An inordinate fondness for inconspicuous brown frogs: integration of phylogenomics, archival DNA analysis, morphology, and bioacoustics yields 24 new taxa in the subgenus Brygoomantis (genus Mantidactylus) from Madagascar Scherz, Mark D. Crottini, Angelica Hutter, Carl R. Hildenbrand, Andrea Andreone, Franco Fulgence, Thio Rosin Köhler, Gunther Ndriantsoa, Serge Herilala Ohler, Annemarie Preick, Michaela Rakotoarison, Andolalao Rancilhac, Loïs Raselimanana, Achille P. Riemann, Jana C. Rödel, Mark-Oliver Rosa, Gonçalo M. Streicher, Jeffrey W. Vieites, David R. Köhler, Jörn Hofreiter, Michael Glaw, Frank Vences, Miguel Megataxa 2022 2022-12-15 7 2 113 311 Scherz & Crottini & Hutter & Hildenbrand & Andreone & Fulgence & Köhler & Ndriantsoa & Ohler & Preick & Rakotoarison & Rancilhac & Raselimanana & Riemann & Rödel & Rosa & Streicher & Vieites & Köhler & Hofreiter & Glaw & Vences, 2022 Scherz & Crottini & Hutter & Hildenbrand & Andreone & Fulgence & Köhler & Ndriantsoa & Ohler & Preick & Rakotoarison & Rancilhac & Raselimanana & Riemann & Rödel & Rosa & Streicher & Vieites & Köhler & Hofreiter & Glaw & Vences 2022 [136,562,1803,1829] Amphibia Mantellidae Mantidactylus Animalia Anura 72 185 Chordata subSpecies ambohimitombi subsp. nov. miloko   Identity and justification.—This lineage of the  M. curtusclade was considered as confirmed candidate species M.sp. 18 by Vieites et al.(2009)due to its mitochondrial divergence in concert with slight but distinct differences in colour pattern, and as M.sp. Ca18 by Perl et al.(2014). It was depicted as ‘  Mantidactylussp. aff. curtus“Ambohitantely”’ by Glaw and Vences (2007). We here consider this lineage provisionally as a subspecies of  M. ambohimitombibased on the following rationale: (i) it belongs to the same general mitochondrial lineage as the nominal subspecies in the 16S tree, and the two are also closely related in the phylogenomic tree; (ii) morphologically, the two lineages are similar to each other, except for a somewhat more distinct dorsal pattern in M. a. miloko ssp. nov.; (iii) based on two samples in our mitochondrial tree, RJS 1877 and ACZC 4254, a very similar mitochondrial haplotype to the one from Ambohitantely also occurs at Ankaratra, suggesting the possibility of past or ongoing gene flow between localities or mitochondrial introgression;and (iv)based on sequences of specimens APR 10803, APR 10638, and APR 10663, the lineage also occurs in Angavokely and Ankazomivady, thus rather close to Ankaratra, suggesting the two lineages may be parapatric and could hypothetically have a hybrid zone. Since the status of the Ambohitantely population as a fully isolated evolutionary lineage is thus not fully verifiable with the data at hand, we consider the status as a subspecies of  M. ambohimitombito be adequate.     Holotype.— ZSM 219/2005(FGZC 2143), adult male, collected by M. Vences, L. du Preez, P. Bora, L. Raharivololoniaina, R.D. Randrianiaina, T. Razafindraibe, E. Randriamitsoon  18 January 2005at Ambohitantely Special Reserve, ‘Jardin Botanique’, at a site about 500 m fromthegeographicalcoordinates 18.1725°S, 047.2768°E,  1580 ma.s.l., Analamanga Region, Madagascar.A 16S barcode sequence of the holotype was obtained in this study and was included in the analysis.  Paratype.—A single  paratype: ZSM 237/2005(FGZC 2172), adult female, with the same collection data as holotype.  Additional material.—The following specimens probably belong to this taxon but are only assigned tentatively and not designated as paratypes because no molecular data is available:  ZMA 6859( two specimenswith field numbers 675 and 676) and ZMA 6860 ( six specimenswith field numbers 910–914, 917), collected by R.M.A. Blommers-Schl̂sser in 1972 at 1500 m ( ZMA 6859) and  2200 ma.s.l. ( ZMA 6860) at Tampoketsa d’Ankazobe.   Diagnosis.— Mantidactylus ambohimitombi milokois a member of the  M. curtusclade, and the sister group of M. a. ambohimitombi+ M. a. marefo.Morphologically, it is very similar to M. a. ambohimitombi. See Table 4for a list of diagnostic morphological characters. The combination of relatively large body size of up to 49 mm, slightly granular skin without clearly defined dorsolateral ridges, and small tympanum diameter of a maximum of 10% of SVL in males, distinguishes M. a. miloko ssp. nov.from species of the other clades. Within the  M. curtusclade,  M. alutus, M. madecassusand  M. paulianihave smaller body sizes and are distinguished by either a usually shorter snout (  M. madecassus, M. pauliani, M. a. marefo), or presence of rather distinct dorsolateral ridges (  M. alutus) ( Table 4);  M. curtususually has a smoother skin and a somewhat shorter snout;  M. bourgatiis morphologically very similar but appears to occur only on the Andringitra Massif. Compared to other subspecies of  M. ambohimitombi, the new subspecies differs from M. a. marefoby its distinct dorsal pattern, lack of bluish colour around the eye, and more pointed snout; and from M. a. ambohimitombiby the usually more distinct dorsal pattern. For detailed distinction from other new species and subspecies described herein, see the respective species accounts. A full list of molecular diagnostic sites in the 16S gene of M. a. milokoin pairwise comparisons to all other  Brygoomantisspecies and subspecies is provided as Supplementary appendix.   Description of the holotype.—Adult male in good state of preservation ( Fig. 9). Tissue sample taken ventrally from right thigh. Femoral gland partly detached to examine their structure internally. Body rather slender. Head slightly wider than body. Snout rounded in dorsal and lateral view. Nostrils directed dorsolaterally, slightly protuberant. Nostrils nearer to tip of the snout than to eye. Canthus rostralis almost not recognisable, slightly concave. Loreal region slightly concave. Tympanum distinct, elliptical, wider than high, its diameter 78% of eye diameter. Supratympanic fold distinct, beginning straight, with a distinct, angular 90° bend at the posterior edge of tympanum towards insertion of forelimb. Tongue ovoid, distinctly posteriorly bifid. Maxillary teeth present. Vomerine teeth distinct in rounded aggregations, positioned posterolateral to choanae. Choanae rounded. Subarticular tubercles single. Outer metacarpal tubercle present, inner metacarpal tubercle present. Fingers without webbing. Relative length of fingers: I=II<IV<III. Finger discs minimally enlarged. Nuptial pads absent. Foot longer than tibia (113%). Lateral metatarsalia separated. Inner metatarsal tubercle present. Outer metatarsal tubercle small but recognisable. Webbing formula: 1(0.25), 2i(1), 2e(0.5), 3i(1.5), 3e(1), 4i(2), 4e(2), 5(0.75). Relative length of toes: I<II<V=III<IV. Skin on the upper surface smooth, with some granules on flanks, and some longitudinal tubercles forming interrupted and weakly expressed dorsolateral folds.Ventral side smooth. Femoral glands distinct but relatively small, with a distal ulcerous macrogland consisting of at least eight large granules and an external central depression, and a weakly expressed proximal granular gland field visible in internal view. Colour in preservative: light brown dorsally with large and contrasted dark brown patches all over the dorsal surface, distinct dark crossbands on limbs, and a light frenal stripe with some dark markings on the upper lip. Ventrally uniformly light grey with alternating dark-light pattern ventrally on the lower lip. Colour in life similar to that in preservative, but more contrasted ( Fig. 16).  Variation.—Variation in measurements is given in Table 5. Too few specimens have been sexed to assess the degree of sexual size dimorphism. Femoral glands distinct and large (but not differing in colour from surrounding ventral skin of thigh) in the male holotype( Fig. 16). In contrast, specimens from the ZMA collection are difficult to sex externally as femoral glands are often indistinct, as seems to be typical for several taxa in the  M. curtusclade, possibly due to seasonal effects.  Natural history.—Specimens were collected around clean highland streams running in open areas between forest fragments. Specimens at Angavokely and Ankazomivady that appear to belong to this taxon based on mitochondrial DNA were found on wet rocks along slow-moving parts of streams.  Calls.—The call of this subspecies has not been recorded.  Tadpoles.—Probably reported from Angavokely by Blommers-Schl̂sser (1979)..   Distribution.—Endemic to the central highlands of Madagascar, north of the distribution of the nominal subspecies ( Fig. 7). This subspecies is currently known from Ambohitantely, and probably also from Angavokely, and Ankazomivady. A mitochondrial haplotype corresponding to this subspecies has also been detected at Ankaratra, but due to the limited information on the sampling event and the absence of specimens (only tissue sampels were collected) this record (which might also represent mitochondrial introgression) requires confirmation. Elevation range: 1520–1735 ma.s.l.   Etymology.—The subspecies name is derived from Malagasy word miloko, meaning ‘painted’, referring to the rather distinct dorsal pattern of well-delimited dark blotches characterizing this subspecies. The subspecies name is used as a noun in apposition. 4010830427 2005-01-18 ZSM M. Vences; L. du Preez; P. Bora; L. Raharivololoniaina; R. D. Randrianiaina; T. Razafindraibe & E. Randriamitso Madagascar 1580 -18.1725 Ambohitantely Special Reserve 7 47.2768 Analamanga Region 72 185 2143 1 holotype 4010830342 ZSM M. Vences; L. du Preez; P. Bora; L. Raharivololoniaina; R. D. Randrianiaina; T. Razafindraibe & E. Randriamitso 72 185 237/2005 1 paratype 4010830371 ZMA R. M. A. Blommers-Schl 1500 Tampoketsa d'Ankazobe 72 185 6859 2 4010830469 ZMA R. M. A. Blommers-Schl 2200 Tampoketsa d'Ankazobe 72 185 6860 6