 12-13 Supranasals in contact 33% 57% 100% 100% 100% 100% 100% Supralabials 8-10 (8.4) 7-9 (8.2) 8-10 (9.0) 8-10 (8.8) 8-10 (9.2) 8-9 (8.7) 8-9 Infralabials 8-9 (8.3) 8-9 (8.3) 8-11 (9.3) 8-9 (9.1) 8-9 (8.5) 9 (9.0) 8-9 Afroedura pundomontana sp. nov. https://zoobank.org/556B8212-21E8-494A-BF1E-7B3021C53BA9 Bocoio Flat Gecko (English) Osga-achatada do Bocoio (Portuguese) Figs 3A-B, 4, 5C—D, Tables 5, 6 Note. According to Branch et al. (2021), historical ma- terial from near Bocoio in Benguela Province, Angola clustered morphologically with A. wulfhaackei. Howev- er, due to the occurrence at lower elevations and being isolated from other known populations of Afroedura it was suggested that the status of this population required further investigation (Branch et al. 2021). Newly-collect- ed material allowed for its re-assessment within a wider phylogenetic framework, and it was determined that it represented a novel lineage, related to A. bogerti and not A. wulfhaackei, as initially hypothesised. It is therefore described below as a new species. Synonym. Afroedura bogerti — Branch et al. 2017: 162; Marques et al. 2018: 178 (in part). Afroedura wulfhaackei— Branch et al. 2021: 66 (in part). Holotype. PEM R24743, adult female, collected at Morro do Pundo, about 25 km west of Bocoio (-12.44389, 13.92250; 946 m a.s.l.), Benguela Province, Angola by Pedro Vaz Pinto on 6 June 2018. Paratypes. (six specimens). *TM 46587-8, TM 465890, adult females, collected 30 km W of Sousa Lara [= Bocoio] (approx. -12.40689, 13.90400; 670 m a.s.1.), Benguela Province, Angola by Wulf Haacke on 28 May 1974; *TM 46589, adult male, collected 30 km W of Sousa Lara [= Bocoio] (approx. -12.40689, 13.90400; 670 m a.s.l.), Benguela Province, Angola, by Wulf Haacke on 28 May 1974; FKH 0688, FKH 0689, adult females, collected from Alto Pundo — Bocoio (-12.44367, 13.92072, 920 ma.s.1.), Benguela Province, Angola by Pedro Vaz Pinto and Afonso Vaz Pinto on 2 September 2021. *Note the locality data presented as ‘3 zse.pensoft.net km west of Bocoio, Benguela Province (12°28'58.0"S, 14°06'24.8"E)’ in Branch et al. (2017, 2021) is in error and we update it according to the original specimen la- bels and catalogue museum register. Etymology. The new species is named in reference to the area where it was found. The region lies on top of a ridge known as Morro do Pundo that translates to the ‘Hills’ or ‘Mountain’ of the Baboons. The name thus comprises two parts: pundo (= baboon) and montana (= mountain). Diagnosis. A member of the greater ‘transvaalica’ group, possessing two pairs of enlarged scansors per digit, and a strongly verticillate and flattened tail (Jacob- sen et al. 2014). As part of the A. bogerti group it dif- fers from other members of the ‘transvaalica’ group by having 78-82 midbody scale rows (versus 97—102 in A. gorongosa, 113—120 in A. loveridgei, 102-119 in A. transvaalica), and rostral excluded from the nostril (in contact in A. gorongosa) [Note: in Branch et al. (2021) it was incorrectly recorded that the rostral is in contact with the nostril in the A. bogerti-group]; with the supra- nasals always being in contact (separated by 1-3 granules in A. gorongosa; always in broad contact in A. loveridgei, usually in broad contact in A. transvaalica ~ 3—18%); and in having 13-15 scales between the anterior borders of the eyes (19-22 in A. gorongosa; 15—19 in A. loveridgei; 15-20 in A. transvaalica) (comparative data fide Branch et al. 2017, 2021). Afroedura pundomontana sp. nov. differs from other members of the A. bogerti group by a combination of the following characteristics (see Tables 5-6): midbody scale rows 78-82 (mean 79.5) (71-72 [mean 71.5] in A. otjihipa sp. nov., 65—67 [mean 66.0] in A. donveae, 69-77 [mean 73.5] in A. bogerti, 73-78 [mean 74.8] in A. praedicta, 73-88 [mean 79.5] in A. wulfhaackei, 73- 86 [mean 80.3] in A. vazpintorum); by the supranasals always being in contact (~33% of the time in A. bogerti; ~57% in A. wulfhaackei; always in contact in A. donveae, Zoosyst. Evol. 98 (2) 2022, 435-453 f eae! : 3 + &.° she a z ert : we ’ a 443 ¢,. Pr. > i yo Ay ee oh ae Figure 3. Live specimens of: A-B. Afroedura bogerti (A. P1-286, not vouchered; B. JURZC0015, not vouchered) from Serra da Namba, Cuanza Sul Province, Angola; C—D. Afroedura pundomontana sp. nov. (FKH0689) from Morro do Pundo, Benguela Prov- ince, Angola. Photos: A, C, D. Pedro Vaz Pinto; B. Javier Lobon-Rovira. A. vazpintorum, A. praedicta and A. otjihipa sp. nov.); each tail verticil comprising 4-5 (mean 4.4) ventral and 5-6 (mean 5.6) dorsal rows of scales (mean 4 ventral and 5 dorsal in A. bogerti, A. praedicta and A. wulfhaackei, 5—6 [mean 5.5] ventral and 6—7 [mean 6.6] dorsal in A. donveae,; 5—6 [mean 5.0] ventral and 6-7 [mean 6.1] dorsal in A. vazpintorum;, 5 ventral and 6 dorsal in A. otjihipa sp. nov.); ventral surfaces greyish with scattered small black spots (similar to A. bogerti, A. praedicta and A. wulfhaackei, immaculate in A. donveae, A. vazpintorum and A. otjihipa sp. nov.). Afroedura pundomontana sp. nov. differs from its sister highland species A. bogerti in having higher numbers of midbody scale counts (78-82 [mean 79.5] versus 73-78 [mean 74.8]), supranasals always in contact (versus ~33% of the time), and the posterior scales of the dorsal W-shapes crossbars dark black (versus same colour as cross bands; Fig. 3); 1t differs from A. wulfhaackei in that the supranasals are always in contact (versus ~57%). Holotype description. Adult male; SVL 46.0 mm; tail 42.3 mm (detached full original tail). Small mid-ventral incision for removal of liver sample. Measurements and meristic characters of holotype are presented in Table 6. Head and body dorsoventrally compressed; HL 12.5 mm, HW 8.3 mm, broadest at posterior level of eye; head 1.51 times longer than wide. Eye large (2.6 mm wide), pu- pil vertical with indented margins; circumorbital scales small and smooth, elongate at upper anterior margin, upper three posterior scales with small upward pointing spines. Snout rounded, 4.9 mm long, longer than distance between eye and ear openings (3.8 mm). Scales on top of snout smooth, rounded; scales at the edge larger than cen- tral ones, with no intervening minute granules. Scales on snout slightly larger in size to those on the back of head or the nape. Scales on eyelids larger than those on the crown, six scales deep from circumorbital scale to crown. Circumorbital scales separated from the larger scales on the eyelids by two rows of smaller scales. Nostril pierced between first supralabial and three nasal scales; rostral excluded from nostril; 1** supralabial narrowly excluded from nostril; supranasal much larger than the postnasals (which are about equal in size) and in broad contact. Nos- trils slightly elevated. Rostral roughly rectangular but with the upper edges slightly elongated due to extensions to the supranasals. Eight supralabials on either side, the labial margin flexing upwards at the rictus (approx. mi- dorbital position), with 3-4 minute scales proximal to the flexure. Nine infralabials on either side, with a small scale proximal to the flexure. At the lip, mental slightly narrow- er than adjacent infralabial; mental only two thirds the width of rostral (1.1 mm versus 1.8 mm respectively), and in contact with three rounded postmental scales. Scales zse.pensoft.net 444 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa Figure 4. Holotype of Afroedura pundomontana sp. nov. (PEM R24743) from Morro do Pundo, 25 km west of Bocoio, Benguela Province, Angola. Scale bar: 1 cm. Photos: Werner Conradie. on throat notably smaller than those on belly, but the scales touching the infralabials are larger. Fourteen scales across the crown at level of front of eyes; 18 scales from ear to eye; 83 scales around midbody. Ear opening deep, oblique and more-or-less round, nearly symmetrical (0.7 x 0.8 mm). Scales on dorsum smooth, closely set but juxta- posed, largest at mid-body, smaller on nape and tail base. Scales on venter flattened, not overlapping, more-or-less ovate at mid-ventrum, about twice the size of lateral gran- ules and about 1.5 times larger than the scales along the backbone. Original tail slightly dorsoventrally flattened and distinctly verticillate (10 distinct verticils in total), with obvious lateral constrictions that become less dis- tinct towards the tip of the tail; each verticil comprising 6 rows of imbricate scales dorsally and 4 rows of imbricate ventrally, with ventral scales approximately twice the size of those on the dorsal surface. Limbs well-developed, hindlimbs slightly longer than forelimbs, no notable mite pockets (dermal crevices inhabited by small ectoparasit- ic mites) at anterior or posterior margin of hind limbs. zse.pensoft.net All digits with a large pair of distal scansors, separated by a large, curved claw, and followed after a large gap (twice the length of terminal scansor) by a smaller pair of scansors; infero-median row of digital scales enlarged transversely, particularly towards the scansors, where the terminal scale adjoining the first pair of scansors may be medially constricted, swollen and scansor-like; seven en- larged subdigital lamellae on 4" toe. Paratype variation. (see Table 6 for more meas- urements and scale counts of type series). SVL 43.4— 57.8 mm; head length 1.19—1.50 times head width; snout 1.20—1.93 times the diameter of eye. Supranasals always in contact; the first upper labial and rostral always enter the nostril, and the width of the rostral at the lip margin is always wider than that of the mental; 2-3 postmental scales; supralabials 9, infralabials 9; scales between anterior edges of eyes 16—19; scales between nostril and anterior edge of orbit 10—13; scales between anterior edge of ear and rear margin of orbit 16—19; scales around mid- body 78-83; subdigital lamellae of 4 toe 7-9; dorsal Zoosyst. Evol. 98 (2) 2022, 435-453 BE 445 Figure 5. Habitat photos of: A-B. Afroedura pundomontana sp. nov.: Morro do Pundo, 25 km west of Bocoio, Benguela Proy- ince, Angola; C—D. Afroedura otjipha sp. nov.: Otjihipa Middleberg, Kunene Region, Namibia. Photos: A—B. Pedro Vaz Pinto; C_D. Francois Becker. Table 6. Measurements (in mm) and scale counts for the type series of Afroedura pundomontana sp. nov. Catalogue Number PEM R24743 TM 46587 TM 46588 TM 46589 TM 46590 FKHO688 FKHO689 Type Status Holotype Paratype Paratype Paratype Paratype Paratype Paratype Sex Female Female Female Male Female Female Female Snout-vent length 46.0 57.8 43.4 57.8 53.4 54.4 57.1 Tail length 42.3 61.5 - 44.1 47.33 - 47.1 Tail condition Original Original Truncated Regenerated Truncated Truncated Partly Regenerated Head length 12.5 13.0 10.6 13.3 12.7 13.1 15.0 Head width 8.3 10.3 8.3 11.2 10.1 10.6 11.9 Snout length 49 Bel 4.1 4.5 5.0 4.6 4.8 Eye distance 2.6 3.5 3.1 3:7 4.0 2.6 2.5 Eye-Ear distance 3.8 4.5 3.5 4.7 3.7 4.4 4.3 Precloacal pores (males) - - - 12 - - - Dorsal rows per tail verticil 4 4 5 - 4 - 5 Ventral rows per tail verticil 6 5 6 - 5 - 6 Scales below 4" toe 7 8 9 8 8 7 7 Midbody scale rows 83 81 78 82 78 78 80 Scales between eyes 14 14 13 13 13 15 15 Scales: nostril to eye 11 10 10 10 11 12 13 Scales: ear to eye 18 15 16 16 16 19 18 Supranasals in contact Yes Yes Yes Yes Yes Yes Yes Supralabials 8 ? ? ? ? 9 9 Infralabials 9 2 iu ? ? 9 9 scale rows per tail verticil S—6; ventral scale rows per tail verticil 4—5. Precloacal pores 12 (single male). Colouration. In life (holotype PEM R24743 [similar to Fig. 3C—D]): Greyish above with five evenly spaced darker crossbars from the occiput to the sacrum, each crossbar consisting of 9-12 dark scales forming a distinct W-shape, that consist anteriorly of a mix of grey and mus- tard scales and posteriorly by more prominent dark grey to black scales; each dark crossbar is separated by a mix of lighter grey scales; head with irregular dark grey blotches on the crown with intervening pale grey and mustard col- ouration; dark grey bar from nostril to the anterior margin zse.pensoft.net 446 of the ear opening; a vague, thin grey canthal stripe, ex- tends on both sides from the nasal region to anterior mar- gins of eye; upper and lower labials light grey anteriorly and beige posteriorly with fine black specks; lateral sides of the body with a mix of light grey and light cream-yel- low; limbs light greyish above with scattered darker grey markings interspersed with cream-yellow; tail with eight dark brown to black crossbands, becoming increasingly more bold towards the tip; iris gold in colour with a nar- row black elliptic pupil with crenulated edge, and black reticulation with light grey intervening blotches; venter uniform greyish with scattered black specks; ventral part of limbs with scattered black specks, more prominent than on the underparts. In preservative (Fig. 4): Dorsum with five evenly spaced dark brown W-shaped crossbars from the occiput to the sacrum with beige intervening blotches; ventrum 1s beige with numerous small scattered black specks on each scale, more prominent posteriorly; tail with eight dark brown to black cross bands. Para- type colouration variation: greyish above with five to six evenly or irregularly spaced darker grey-black W-shaped crossbars from the occiput to the sacrum, anterior part of these crossbars much darker than the posterior part, which is scattered with mustard coloured scales; lateral sides of body with a mixture of darker grey and mustard coloured scales; limbs and tail with grey blotches, with scattered mustard coloured scales; ventrum uniform grey- ish with scattered black specks. Natural history and habitat. (Fig. SA—B). A rupi- colous species found in rugged landscape between 600 to 1,000 m a.s.l. No details are available regarding the conditions under which the historical material was col- lected, but the new material was collected during the day, underneath vertical flakes in large granite boulders. On both occasions, several individuals were sheltering under the same flake. They were found in rocky outcrops in an- thropogenically disturbed mixed escarpment woodland, characteristic of the ecotone between the arid coastal plain and the inland mesic Angolan plateau. The presence of shrubs and small trees surrounding the granite outcrops suggests that these geckos might forage at night in the vegetation as reported for other Angolan species (Branch et al. 2021). Distribution and conservation. This species is cur- rently known only from central Benguela Province, An- gola (Fig. 2). It was collected at three localities in close proximity to one another, on the first elevation step of the Angolan escarpment, inland from the town of Lobito. The species may be more widely distributed as our pre- dicted mapping indicates but, so far, surveys conducted on the coastal plain and elsewhere along the escarpment did not produce additional material. Even around the type locality, the species proved to be uncommon and quite difficult to find, partly due to the inaccessible topogra- phy, but apparently also due to scarcity of granite flakes. Populations in isolated granite outcrops may be threat- ened by removal of rock flakes for construction of homes and other buildings. In accordance with IUCN Red List zse.pensoft.net Werner Conradie et al.: Two new species of Afroedura for southwestern Africa Guidelines (IUCN 2022) we propose this species to be classified as Data Deficient (DD) at this stage. Afroedura otjihipa sp. nov. https://zoobank.org/B38 1 8B24-7FF0-49B6-B6F9-F580BF50C0CO Otjihipa Flat Gecko (English) Otjihipa Platgeitjie (Afrikaans) Figs 5C—D, 6C—D, 7, Tables 5, 7 Synonym. Afroedura cf. bogerti — Branch 1998: 232; Griffin 2002: 20, 2003:10; Herrman and Branch 2013: 5. Holotype. NMNW R11253, adult female, collect- ed from Otjihipa Middleberg (-17.28314, 12.66506, 1,900 m as.l.), Kunene Region, Namibia, by Morgan Hauptfleisch, Francois Becker, Vera De Cauwer, Wessel Swanepoel and Ernst van Jaarsveld on 23 April 2021. Paratype. NUNW R11245, adult male (paired with female NMNW R11253 in same rock crack). Same col- lection details as holotype. Etymology. The new species is named in reference to the area it was collected, namely Otjihipa Mountains in northern Namibia. Diagnosis. A member of the greater ‘transvaalica’ group as it possesses two pairs of enlarged scansors per digit and a strongly verticillate and flattened tail (Jacob- sen et al. 2014). Part of the A. bogerti group which dif- fers from other members of the ‘transvaalica’ group by having less than 72 mid-body scale rows (vs. 97-102 in A. gorongosa, 113-120 in A. loveridgei, 102-119 in A. transvaalica), rostral excluded from the nostril (in contact in A. gorongosa), supranasals always in contact (separated by 1-3 granules in A. gorongosa; always in broad contact in A. /overidgei; usually in broad contact in A. transvaalica ~ 3—-18%); and 15—16 scales between an- terior borders of the eyes (19-22 in A. gorongosa, 15-19 in A. loveridgei, 15—20 in A. transvaalica) (comparative data from Branch et al. 2017, 2021). Afroedura otjihipa sp. nov. differs from other members of the A. bogerti group by a combination of the following characteristics (see Tables 5 and 7): 65-67 (mean 66.0) mid-body scale rows (64—78 [mean 72.8] in A. donveae, 69-77 [mean 73.5] in A. bogerti, 73-78 [mean 74.8] in A. praedicta, 78-82 (mean 79.5) in A. pundomontana sp. nov.; 76-88 [mean 79.3] in A. wulfhaackei, 73—86 [mean 80.3] in A. vazpintorum);, supranasals always in contact (similar to.4. donveae, A. vazpintorum, A. praedictaand A. pundomontana sp. nov.; in contact in ~ 33% of A. bogerti; in contact in ~ 57% of A. wulfhaackei), each tail verticil comprises 5 ventral and 6 dorsal rows of scales (mean 4 ventral and 5 dorsal in A. bogerti, A. praedicta and A. wulfhaackei, 4—5 (mean 4.4) ventral and 5—6 (mean 5.6) dorsal in A. pundomontana sp. nov.,; 5—6 [mean 5.5] ventral and 6—7 [mean 6.6] dorsal in A. donveae; 5—6 [mean 5.0] ventral and 6—7 [mean 6.1] dorsal A. vazpintorum), ventral surfaces light cream and almost immaculate, with some scattered dark spots near lateral edges (similar to A. donveae and A. vazpintorum, greyish Zoosyst. Evol. 98 (2) 2022, 435-453 Table 7. Measurements (in mm) and scale counts for the type series of Afroedura otjihipa sp. nov. Catalogue Number NMNW R11253 NMNW R11245 Type Status Holotype Paratype Sex Female Male Snout-vent length 57.9 59.9 Tail length - - Tail condition Truncated Regenerated Head length 13.6 15.9 Head width 13.2 13.3 Snout length 5.7 6.0 Eye distance 3.2 3.8 Eye-Ear distance 4.8 5.4 Precloacal pores (males) - 12 Dorsal rows per tail verticil 5 5 Ventral rows per tail verticil 6 6 Scales below 4" toe 8 8 Midbody scale rows 67 65 Scales between eyes 14 14 Scales: nostril to eye 10 11 Scales: ear to eye 12 13 Supranasals in contact Yes Yes Supralabials 8 9 Infralabials 9 8 with black spots in A. bogerti, A. wulfhaackei, A. praedicta and A. pundomontana sp. nov.); larger average adult size 58.2 mm SVL (versus 57.6 mm in A. donveae, 51.7 mm in A. wulfhaackei, 51.3 mm in A. vazpintorum, 50.3 mm in A. pundomontana sp. nov., 50.0 mm in A. bogerti, 49.9 mm A. praedicta), and by having very distinct black-and-white tail banding (similar to A. donveae). Afroedura otjihipa sp. nov. differs from its sister lowland species A. donveae in having a brown or copper coloured (versus black) iris, a relatively broader head (mean HL/HW 1.1 versus 1.3), and in dorsal colour pattern (Fig. 6): in A. otjihipa sp. nov. it is dominantly dark brown, the yellow appearing as small asymmetrical, irregular patches, and as irregular borders of four paired, asymmetrical, irregular, roughly triangular brown blotches, which merge at the scapular and sacral regions to form two additional bands (versus roughly symmetrical brown patterns on a mostly yellow background in A. donveae). Holotype description. Adult female: SVL 57.9 mm; tail regenerated, with a small mid-ventral incision for the removal of liver sample. Measurements and mer- istic characters of holotype presented in Table 7. Head and body dorsoventrally depressed; HL 13.6 mm, HW 13.2 mm, head broadest posterior level of eye and 1.02 times longer than wide. Eyes large (3.2 mm wide), pupil vertical with indented margins; circumorbital scales small and smooth, bottom posterior scales with small upward pointing spines. Snout rounded, 5.7 mm long, longer than distance between eye and ear openings (4.8 mm). Scales on top of snout smooth, rounded, similar in size, with no intervening minute granules. Scales on snout slightly larger than those on back of head or nape. Scales on eye- lids larger than those on the crown, 5 scales deep from circumorbital scale to crown. Nostril pierced between first supralabial and three nasal scales; rostral narrowly excluded from nostril; supranasals much larger than the smaller postnasals, ventral postnasal being about half the 447 size of its dorsal counterpart, and all in broad contact with one other. Nostrils very slightly elevated. Rostral roughly rectangular, but with its upper edges elongated due to ex- tensions toward the nostril, and the central point extends between the nasals. Seven supralabials on each side, the labial margin flexing upwards at the rictus (approx. mid-orbital position), with 1—2 elongate scales proximal to the flexure and several minute scales along the flex- ure proximal to these. Seven infralabials on either side. At the lip, mental scale slightly narrower than adjacent infralabial, mental only two thirds the width of rostral (1.1 mm versus 1.8 mm respectively) and in contact with three postmental scales; mental similar in size and shape to the surrounding gular scales, the central one of which is distinctly smaller. Scales on throat much smaller than those on belly, scales touching infralabials larger. Four- teen scales across the crown at level of front of eyes; 10 scales between nostril and front of eye; 12 scales from ear to eye; 67 scales around mid-body. Ear opening deep, oblique and roughly oval, less than half as high as wide (0.42 x 0.95 mm respectively). Scales on dorsum smooth, non-overlapping, largest at mid-body, smaller on nape and tail base. Scales on ventrum flattened, not overlapping, roughly twice the size of lateral granules and 1.4 times the size of scales along the dorsal mid-line. Regenerat- ed tail dorsoventrally flattened, roughly as broad as the neck, with ventral scales larger than those on the dorsal surface. Limbs well-developed, hindlimbs slightly longer than forelimbs; all limbs without obvious mite pockets at posterior or anterior margin of limb insertions. All digits with a large pair of distal scansors, separated by a curved claw, notably smaller on the fingers than toes, and fol- lowed after a gap (about the width of terminal scansor) by a smaller pair of scansors; infero-median row of digital scales slightly enlarged transversely, the distal two rows being paired in both digits and toes, where the terminal scale adjoining the first pair of scansors may be swollen and scansor-like; 6 enlarged central and two paired dis- tal scale rows under 3" toe, while other toes have paired scale rows, 8 under the 4" toe. Paratype variation. SVL 59.9 mm adult male, tail truncated, precloacal pores 12. Measurements and mer- istic characters of paratype are presented in Table 7. The paratype is very similar to the holotype with regard to scalation. Colouration. In life (holotype NMNW_ R11253, Fig. 7C—D): dark brown with yellowish patterns, fading to whitish on limbs and top of head; yellow patterns are irregular, asymmetrical patches and spots along the body, symmetrical paired spots around the nape and near the tail base; there is a thin, irregular, broken or continuous yellow bar on the nape; another broken, irregular yellow bar across the scapular region to the shoulders; three asymmetrical yellow double-bars which may present as pairs of medially-angled triangles posteriorly, across the back, each with an irregular dark brown core; anoth- er broken yellow bar or collection of symmetrical spots around the sacrum; head dark brown with yellow blotches zse.pensoft.net Werner Conradie et al.: Two new species of Afroedura for southwestern Africa Figure 6. Live specimens of: A—-B. Afroedura donveae from Omahua, Namibe Province, Angola (not sampled); C—D. Afroedura otjihipa sp. nov. (holotype female, NUNW R11253) from Otjihipa Middleberg, Kunene Region, Namibia. Photos: A—B. Javier Lobon-Rovira; C—D. Francois Becker. on the crown with intervening pale yellow colouration; dark brown bar from nostril across the upper margins of the ear opening, connecting with dark brown lateral bar on the neck; a thin pale yellow canthal stripe extends on both sides from the nasal region to anterior margins of eye, continuing posteriorly from the eye onto the nape; skin above eyes copper blue with dark brown spots; upper and lower labials light grey with dense brown speckling, denser anteriorly and on supralabials; lateral sides of the body with a mix of dark brown and yellowish blotches, as a continuation of the dorsal patterns; limbs dark brown above with scattered light grey markings; tail (regener- ated) with an asymmetrical chequered pattern of dark brown and light grey; iris copper in colour with a nar- row black elliptic pupil with crenulated edge, and black reticulation; venter uniform beige with scattered brown specks mostly on lateral edges; ventrally, limbs with scattered brown spots, mostly near lateral surfaces. In preservative: yellowish patterns faded to light grey, dark brown to grey-brown, and eyes faded to bluish grey, with original colouration of pupils and iris no longer evident. Paratype colouration: Similar colouration and patterning as to the holotype, but the yellow bands and patterns are more clearly defined: the bar on the nape is nearly contin- uous, that on the scapular region has a clear dark brown core, and three pairs of asymmetrical, medially-pointing, zse.pensoft.net dark brown, triangular blobs are clearly outlined by ir- regular yellow lines; no clear bar near the tail base, but a collection of symmetrical spots. The original tails are not present on the preserved specimen, but were observed briefly in life before capture. The original tails of another pair of individuals in a nearby rock crack (not caught) were also observed. Tail bars could not be counted, but bold black-and-white banding was clearly visible. Natural history and habitat. A rupicolous species living in narrow rock crevices in relatively small sand- stone outcrops in arid woodland savannah (Fig. 5C—D), at elevations of 1,800—1,900 m a.s.]. in the Otjihipa Moun- tains. It was not found in the dolomite formations near the type locality, despite greater search time dedicated to those areas. The rock cracks where they were found were smaller than is typical for this group and were sim- ilar throughout this surface formation. Congeners in the A. bogerti group are normally found only in deep rock cracks in and amongst large boulders. Such habitat fea- tures were present in the surveyed area, but only in do- lomite formations. The much less crevice-rich sandstone formation, with thin, straight cracks formed between the sandstone strata, appeared to be favoured syntopically by A. otjihipa sp. nov. and Cordylus namakuiyus. Distribution and conservation. Currently known from a single sandstone ridge on Otjihipa Middleberg in Zoosyst. Evol. 98 (2) 2022, 435-453 449 Figure 7. Holotype of Afroedura otjihipa sp. nov. (NMNW R11253) from Otjihipa Middleberg, Kunene Region, Na- mibia. Scale bar: 1 cm. Photos: Francois Becker. the extreme north-west of the Kunene Region, Namibia (Fig. 2). The species remains poorly known, but it is prob- ably stable in numbers as the local habitat 1s currently not threatened and is topographically unsuitable for human habitation. It likely occurs more broadly across the Otyji- hipa Mountain range. In accordance with IUCN Red List Guidelines (IUCN 2022) we propose this species to be classified as Data Deficient (DD) at this stage, but due to the remoteness of the locality and because no notifiable threats exist, it could be listed as Least Concern. Updated key to the Afroedura bogerti-group (updated from Branch et al. 2021) 1 MIKCIOGIE's Cales OWS JM Ore i eag.o unc cnc cony ese ose e rains ectere seme mobs ya ack aren oeieae deeee tac ove stsne seetes een ee natant ena eaenesee ees 2 — Midbody scale rows less than 95; occurs in northern Namibia and Angola ..................cccccsscceeceecceeseeeaeccussentenecsacaseceens 4 Zz RoOSthal "Still NM DOrGenineaO Sui Airlie ager a oat tenn ea ale lod he ce Ret eta ee ass el kee eo 22 tales S Rostralruisual i SxChuide ch tren (MOST snus nczpe kau ees oe sce dane deeb apne xy adleste nena iden anv was eee dace ddlasas eee vent teed ans anenes A. loveridgel Anterior nasals in contact (very rarely separated); scales around midbody: South Africa 102-118 (mean 109), northern PALACIO Fa) CN LCA NG Fes ALLIS ei ge Fey WL 4 a a OR a Sl oer gn mets RP Oa ea A. transvaalica Anterior nasals separated by 1-3 granules; scales around midbody 99-101 (average 100)............. ccs A. gorongosa zse.pensoft.net 450 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa 4 Each tail verticil usually comprising 5 ventral and 6 dorsal rows of scales; anterior nasals always in contact; ventrum Fig 8) Osea | st Seg eC Aan ok RE | te Sle OR: 3 5 RR «CRD fe 5 - Each tail verticil usually comprising 4 ventral and 5 dorsal rows of scales; anterior nasals not always in contact; ventrum SLEVISMAVENTeSiel | IOVACK ASO SCK Su. ces sc. amit: hol RE ES eas RS ei le BRO soos le SS ea Ee 6 5 Midbody scales 78-82 (mean 79.5) larger average adult size 57.1 mm SVL; precloacal pores 11-12 (mean 11.5) in Males bole coLlourallon MOlLaCkK LAS OC CUESA IMAM BOlae..c5. ccttranetss -heleeiagineas delysincs duis ciecbebionns sesteibodedas mtkebataes caalidemens= A. donveae - Midbody scales 65-67; larger average adult size 59.9 mm SVL; precloacal pores 12 in males; bold colouration, golden WS OCCUrS Thrall tics ace. uaa wee edie eeees aces teeusrel meee. ern ieen dace Rnntehe ca ue enters sr ees ae nag eae emede staat A. otjihipa sp. nov. — Midbody scales 73-86 (mean 80.3); smaller average adult size 48.6 mm SVL; precloacal pores 9-11 (mean 10.2) in res Cll eG OUIPAULO Meee dare e i a hn eine at se tc ee a eth bk won ats te eee a, Ula GDh se | A. vazpintorum 6 Anterior nasals always in contact, restricted to first elevation step and isolated inselbergs below the Angolan escarp- (A=) 2 Ree eee ne er Re, Rind eo WA Rs ght ce a oe RD Niet ae ge sea aero arenes Ae, Ae ee ee a oe fi - Anterior nasals not always in contact, restricted to above the Angolan escarpment.............::cccceeeceeeeeeeeeeeeeeeeeeeeeeeaeeees 8 7 Restricted to Serra de Neve Mountains, Namibe Province.........cccccc ccc cec ec cce cece cece eee e cease eeeaeaeeeeeeeegeeeeaeaeaeaeaees A. praedicta - Occurs on the first elevation step of the Angolan escarpment, Benguela Province..................:5 A. pundomontana sp. nov. 8 Wiel Oy -SCaleS 697: CITA Sto) sxc sss vor eee pate PORE rae 7a gia OR 9 > UN arg URL PERS Pana ra ght BA Pee a SERIE A. bogerti - Midbecy-sSCalesto=Sesiihicat~ A973, a. eet ry peek sree ered oy pus Ars bor Merwe 8d ows, 08 res were bes Pelee Prom bres ew ora oe A. wulfhaackel Discussion be confirmed in future studies which utilise all known The addition of newly collected material from northern Namibia and western Angola, as well as the inclusion of more genes, improved the phylogenetic support values and relationships between species. The topology recovered remained very similar to those found in Branch et al. (2017, 2021), except this study resulted in the addition of two novel lineages here described as new species: A. pundomontana sp. nov. and A. ofjihipa sp. nov. To avoid a potential overestimation of the number of taxa in a dataset, we took a conservative approach, recognising only those lineages where consistent morphological differences were evident and where genetic differences exceeded 5% in the /6S gene as representing valid species. Deep genetic sub-structuring persists within A. wulfhaackei and A. vazpintorum and it will require finer-scale genetics to resolve their taxonomic status. Afroedura is a genus of flat geckos that is restricted to rocky and mountainous habitat and inhabits both the eastern and western southern Africa escarpments, while the central Kalahari and Zambezi basin regions (generally fluvial lowlands and few or no mountains or inselbergs) appear to be devoid of these flat geckos (Branch 1998; Jacobsen et al. 2014). However, there are surprising links between the eastern and western populations, with A. transvaalica, A. loveridgei and A. gorongosa from eastern Zimbabwe and adjacent Mozambique highlands being sister species to the A. bogerti-group in western Angola and adjacent Namibia (Jacobsen et al. 2014; Branch et al. 2017, 2021). Preliminary dating analyses (results not shown here) suggest that the initial split between the two main A. bogerti lineages (praedicta + otjihipa + donveae + vazpintorum and pundomontana + bogerti + wulfhaackei) occurred between the late Pliocene and early Pleistocene, thus suggesting that rocky habitats throughout the southern inland parts of Africa may have, historically, been connected between eastern and western extremes. These preliminary dating findings still need to zse.pensoft.net species of the genus Afroedura. The Afroedura bogerti-group is endemic to the central highlands and south-west coastal regions of Angola, with One species endemic to the Otjihipa Mountains, south of the Kunene River, in the northern Kunene region of Namibia. This group does not extend further south into the Namibian portion of the western escarpment. The only other species of Afroedura known to occur intermittently along the highlands of Namibia are taxa in the A. africana- group, which are closely related to South African species, and A. tirasensis, the phylogenetic relationship of which is still unknown (Jacobsen et al. 2014). Our results show that the Afroedura bogerti-group Speciation 1s driven by the complex landscape mosaic of rocky/mountainous and flat lowland habitats, under the influence of the steep climatic gradient characteristic of the Angolan escarpment region and exacerbated by Pleistocene climatic events. The basal split in the group is between the clade inhabiting higher rainfall mountains in the north-eastern extreme of its range (4. bogerti, A. pundomontana sp. nov. and A. wulfhaackei), and the clade inhabiting the predominantly more arid coastal regions in the southwest (A. donveae, A. praedicta, A. otjihipa sp. nov. and A. vazpintorum). This could be associated by a major ecological landscape transformation, such as with the expansion of C, grasslands, which occurred in the period between 2.0 and 1.75 Mya, and which has led to strong evolutionary pressure and species turnover in other African fauna (Bibi and Kiessling 2015). The basal split within the inland highland lineages (A. bogerti + A. wulfhaackei + A. pundomontana sp. nov.) is between the strictly higher-lying inland species (A. wulfhaackei) and the subgroup formed by A. bogerti + A. pundomontana sp. nov. The latter subgroup further split into northern (A. bogerti) and western (A. pundomontana sp. nov.), lineages, currently present on an isolated moun- tain to the north and on the intermediate elevational step of the central escarpment, respectively. Although nomi- Zoosyst. Evol. 98 (2) 2022, 435-453 notypical A. bogerti is currently only known to occur at Namba Mountain, and is geographically and ecologically closer to some populations of A. wulfhaackei, the genetic results suggest a past link maintained along the Angolan western escarpment. Namba Mountain is unique in con- taining more extensive forested habitats than any other Angolan highland, and clear phylogenetic relationships between Namba and the western escarpment have been revealed in various faunistic groups, such as rupicolous dwarf toads of the genus Poyntonophrynus (Baptista et al. in prep.). More recently, several A. wulfhaackei pop- ulations may have become isolated on scattered moun- tain tops or granitic outcrops in the central highlands, leading to the evolution of the four subclades already identified. This study suggests that these subclades are a consequence of an ongoing incipient speciation process. Moreover, the habitat, ecological niche and morphologi- cal conservatism, seem to be consistent with non-adap- tive radiation, similar to what has been reported for other reptile radiations (e.g. Reaney et al. 2018). On the other hand, the south-western lineages (A. donveae + A. praedicta + A. otjihipa sp. nov. + A. vazpintorum) have a more complex history of contraction, recolonisation and secondary contact, probably due to climatic changes exerted ona dynamic and highly heterogeneous landscape. The former three species seem very localised in their distributions, with high levels of specialisation. Afroedura praedicta and A. otjihipa sp. nov. are present only on two inselbergs separated by over 400 km of arid lowlands, contrasting with A. donveae, which occurs in the Angolan Kaokoveld desert and is the only species within the southwestern lineages, exclusively found in lowlands. While A. donveae and A. praedicta are associated with isolated large granite boulders, A. otjihipa sp. nov. lives among small rocks and vegetation. This suggests that the speciation of this group was likely caused by vicariance following the severe contraction of a once widespread ancestral taxon. Geographically intermediate populations between northern A. praedicta and southern A. donveae likely disappeared in response to extreme climatic and habitat changes, creating a large unoccupied area in between, while A. ofjihipa sp. nov. may reflect a relatively recent colonisation (see Fig. 2). A deep split has been revealed within A. vazpintorum, leading to two subclades, and contemporaneous with the separation between A. praedicta and A. donveae + A. otjihipa sp. nov., when extreme environmental conditions may have forced various Afroedura populations to become isolated on mountain tops. Pinpointing the geographical origins of both subclades may not be possible, but their geographical distribution, together with the molecular results, suggest that when environmental conditions became suitable, A. vazpintorum expanded its range and populated the lowlands, eventually becoming the only widespread lineage within the genus. Although we lack occurrence records connecting coastal and inland subclades of A. vazpintorum populations, secondary contact is demonstrated by the presence of mitochondrial 451 introgression of the coastal subclade in two of the three southern highland sites surveyed, including at Bimbe. The relationships between the coastal and inland lineages of A. vazpintorum need further investigation by using more informative nuclear markers, and an increased survey effort at the base of the southern escarpment, to test for current gene flow isolation. Although the southwestern regions of Angola include some dramatic and heterogeneous topographic features and may have experienced geomorphological transformation throughout the Pleistocene, such as the gradual escarpment uplift (Feio 1981), it is likely that speciation in Angolan Afroedura was driven mostly by climatic factors. Our preliminary dating estimates seem to indicate that most node splits within the A. bogerti complex seem to have occurred in the Mid-Pleistocene Transition, a period when the change in orbital cycles led to shifts towards increasingly variable and drier climate in Africa, consequently promoting a speciation pulse (deMenocal 2004). All remaining splits recovered in our study, originating with A. donveae and A. otjihipa sp. nov. and leading to the diversification into subclades within A. wulfhaackei, must be of much younger origin but have very likely also resulted from strong environmental pressure. These diversification episodes could have been driven by glacial cycles, but linking each node split to specific climatic events will probably remain impossible. Many species occur in largely undisturbed remote ar- eas with little human interference or development, and populations can therefore be considered stable. Howev- er, most species are range-restricted, and future devel- opments could quickly change their conservation status. A steep climatic gradient influenced by a steep-sloped, west-falling escarpment and the influence of the Atlantic Ocean, may render these specialised and topographical- ly-isolated, high-altitude habitats particularly sensitive to climate change. The impacts of climate change on species endemic to high elevation have been found to be dispro- portionately high (Dirnbock et al. 2011). Therefore, sur- veys to monitor population size and determine population trends over time as associated with climatic changes, are a priority for the range-restricted species. Additional ex- ploratory surveys to improve the accuracy of projected Species ranges, particularly in the poorly-sampled Na- mibian Kaokoveld, are also needed. Acknowledgements We thank Instituto Nacional da Biodiversidade e Conser- vacao (INBC) of the Ministry of Culture, Tourism and En- vironment (MCTE, Luanda, Angola), and especially the director of INBC, Dr Albertina Nzuzi, for issuing research export permits. We also thank Fernanda Lages and Vladi- mir Russo for co-ordinating institutional relationships and facilitating work authorisations, and Afonso Vaz Pinto for collecting efforts at Bocoio. We acknowledge the import- ant role played by Rolf Becker, Ansie Bosman, Wessel zse.pensoft.net 452 Werner Conradie et al.: Two new species of Afroedura for southwestern Africa Swanepoel and Vera De Cauwer in conceiving and organ- ising the helicopter mountain-top surveys in Namibia and southern Angola, and Fernanda Lage and Vladimir Russo for co-ordinating institutional relationships and facilitating work authorisations in this regard. The SCIONA - Biodi- versity Survey of Mountain Tops in the Kaokoveld Centre of Endemism was done in partnership with the Ministry of Culture, Tourism and Environment of the Republic of An- gola and the Ministry of Environment, Forestry and Tour- ism of Namibia, and was funded by the European Union under grant agreement FED/2017/394-802. We thank Chad Keates for the generation of additional genes for newly collected material from northern Namibia, through the use of infrastructure and equipment provided by the NRF-SAIAB Aquatic Genomics Research Platform.. JLR were supported by Funda¢ao para a Ciéncia e Tecnologia (FCT) and BIOPOLIS (PD/BD/140808/2018 and BIOP- OLIS 2022-18, respectively). We thank Lemmy Mashini and Adriaan Jordaan for access to material housed in the Ditsong National Museum of Natural History (Pretoria, South Africa). This work was supported by the Europe- an Union’s Horizon 2020 Research and Innovation Pro- gramme under the Grant Agreement Number 857251 and by National Funds through FCT-Fundag¢Ao para a Ciéncia e Tecnologia, under the scope of project UIDP/50027/2020. References Bauer AM, de Silva A, Greenbaum E, Jackman T (2007) A new species of day gecko from high elevation in Sri Lanka, with a preliminary phylogeny of Sri Lankan Cnemaspis (Reptilia, Squamata, Gekkoni- dae). Mitteilungen aus dem Zoologischen Museum in Berlin 83: 22-32. https://doi.org/10.1002/mmnz.200600022 Bauer AM, Jackman TR, Greenbaum E, Giri VB, de Silva A (2010) South Asia supports a major endemic radiation of Hemidactylus geckos. Molecular Phylogenetics and Evolution 57(1): 343-352. https://doi.org/10.1016/j.ympev.2010.06.014 Benson DA, Cavanaugh M, Clark K, Karsch-Mizrachi I, Lipman DJ, Ostell J, Sayers EW (2013) GenBank. Nucleic Acids Research 41(D1): D36—D42. https://doi.org/10.1093/nar/gks1195 Bibi F, Kiessling W (2015) Continuous evolutionary change in Plio-Pleistocene mammals of eastern Africa. Proceedings of the National Academy of Sciences of the United States of America 112(34): 10623-10628. https://doi.org/10.1073/pnas. 1504538112 Bittencourt-Silva G, Conradie W, Siu-Ting K, Tolley KA, Channing A, Cun- ningham M, Farooq HM, Menegon M, Loader SP (2016) The phyloge- netic position and diversity of the enigmatic mongrel frog Nothophryne Poynton, 1963 (Amphibia, Anura). Molecular Phylogenetics and Evo- lution 99: 89-102. https://do1.org/10.1016/j.ympev.2016.03.021 Branch WR (1998) Field Guide to Snakes and other Reptiles of South- ern Africa. Struik, Pretoria, 236 pp. https://doi.org/10.2307/1447414 Branch WR, Guyton JA, Schmitz A, Bare; MF, Naskrecki P, Farooq H, Verburgt L, Rodel M-O (2017) Description of a new flat gecko (Squamata: Gekkonidae: Afroedura) from Mount Gorongosa, Mo- zambique. Zootaxa 4324(1): 142-160. https://doi.org/10.11646/ zootaxa.4324.1.8 zse.pensoft.net Branch WR, Schmitz A, Lobon-Rovira J, Baptista NL, Anténio T, Con- radie W (2021) Rock island melody: A revision of the Afroedura bogerti Loveridge, 1944 group, with the description of four new endemic species from Angola. Zoosystematics and Evolution 97(1): 55-82. https://doi.org/10.3897/zse.97.57202 Briscoe NJ, Kearney MR, Taylor CA, Wintle BA (2016) Unpacking the mechanisms captured by a correlative species distribution model to improve predictions of climate refugia. Global Change Biology 22(7): 2425-2439. https://doi.org/10.1111/gcb.13280 Burbrink FT, Lawson R, Slowinski JB (2000) Mitochondrial DNA Phy- logeography of the Polytypic North American Rat Snake (Elaphe obsoleta): A Critique of the Subspecies Concept. Evolution 54(6): 2107-2118. https://doi.org/10.1111/j.0014-3820.2000.tb01253.x Busschau T, Conradie W, Daniels S (2019) Evidence for cryptic diversifi- cation in a rupicolous forest-dwelling gecko (Gekkonidae: Afroedura pondolia) from a biodiversity hotspot. Molecular Phylogenetics and Evolution 139: e106549. https://doi.org/10.1016/j.ympev.2019.106549 Candau JN, Fleming RA (2005) Landscape-scale spatial distribution of spruce budworm defoliation in relation to bioclimatic conditions. Canadian Journal of Forest Research 35(9): 2218-2232. https://do1. org/10.1139/x05-078 de Queiroz K (1998) The general lineage concept of species, species crite- ria, and the process of speciation: A conceptual unification and termino- logical recommendations. In: Howard DJ, Berlocher SH (Eds) Endless forms: Species and speciation. Oxford University Press, Oxford, 1-19. deMenocal PB (2004) African climate change and faunal evolution during the Pliocene-Pleistocene. Earth and Planetary Science Letters 220(1—2): 3-24. https://doi.org/10.1016/S0012-821X(04)00003-2 Dirnbock T, Essl F, Rabitsch W (2011) Disproportional risk for habitat loss of high-altitude endemic species under climate change. Global Change Biology 17(2): 990-996. https://doi.org/10.1111/).1365- 2486.2010.02266.x Enriquez-Urzelai U, Kearney MR, Nicieza AG, Tingley R (2019) Integrating mechanistic and correlative niche models to unravel range-limiting processes in a temperate amphibian. Global Change Biology 25(8): 2633-2647. https://doi.org/10.1111/gcb. 14673 Feio M (1981) O relevo do sudoeste de Angola; estudo de geomorfo- logia. Memorias da Junta de Investigacao Cientifica do Ultramar, Lisboa, 32 pp. Fick SE, Hiymans RJ (2017) WorldClim 2: New 1-km spatial resolu- tion climate surfaces for global land areas. International Journal of Climatology 37(12): 4302-4315. https://doi.org/10.1002/joc.5086 Griffin M (2002) Annotated Checklist and Provisional National Conser- vation Status of Namibian Reptiles. Technical Reports of Scientific Services. Number 1. Directorate of Scientific Services, Ministry of Environment and Tourism, Windhoek, 180 pp. https://www.lacerta. de/AF/Bibliografie/BIB_ 4908 .pdf Griffin M (2003) Annotated checklist and provisional national con- servation status of Namibian reptiles. Namibia Scientific Society, Windhoek, 169 pp. Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium 41: 95-98. Herrman H-W, Branch WR (2013) Fifty years of herpetological re- search in the Namib Desert and Namibia with an updated and an- notated species checklist. Journal of Arid Environments 93: 94-115. https://doi.org/10.1016/j.jaridenv.2012.05.003 Zoosyst. Evol. 98 (2) 2022, 435-453 IUCN (2022) The IUCN Red List of Threatened Species. Version 2022- 1. https://www.iucnredlist.org [Accessed on 28 September 2022] Jacobsen NHG, Kuhn AL, Jackman TR, Bauer AM (2014) A phylogenet- ic analysis of the southern African gecko genus Afroedura Loveridge (Squamata: Gekkonidae), with the description of nine new species from Limpopo and Mpumalanga provinces of South Africa. Zootaxa 3846(4): 451-501. https://doi.org/10.11646/zootaxa.3846.4. 1 Kumar S, Stecher G, Li M, Knyaz C, Tamura K (2018) MEGA X: Molecular Evolutionary Genetics Analysis across computing plat- forms. Molecular Biology and Evolution 35(6): 154-1549. https:// doi.org/10.1093/molbev/msy096 Lanfear R, Frandsen PB, Wright AM, Senfeld T, Calcott B (2016) PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic anal- yses. Molecular Biology and Evolution 34: 772—773.https://do1. org/10.1093/molbev/msw260 Makhubo BG, Tolley KA, Bates MF (2015) Molecular phylogeny of the Afroedura nivaria (Reptilia: Gekkonidae) species complex in South Africa provides insight on cryptic speciation. Molecular Phylogenetics and Evolution 82: 31-42. https://doi.org/10.1016/). ympev.2014.09.025 Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, Lanfear R (2020) IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Molecular Biology and Evolution 37(5): 1530-1534. https://doi. org/10.1093/molbev/msaa015 Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ (2015) IQ-TREE: A Fast and Effective Stochastic Algorithm for Estimating Maxi- mum-Likelihood Phylogenies. Journal of Molecular Biology and Evolution 32(1): 268-274. https://doi.org/10.1093/molbev/msu300 Palumbi SR, Martin AP, Romano SL, McMillan WO, Stice L, Grabows- ki G(1991) The Simple Fool’s Guide to PCR. Dept. ofZoology, Uni- versity of Hawaii, Honolulu. Rambaut A. 2014. FigTree version 1.4.4. http://tree.bio.ed.ac.uk/soft- ware/figtree/ Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA (2018) Pos- terior summarisation in Bayesian phylogenetics using Tracer 1.7. 453 Systematic Biology 67(5): 901-904. https://doi.org/10.1093/sysbio/ syy032 Reaney AM, Saldarriaga-Cordoba M, Pincheira-Donoso D (2018) Macroevolutionary diversification with limited niche disparity in a species-rich lineage of cold-climate lizards. BMC Evolutionary Bi- ology 18(1): 1-12. https://doi.org/10.1186/s12862-018-1133-1 Ronquist F, Teslenko M, Van der Mark P, Ayres DL, Darling A, Hohna S, Larget B, Liu L, Suchard MA, Huelsenbeck JP (2012) MrBayes 3.2: Efficient Bayesian Phylogenetic Inference and Model Choice Across a Large Model Space. Systematic Biology 61(3): 539-542. https://doi.org/10.1093/sysbio/sys029 Stanley EL, Bauer AM, Jackman TR, Branch WR, Mouton LFN (2011) Between a rock and a hard polytomy: Rapid radiation in the rupi- colous girdled lizards (Squamata: Cordylidae). Molecular Phy- logenetics and Evolution 58(1): 53-70. https://doi.org/10.1016/). ympev.2010.08.024 Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG (1997) The ClustalX windows interface: Flexible strategies for mul- tiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24(24): 4876-4882. https://doi.org/10.1093/ nar/25.24.4876 Uetz P, Freed P, HoSek J [Eds] (2022) The Reptile Database. http:// www .reptile-database.org [Accessed 11 April 2021] Wiens JJ, Kuczynski CA, Townsend T, Reeder TW, Mulcahy DG, Sites Jr JW (2010) Combining Phylogenomics and Fossils in Higher-Lev- el Squamate Reptile Phylogeny: Molecular Data Change the Place- ment of Fossil Taxa. Systematic Biology 59(6): 674-688. https://doi. org/10.1093/sysbio/syq048 Yang XQ, Kushwaha SPS, Saran S, Xu J, Roy PS (2013) Maxent mod- eling for predicting the potential distribution of medicinal plant, Justicia adhatoda L. in Lesser Himalayan foothills. Ecological En- gineering 51: 83-87. https://doi.org/10.1016/j.ecoleng.2012.12.004 Zhang D, Gao F, Jakovli¢ I, Zou H, Zhang J, Li WX, Wang GT (2020) PhyloSuite: An integrated and scalable desktop platform for stream- lined molecular sequence data management and evolutionary phy- logenetics studies. Molecular Ecology Resources 20(1): 348-355. https://doi.org/10.1111/1755-0998.13096 zse.pensoft.net