Concordant patterns of genetic, acoustic, and morphological divergence in the West African Old World leaf‐nosed bats of the <i>Hipposideros caffer</i> complex

Levels of biodiversity are globally underestimated, especially in tropical ecosystems. This is particularly so for bats compared to other mammalian taxa, due morphological conservatism. Here, we investigate West African the Hipposideros caffer complex, an insectivorous bat group occurring throughout Afrotropics. From samples collected Ghana, aim resolve cryptic diversity identified by mitochondrial (mt) DNA using nuclear genetic, acoustic, and external morphometric data. We confirmed presence four previously established mtDNA lineages within H. complex found significant genetic divergence among based on microsatellite data differences frequencies echolocation calls measures. these new data, conclude that Africa consists at least distinct species. While small-sized species from coastal savanna could be assigned tephrus, taxonomic identity three sympatric, similarly sized pertaining ruber forest zone Central Ghana yet assessed. Die Artenvielfalt vieler Ökosysteme wird generell oft unterschätzt, insbesondere den Tropen. Dies trifft besonders für Fledermäuse zu, bei denen sich, im Vergleich zu anderen Säugetiergruppen, zahlreiche morphologisch sehr ähnliche Arten finden. Wir untersuchten die westafrikanischen Vertreter des - Komplexes, einer der gesamten Afrotropis verbreiteten Gruppe insektivorer Fledermausarten. Basierend auf Material aus wir vier über zuvor bereits identifizierte Linien mittels Kernmikrosatelliten-Daten sowie akustischer und morphometrischer Daten. Unsere Ergebnisse bestätigen Existenz von mtDNA-Linien innerhalb caffer-Komplexes zeigen nun auch anhand Kernmikrosatellitendaten eine signifikante genetische Divergenz einzelnen Linien, Unterschiede Echoortungsfrequenzen morphometrischen Parametern. neuen Daten belegen, dass caffer-Komplex Westafrika mindestens besteht: Während relativ kleine Individuen Küstensavanne Ghanas wohl zur Art tephrus gehören, kann taxonomische Identität drei ähnlich großen, sympatrischen Waldzone Zentralghana, aktuell zugeordnet werden, bislang noch nicht abschließend beurteilt werden. Species richness still relatively poorly known likely far greater than currently described (Fahr & Kalko, 2011; Herkt et al., 2016; Monadjem 2010; 2013; 2019; Taylor 2012). The difficulty resolving taxonomy partly explained a higher incidence this with many taxa (Jones Barlow, 2004). considered if they share high similarity have been traditionally form single species, but actually lines evidence (e.g. ecological, behavioral, morphological; Bickford 2007). In bats, constraints flight echolocation, relative importance auditory versus visual cues, mean often goes undetected 2004; Kingston 2001). recent years, molecular, analyses revealed number unsuspected (Jones, 1997; Jones 2001; Russo Jones, 2000; Thabah 2006). Old World leaf-nosed genus Gray, 1831, also as roundleaf highly speciose family Hipposideridae, which exhibits level (Foley 2015, 2017; Guillén-Servent Francis, 2006; 2013, Thong 2012; Vallo 2008). occurs sub-Saharan southwestern part Arabian Peninsula, comprises two recognized, though cryptic, species: Sundevall's (Sundevall, 1846) Noack's (Noack, 1893). usually associated dry open woodland, while inhabits rain wet woodland (Monadjem 2008; Wright, 2009). Although several subspecies recognized each their definition rather vague substantial variability. It has suggested further exists (Patterson 2020; Simmons, 2005; 2013). Previous analysis existence deeply divergent (Vallo 2008), split into multiple sublineages 2020), possibly constitute separate acknowledged. major were labeled A, B, C, D al. (2008). Lineage A comprised forms smaller morphotype, disjunct distribution areas tephrus. However, study Patterson (2020) showed distinction underestimates proper lineage; hence, previous designation adopted here particular sublineage northern belt. correspond larger morphotype (forearm length >48 mm; 2011). E, introduced later (2013) certain clades, falls lineage (cf. 2008, former endemic lamottei Mount Nimba belong represents another challenge complex. Further complexity shown (2020). phylogeny corroborates basic division main (2008), newly East together older redefined proposed putative limited nature maternal inheritance. Taxonomic classifications can result oversplitting (Lausen gene flow occur even clades (Avise, Castella Demos Petit Mayer, 1999). must therefore (Mayer Helversen, Microsatellites, bi-parentally inherited, powerful tools flow, hybridization, or introgression between lineages. Additional information, such call frequency used divergence. may morphologically conserved, acoustic indicative show constant component sensitive fovea fine-tuned discrimination Variation populations elsewhere continent (Guillén-Servent 1993; Webala 2019). resolution interest because discovered host coronaviruses (CoVs) zoonotic potential (Pfefferle Coronaviruses include viruses concern human public health, including SARS-CoV (Li 2005), MERS-CoV (Ithete 2013), 2019 novel coronavirus SARS-CoV-2 causes COVID-19 (Zhu 2020). Each origin Li Shereen Contact humans those common parts (Anti 2015), practices bushmeat consumption cave visitation directly linked outbreaks Leroy Knowledge important understanding virus ecology transmission dynamics, prediction emerging diseases 2015). aims some uncertainty processes maintain bats. assign sensu (2008) use microsatellites, measurements test hybridization (N = 1305) sampled August 2010 July 2012 eight colonies located caves, abandoned mines buildings (Figure 1, Table S1). Most sampling sites localized moist semi-deciduous (Kwamang, Likpe Todome) transition Guinean (Buoyem, Forikrom) administrative units Ashanti Volta Regions, Brong-Ahafo Region, respectively. forest–savanna zones characterized bimodal rainfall pattern minor seasons March September, annual averaging 1500 mm gradually diminishing increasing latitude (Abbam 2018). natural environment areas, surrounding sites, mostly transformed agricultural landscape (Pappoe 2010). One site (Elmina) was Region. arid lower average 800 Permission capture obtained Wildlife Division Forestry Commission Ministry Lands, Mines. Bats captured mist nets set entrances along trails one hour after dusk dawn. then sexed, aged, forearm mass measured vernier calipers 0.05 precision electronic balance 0.1 g precision. Individuals aged degree fusion fourth epiphyseal joint (Brunet-Rossinni Wilkinson, Two 3-mm wing membrane biopsy punch stored molecular-grade ethanol extraction. Sampled individuals released locality capture. Total genomic isolated 267 tissue (Appendix 1) QIAamp micro kit (QIAGEN) GenCatch Genomic Extraction (Epoch Life Science). 259 1), complete cytochrome b (cytb, 1140 bp) amplified polymerase chain reaction (PCR) primers L14724 (5′-CGAAGCTTGATATGAAAAACCATCGTTG-3′) H15915 (5′-AACTGCAGTCATCTCCGGTTTACAAGAC-3′; Irwin 1991) slightly modified versions L14724ag (5′-ATGATATGAAAAACCATCGTTG-3′) H15915ag (5′-TTTCCNTTTCTGGTTTACAAGAC-3′; Reactions performed following Taq (Promega) Combi PPP Mastermix (Top-Bio, Czech Republic). PCR products ca. 1200 bp long purified QIAquick purification Geneaid Fragments extraction (Geneaid), sequenced commercially (Macrogen, Seoul, Korea) respective amplification forward primers, Big-Dye Terminator sequencing chemistry (Applied Biosystems) ABI 3730xl sequencer. Sequences checked, manually edited, aligned Sequencher 5.1 (Gene Codes) Bioedit 7.1 (Hall, Aligned cytb sequences first shortened shortest sequence narrowed down haplotypes DnaSP 5.10 (Librado Rozas, For assignment Ghanaian comparison published regions, selected (2008, 2011) (2020), outgroup jonesi R. landeri 2, S2, Alignment All submitted GenBank under accession numbers MT051541–MT051580. Bayesian MrBayes 3.2 (Ronquist Huelsenbeck, 2003) reconstruction phylogenetic tree coupled MCMC runs 2,500,000 generations, 100 generations burn-in 25% trees. analyzed general time-reversible model gamma substitution rates categories portion invariable (GTR + G I), most appropriate evolution MrModeltest2 (Nylander, 2004) Akaike Information Criterion (AIC). visualized FigTree 1.4.2 (Rambaut, 2014). extracted genotyped nine loci: Hr2, Hr3, Hr5, Hr6, Hr7, Hr8, Hr10, Hr11, Hr12 (Baldwin 2014), aff. (sensu Primers 5’-fluorochrome bases (VIC, FAM, NED, PET) loci protocols outlined Baldwin (2014). electrophoresed ABI3130 Genetic Analyzer Biosystems). Allele sizes determined via manual inspection software Peak Scanner 1.0 Biosystems), followed automated binning Tandem 1.09. reanalyzed randomly chosen 20% ensure integrity. Microsatellites examined deviation Hardy–Weinberg equilibrium (HWE) linkage disequilibrium Genepop 4.2 (Raymond Rousset, 1995). structure performing discriminant principal components (DAPC; Jombart 2010) k-means cluster adegenet package R (R Core Team, looked tests program Geneclass2 2.0 (Piry individual, likelihood belonging its calculated genotype, probability 1000 Monte Carlo Markov permutations per population. Structure 2.3 (Pritchard 2000) evaluate Using admixture independent allele frequencies, identities POPID priors, 10 replicate 80,000 iterations 10,000 k 2 7. ∆k value assumed Harvester (Earl vonHoldt, 2012), Evanno (2005). Echolocation handheld recorded ultrasound detector (Pettersson D1000X; Pettersson Elektronik AB) rate 400 kHz. Calls Matlab 8.0 (Mathworks Inc). adult file, only loudest, non-overloaded analysis. Only minimum duration 4 ms assessed component. fast Fourier transformation (FFT) Hamming window 1024 (2.6 ms) overlap 98.4%. resulted maximum 390 Hz time 0.16 ms. FFT procedure progressively CF until amplitude dropped 22 dB below maximum. peak individual statistical 189), characteristics 1305), 1301) different subsets (Table 1; Appendix 1). Averages character Kruskal–Wallis tests. Post hoc Bonferroni-adjusted Mann–Whitney–Wilcoxon pairwise comparisons. Generalized linear mixed models (GLMM) examine association frequency, length, mass, nlme (Pinheiro GLMMs included sex fixed effects effect model. Random all reproductive status Frequency (kHz) n 189 Forearm Length (mm) 1305 Mass (g) 1301 them 12 Elmina geographical exclusivity obvious therein population, presented text. resulting alignment 782 42 them, Hap 8 14, corresponded registered HQ343266 HQ343265, 40 (MT051541–MT051580; reconstructed 2) clustered lineages: haplotype representing placed six 45 16 37 19 169 D. Pairwise grouped varied 8.2% 10.4%. B compact clade internal reaching up 0.3%, C subclades differing 1.9–2.8% 1.8–2.2, As geographic occurred coast region. abundant seven Ashanti, Brong-Ahafo, is, except Elmina. Three lineages, D, sympatry Buoyem caves region, well regions. Interestingly, conformed geographically delimited west east Lake (sublineages D3 D4, Figure 2). No noted bearing haplotypes, 28 29, clustering Cameroon, Kwamang haplotypes. Nine S3) no (n 12, HO 0.677, HE 0.663, FIS −0.023). There null alleles locus Hr1 Hr13 removed analyses. remaining loci, consistent departures HWE detected Bonferroni correction. K-means clusters. DAPC scatterplot shows separation groups 3). Lineages clearly delineated other. slight 95% confidence ellipses C. Assignment correctly 264 (quality index 98.2%). analysis, modal clusters (∆k 968.1 19.5 3, 140.3 4, 0.9 5); however, remained when more 4). At clusters, equal Of captured, had weighed, recordings taken subset Average lineage, sex, 1. highest largest lowest frequency. Univariate (Tables 1 5). slight, significant, males calling having shorter forearms, observed not Differences ranged 1.5 kHz (C females males) 28.8 (A females; PF FL M analyzing Our shared roosts deciduous transitional zones. colony encountered habitat harbored animals, although roosting much abae (H. Baldwin, unpublished data). corresponds Sudanian Senegal Morocco Yemen. Its occurrence tropics related narrow strip extension Dahomey Gap perhaps regarded southernmost point range. c. Kenya. Patterns mito-nuclear discordance frequently problematic relationships, potentially insufficient signal introns (Demos 2019) diversifying selection metabolic pathways, oxidative phosphorylation system encoded genes (Sunnucks 2017). contrast, strongly concordant patterns resolved thereby providing robust support described. Furthermore, given demonstrates isolation another, satisfies concept (Baker Bradley, 2006) biological (Dobzhansky Dobzhansky, 1937; Mayr, 1942). size provide phenetic support, groups, characters alone diagnostic. discriminated alone. helpful discriminating there overlap. reported Mt (51.1 ± 2.0, 4) very similar here, (9.1 0.4, =3; Other example skull measurements, require investigation. Significant divergences exist sufficient discriminate reasonable respect populations. variation Call roughly inversely correlated size, difference reflecting B); unlikely (Russo Small non-adaptive, arisen drift founder populations, contact re-established prevent interbreeding primary sense it subject strong pressures. parameters sister arise environmental preferences 2000), dietary niche partitioning (Heller 1989; Puechmaille 2011), social al.,. 2007, 2018), jamming avoidance 2007), possible explanations mutually exclusive. Given strength evidence, suggest should others shown, deep DNA, required assess delimitation illuminate extent (part 2020) unambiguously enough closely 3 South s.str. conservatively (morpho)species name ruber. interim, propose provisional cf. (see 2011 later). According our (B, D) fall lamottei, latter ranges seems quite probable includes taxon Moreover, none “caffer” (caffer 5–8) A). Labeling unique montane savannah habitats Mt. represented phylogenetically Nonetheless, knowledge various limited, avoiding splitting available Data retain signatures ancient splits confound incomplete sorting present (Nichols, see 5 Microsatellite enable measure contemporary finding lack sympatric thus implies area. supporting proposal represent provisionally designate caffer, investigation entire range will necessary understand full would pertinent know whether coronaviruses. information dispersal, population connectivity, behavior valuable predicting preventing humans, grateful Christian Drosten, Isaac Mawusi Adanyeguh, Lucinda Kirkpatrick, Anna Vogeler, Mac Elikem Nutsuakor, David Ofori Agyei, Sarah Koschnicke, Julia Morrison, Emmanual Asare, Eunice Okyere, Kenneth Quansah, Emmanuel Essoun, Paul Marfo, Kennedy Darkwa, Stefan Klose, Sandra Junglen, guides Justice, Paul, Kwame Takye, Winfred, Samuel Thomas Kruppa fieldwork logistical organization conduct. thank communities, community leaders, administrators staff Division, Kumasi Centre Collaborative Research Tropical Medicine. Thank you Paolo Momigliano Kirkpatrick discussion, Siobhan Dennison Kate Umbers comments manuscript. Funding provided DFG German Foundation (KA 1241/18-1; TS 81/10-1). Open Access funding enabled organized Projekt DEAL. S1. Sampling localities study. S2. List comparative final corresponding voucher specimen identification. S3. scores (bp) Please note: publisher responsible content functionality any supplied authors. Any queries (other missing content) directed author article.