A new microendemic gecko from the small forest fragments of south-eastern Madagascar (Squamata, Gekkonidae, Paragehyra) Author Belluardo, Francesco 0000-0002-3967-2686 EnviXLab, Department of Biosciences and Territory, University of Molise, Contrada Fonte Lappone s. n. c., 86090 Pesche, Italy & CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, InBIO Laboratório Associado, Universidade do Porto, Campus de Vairão, Rua Padre Armando Quintas, 4485 - 661 Vairão, Portugal & Departamento de Biologia, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre s / n, 4169 - 007 Porto, Portugal Author Piccoli, Costanza 0000-0002-1234-5706 CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, InBIO Laboratório Associado, Universidade do Porto, Campus de Vairão, Rua Padre Armando Quintas, 4485 - 661 Vairão, Portugal & Departamento de Biologia, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre s / n, 4169 - 007 Porto, Portugal Author Lobón-Rovira, Javier 0000-0003-4380-9427 CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, InBIO Laboratório Associado, Universidade do Porto, Campus de Vairão, Rua Padre Armando Quintas, 4485 - 661 Vairão, Portugal & Departamento de Biologia, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre s / n, 4169 - 007 Porto, Portugal Author Oliveira Alves, Ivo https://orcid.org/0009-0000-1187-4211 CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, InBIO Laboratório Associado, Universidade do Porto, Campus de Vairão, Rua Padre Armando Quintas, 4485 - 661 Vairão, Portugal & Departamento de Biologia, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre s / n, 4169 - 007 Porto, Portugal & BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairão, Rua Padre Armando Quintas, 4485 - 661 Vairão, Portugal Author Rasoazanany, Malalatiana Departamento de Biologia Animal, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749 - 016 Lisboa, Portugal Author Andreone, Franco 0000-0001-9809-5818 Mention Zoologie et Biodiversité Animale, Domaine Sciences et Technologieìiscussions, Université d’Antananarivo, B. P. 906, 101 Antananarivo, Madagascar Author Rosa, Gonçalo M. 0000-0002-8658-8436 Museo Regionale di Scienze Naturali, Via G. Giolitti 36, 10123 Torino, Italy & IMIB Biodiversity Research Institute (CSIC, Universidad de Oviedo, Principality of Asturias), c / Gonzalo Gutiérrez Quirós, 33600 Mieres, Spain & BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairão, Rua Padre Armando Quintas, 4485 - 661 Vairão, Portugal Author Crottini, Angelica 0000-0002-8505-3050 Institute of Zoology, Zoological Society of London, Outer Circle, Regent’s Park NW 1, 4 RY London, UK text ZooKeys 2025 2025-06-02 1240 1 38 journal article 310148 10.3897/zookeys.1240.151016 1c58645d-65d5-44a7-9a6c-796abedc47f0 07E50361-046B-4D17-8CF0-6BA4CD04DD73 Paragehyra tsaranoro sp. nov. Figs 3 , 5 , 6 , Tables 1 , 4 , Suppl. material 2 Note. The species has been reported as Paragehyra sp. aff. felicitae “ Tsaranoro ” CCS in Belluardo et al. (2021 a ). Type material. Holotype . Madagascar • 1 adult ♂ ; south-eastern Madagascar, Haute Matsiatra Region , Fianarantsoa province , ca 32 km south of the town of Ambalavao , Tsaranoro , Forêt Sacrée ; 22°04'57.65"S , 46°46'33.56"E ; 927 m a. s. l. ; 6 Dec. 2018 ; F. Belluardo , J. Lobón-Rovira , M. Rasoazanany leg.; boulder in semi-deciduous forest close to the forest edge ; GenBank: MZ 285407 ; ZSM 11/2023 : ACZCV 765 . Figs 3 A , 5 . Paratypes . Madagascar • 1 adult ♀ ; south-eastern Madagascar, Haute Matsiatra Region , Fianarantsoa province , western slopes of Andringitra National Park , Iantaranomby ; 22°07'42.60"S , 46°50'52.90"E ; 1,610 m a. s. l. ; 19 Nov. 2018 ; F. Belluardo , J. Lobón-Rovira , G. M. Rosa leg.; boulder in open environment with scattered palm trees ; GenBank: MZ 285406 , PV 389999 ; ZSM 10/2023 : ACZCV 600 . Fig. 3 . Madagascar • 1 juv. ; south-eastern Madagascar, Haute Matsiatra Region , Fianarantsoa province , Ambatomainty , Tsiomba ; 22°00'10.37"S , 46°47'16.08"E ; 960 m a. s. l. ; 8 Dec. 2018 ; F. Belluardo , J. Lobón-Rovira , M. Rasoazanany leg.; large boulders in open environment surrounded by scattered trees ; GenBank: MZ 285516 , PV 383222 ; ZSM 12/2023 : ACZCV 809 . Madagascar • 1 adult ♂ ; same data as preceding; 22°00'09.40"S , 46°47'16.22"E ; 947 m a. s. l. ; same data as preceding; same data as preceding; same data as preceding; GenBank: MZ 285517 , PV 383221 , PV 389997 ; UADBA uncatalogued: ACZCV 810 . Madagascar • 1 unsexed adult ; south-eastern Madagascar, Haute Matsiatra Region , Fianarantsoa province , Tsaranoro Forêt Sacrée ; 22°04'51.71"S , 46°46'36.88"E ; 909 m a. s. l. ; 6 Dec. 2018 ; F. Belluardo , J. Lobón-Rovira , M. Rasoazanany leg.; large boulders within semi-deciduous forest ; GenBank: MZ 285518 , PV 383219 , PV 389998 ; UADBA uncatalogued: ACZCV 771 . Madagascar • 1 unsexed adult ; same data as preceding; 22°05'05.03"S , 46°46'30.54"E ; 933 m a. s. l. ; same data as preceding; same data as preceding; same data as preceding; GenBank: MZ 285519 , PV 383220 , PV 390005 ; UADBA uncatalogued: ACZCV 770 . Madagascar • 1 unsexed adult ; same data as preceding; 22°05'07.58"S , 46°46'30.32"E ; 983 m a. s. l. ; same data as preceding; same data as preceding; same data as preceding; GenBank: MZ 285520 , MZ 285408 ; UADBA uncatalogued: ACZCV 779 . Type locality. Tsaranoro Forêt Sacrée (south-eastern Madagascar , Haute Matsiatra Region, Fianarantsoa province , ca 32 km south of the town of Ambalavao), 22°04'57.65"S , 46°46'33.56"E , 927 m a. s. l. A semi-deciduous forest fragment of ca 46 ha within Tsaranoro Valley Forest reserve (Figs 1 , 2 ). Diagnosis. The species is assigned to the genus Paragehyra based on genetic distances at the COI and 16 S markers, the presence of two diagnostic morphological derived characters of the genus relative to the other gekkonids: the asymmetrical relationship of the claw and toe-pad on digit I and the uniscansorial distal pad on digit I separated from enlarged basal scales by a series of smaller scales ( Nussbaum and Raxworthy 1994 ; Crottini et al. 2015 ). The inter-specific diagnosis is detailed in the following lines (Fig. 6 ; Tables 1 , 4 ). Paragehyra tsaranoro sp. nov. has 12 distinct longitudinal rows of enlarged tubercles on the body dorsolateral surface (character BT ), enlarged tubercles on the entire dorsal surface of hind limbs and on distal dorsal portions of forelimbs (character TDL ), enlarged tubercles organised in transversal rows that encircle tail dorsolateral surface (character TT ), 6-8 enlarged infralabial scales gradually decreasing in size in posterior direction (character I), lateral chin scales extending forward along each side, excluded from contact with first line of infralabials and first postmentals, in contact with second line of infralabials (character C), 3-6 ventral scales on the distal (tibial) segments of hind limbs enlarged into plates especially in the most distal part (character VET ), five or six large scales on preaxial ventral portion of pes between end of tibia and base of digit I (character SPP ), two or three small subdigital scales between enlarged basal scales and terminal distal pad on digit I of manus and pes (character SS ), 4-6 subdigital scales on claw-bearing segment of digits II-V of manus and pes with the distalmost scale markedly larger than the others (character SSC ), smooth dorsal scales on body surface smaller than ventral scales (character DO ), smooth unpigmented or poorly pigmented ventral scales on body surface (character VE ), smooth and pigmented subcaudal scales (character SC ), triangle-shaped mental scale (character ME ), large first postmental scales in contact for more than the 50 % of their length (character 1 PM ). Among species of the genus Paragehyra , P. felicitae is the most similar to P. tsaranoro sp. nov. , from which it differs in the following four morphological characters: TDL , enlarged tubercles present on the entire dorsal surface of hind limbs and only on distal dorsal segments of forelimbs (vs enlarged tubercles present on the entire dorsal surface of both hind limbs and forelimbs); C, lateral chin scales extend forward along each side, excluded from contact with first line of infralabials and first postmentals, in contact with second line of infralabials (vs lateral chin scales extend forward along each side, excluded from contact with first line of infralabials and first postmentals, in contact with second line of infralabials or lateral chin scales extend forward along each side, excluded from contact with first infralabials, in contact with first postmentals and second infralabials); SPP , five or six large scales on preaxial ventral portion of pes between end of tibia and base of digit I (vs generally 6-9 large scales on preaxial-ventral portion of pes between end of tibia and base of digit I); SSC , 4-6 subdigital scales on claw-bearing segment of digits II-V of manus and pes with the distalmost scale markedly larger than the others (vs 4-6 large and subequal subdigital scales on claw-bearing segment of digits II-V of manus and pes). Paragehyra tsaranoro sp. nov. differs from P. petiti in the following five characters: BT , 12 (vs 10) distinct longitudinal rows of enlarged tubercles on dorsolateral body surface; I, 6-8 enlarged infralabials that diminish gradually in size posteriorly (vs first 4-6 infralabials much larger than remainders); SSC , enlarged tubercles present on the entire dorsal surface of hind limbs and only on distal dorsal segments of forelimbs (vs 4-6 large and subequal subdigital scales on claw-bearing segment of digits II-V of manus and pes); SC , smooth and pigmented subcaudal scales (vs smooth and unpigmented subcaudal scales); ME , triangle-shaped mental scale (vs bell-shaped mental scale). Paragehyra tsaranoro sp. nov. differs from P. austini for the following eight characters: BT , 12 distinct longitudinal rows of enlarged tubercles on dorsolateral body surface (vs absence of any tubercle on dorsolateral body surface); TDL , enlarged tubercles present on the entire dorsal surface of hind limbs and only on distal dorsal segments of forelimbs (vs absence of enlarged tubercles on limbs dorsal surface); TT , transverse rows of enlarged tubercles that encircle tail dorsolateral surface (vs absence of tubercles on tail dorsolateral surface); VET , 3-6 ventral scales on the distal (tibial) segments of hind limbs enlarged into plates especially in the most distal part (vs ventral scales on distal (tibial) segment of hind limb slightly larger compared with scales on the proximal segments of hind limbs); SPP , five or six large scales on preaxial ventral portion of pes between end of tibia and base of digit I (vs generally 7 small scales on preaxial-ventral portion of pes between end of tibia and base of digit I); SS , two or three small subdigital scales between enlarged basal scales and terminal distal pad on digit I of manus and pes (vs 3 or 4 small subdigital scales between enlarged basal scales and terminal distal pad on digit I of manus and pes); DO , smooth dorsal scales on body surface smaller than ventral scales (vs smooth dorsal scales on body surface equal in size than ventrals); VE , smooth unpigmented or poorly pigmented ventral scales on body surface (vs smooth pigmented ventral scales on body surface). Paragehyra tsaranoro sp. nov. differs from P. gabriellae in the following 11 characters: BT , 12 distinct longitudinal rows of enlarged tubercles on dorsolateral body surface (vs presence of small tubercles not arranged in distinct rows); TDL , enlarged tubercles present on the entire dorsal surface of hind limbs and only on distal dorsal segments of forelimbs (vs absence of enlarged tubercles on limbs dorsal surface); TT , transverse rows of enlarged tubercles that encircle tail dorsolateral surface (vs presence of enlarged tubercles on tail not organised in transverse rows); C, lateral chin scales extend forward along each side, excluded from contact with first line of infralabials and first postmentals, in contact with second line of infralabials (vs lateral chin scales extend forward along each side in contact with first and second infralabials and with first postmentals or lateral chin scales excluded from contact with first infralabials, in contact with first postmentals and second infralabials); VET , 3-6 ventral scales on the distal (tibial) segments of hind limbs enlarged into plates especially in the most distal part (vs ventral scales on distal (tibial) segment of hind limb of normal size compared with scales on the proximal segments of hind limbs); SPP , five or six large scales on preaxial ventral portion of pes between end of tibia and base of digit I (vs 7-9 small scales along preaxial-ventral border of pes between end of tibia and base of digit I); SS , two or three small subdigital scales between enlarged basal scales and terminal distal pad on digit I of manus and pes (vs 7-10 small subdigital scales between enlarged basal scales and terminal distal pad on digit I of manus and pes); SSC , 4-6 subdigital scales on claw-bearing segment of digits II-V of manus and pes with the distalmost scale markedly larger than the others (vs numerous small scales increasing gradually in size distally); DO , smooth dorsal scales on body surface smaller than ventral scales (vs smooth dorsal scales on body surface equal in size than ventrals); VE , smooth unpigmented or poorly pigmented ventral scales on body surface (vs smooth pigmented ventral scales on body surface); 1 PM , large first postmental scales in contact for more than the 50 % of their length (vs large first postmental scales in contact for 50 % or less of their length). Description of the holotype . Figs 3 A , 5 , 6 ; Table 1 ; Suppl. material 2 . The holotype is an adult male in well-preserved condition with intact and original tail. Hemipenes are not everted. Digit III of the right pes was clipped and stored in 96 % ethanol as tissue sample for molecular analyses. The specimen has flattened body and head. Head width is slightly lower than body width ( 11.95 mm vs 13.65 mm ). The head snout is rounded, with HL and HW that are 0.32 and 0.21 × SVL , respectively. Head is 1.52 longer than wider ( HL / HW ), 3.3 × longer than deeper ( HL / HD ), and 2.17 larger than deeper ( HW / HD ). Eye with vertical pupil and ED that is 0.29 × relative to SnL and 0.66 × the IOD . Ear openings are elliptical in vertical direction with a diameter on the horizontal axis ( EO ) that is 0.33 × the ED and 5.05 mm distant from the eye ( ETD ). Forelimb and hind limb lengths are 0.32 and 0.48 × SVL , respectively. Forelimbs and hind limbs lengths are 0.72 and 1.08 × the AGD , respectively. Forelimb when extended forward reaches nostril, when extended posteriorly reaches three quarters of distance to groin, hind limb reaches anterior axilla. Tail length is 1.24 × SVL . The tail is subcylindrical and dorsoventrally flattened at its base with a pointed and narrow tip. A concave groove is present between nasal scales, which are not in direct contact, separated by one IN . Nostrils in contact with rostral, nasals, and four postnasals, not in contact with first supralabials. Quadrangular rostral scale less wide than mental scale with an incomplete, dorsal vertical groove extending downward approximately one-half the distance from dorsal edge to lip. Nine enlarged supralabial scales (mostly rectangular) are present on both right and left sides ( SL ), and eight and seven infralabial scales (I) are on the right and left sides, respectively. Infralabials gradually decrease in size posteriorly. The first six supralabials are squared and have equal size, while the last three scales are much smaller and elongated in posterior direction. Triangle-shaped ME laterally in contact with the first line of infralabials. Posterior to the mental scale are two large 1 PM with irregular pentagonal shapes and in reciprocal contact for more than 50 % of their length. The first postmentals are bordered posteriorly by a line of six smaller polygonal scales ( 2 PM , which are clearly distinguishable from the much smaller and mostly granular chin scales that are posterior to these) and in contact with the first line of I. Scalation of the contact between chin scales with postmentals and infralabials (C) is variable between left and right sides. On the left side, chin scales extending laterally are excluded from contact with the first line of infralabials and first postmentals and are in contact with the second line of elongated infralabials (defined as a distinguishable line of scales that are larger than chin scales and parallel to the first infralabials). On the right side, chin scales extending laterally are in contact with a third line of elongated infralabials but excluded from contact with the first and second lines of infralabials and first postmentals. Throat scales small, circular, and largely juxtaposed. Scales just below posterior infralabials enlarged. Throat covered by small granular scales. Scales on the dorsal body surface are mostly small granular with smooth surface, smaller than ventral scales (approximately half dimension on the horizontal axis). Dorsal enlarged body tubercles (~ three times bigger than scales) with rounded or subconical shape that are arranged into 12 distinct parallel longitudinal rows ( BT ), equally divided between the two body sides. Dorsal scales on limbs are mostly small with granular shape and smooth surface. Body tubercles extend on the entire surface of hind limbs and only on the distal portions of forelimbs ( TDL ). Hind limbs tubercles are larger than tubercles on forelimbs and larger than the dorsal tubercles organized in the 12 parallel longitudinal rows. A few scattered tubercles are also found on the posterior-lateral portion of the head and the neck. Dorsal scales on the tail are small and granular with smooth surface and of the same size as the dorsal scales on body. Eight dark whorls with one or two transverse rows (ring) of enlarged tubercles of each of the first five dark whorl are alternated until the tail tip with seven paler whorls with two transverse rows (rings) of enlarged tubercles, one at the anterior and one at the posterior border of each of the first four pale whorl. Enlarged tubercles on tail are larger than the dorsal tubercles organised in the 12 parallel longitudinal and the enlarged tubercles present of limbs. Whorl 1 and 2 have 12 enlarged tubercles, whorls 3 and 4 have six enlarged tubercles, whorls 5 and 6 have four enlarged tubercles, whorl 7 has three enlarged tubercles, whorl 8 has two enlarged tubercles, whorl 9 has a single enlarged tubercle. The two distalmost pale whorls are lighter than the previous ones. Dark whorls size increases in posterior direction. Pale whorl size increases in posterior direction until the fourth pale whorl, with the remaining three pale whorls gradually decreasing towards the tail tip. Transverse rows of enlarged tubercles with rounded or subconical shape encircle the tail dorsolateral surface ( TT ). Tubercle size gradually decreasing in posterior direction. Scales on the body ventral surface are mostly regular rhombus-shaped and juxtaposed with smooth surface ( VE ). They are unpigmented or poorly pigmented (pigments are visible only under stereo microscope). Slightly imbricate, cycloid scales begin behind throat and cover chest and belly. Ventral surface of forelimbs covered with granular scales on proximal segment and imbricate cycloid scales on distal segment. Hind limbs have mostly large polygonal and slightly imbricate scales with smooth surface and unpigmented to the naked eye, as described above for the ventral body surface. Some pale scale pigmentation is only visible on belly. Ventral surface of pelvis and thigh (proximal portion of hind limbs) covered with imbricate cycloid scales. Scales on the hind limbs distal (tibial) portion are larger than on the proximal portion. They gradually increase in size towards pes, and the three distalmost form large plates ( VET ) as large as half the limb width. Subcaudal scales are enlarged transversely, smooth, and pigmented, especially in their anterior and posterior borders ( SC ). In the distalmost part of the tail, they tend to be pigmented also in their central portions and follow the alternated dark and pale colour pattern of the dorsal whorls described above. In the proximal portion of the tail, there are three central longitudinal rows of large imbricate and mostly cycloid scales and, starting from the section corresponding to the second dark dorsal whorl and in posterior direction, they form a unique longitudinal row of imbricate undivided plates almost as large as the tail width. These central longitudinal rows are laterally bordered on both sides by 1 to 3 longitudinal rows of smaller imbricate cycloid scales. Scales immediately adjacent to cloacal opening much smaller than surrounding scales. Precloacal or femoral pores are absent. Preaxial border of palm and digit I of manus covered with large scales. The pes palm is covered with mostly granular scales in their central part and relatively large cycloid imbricate scales on the lateral portions. In particular, on the pes preaxial ventral portion between the end of the tibia and the base of digit I there is a longitudinal row of five (right) and six (left) large imbricate cycloid scales with smooth surface on pes ( SPP ). Subdigital scalation of digit I of manus and pes organised into a basal longitudinal row of relatively large imbricate scales on the proximal portion, followed by two intermediate parallel rows of smaller juxtaposed scales (composed of three scales on left manus and pes; SS ), and a distal enlarged rectangular pad with an extremely small claw (only visible with a stereo microscope) on top of it. Mostly imbricate cycloid scales on basal portions of digits II-V of manus and pes followed by rows of undivided and transversely enlarged pads covering the distal two-thirds of digits. Lamellae cover the most distal pads of digits II-V with the following numbers on the left side: 6, 7, 7, 7 pads (manus), 8, 8, 7, 7 pads (pes) ( SLMP ). Claw-bearing segments on top of the subdigital lamellae composed of relatively large imbricate rectangular scales organised into longitudinal rows with some degree of overlap between them. In the central row, the distalmost scale is markedly larger than the others and composed of the following scale numbers on left digits II-V: 5, 5, 5, 5 (manus) and 5, 6, 5, 5 (pes) ( SSC ). Comparative finger and toe length in manus is 1 <2 <5 <3 <4, in pes 1 <2 <3 <4 <5. Colouration. Colouration after six years in ethanol is slightly paler but with an identical pattern to that at the time of collection (Figs 3 A , 5 , 6 ). Dorsolateral surface of the head, body, limbs, and tail with a pale grey ground colour and dark brownish transversal bands and blotches. Head with large posterior blotches and scattered dark brownish dots and linear marks until the mouth tip. A dark brownish line connects ear openings to the eyes posteriorly and reaches half of the distance between the eye anterior parts and nostrils, on both sides. Brownish longitudinal vertebrate line on the dorsal body surface. Five transversal dorsolateral dark brownish blotches on the dorsal body surface with a paler greyish spot in their centre. Hind limbs with transverse bands that are more defined in their distal (tibial) rather than proximal portions, where they can be more similar to blotches. Forelimbs with less defined transverse bands with interspersed brownish blotches. Dorsal surface of digits with two pale grey and two dark brownish bands alternated from the base until the tip; in life colouration, the second pale band towards the distal portion is white. In the preserved specimen the tail has alternated dorsolateral pale grey and dark brownish transverse bands. The ventral surface of the head, body, limbs, and tail is uniformly pale brownish tending to whitish, for the presence of a sparse dotted dark pigmentation visible only under stereo microscope (character VE ; Fig. 6 ). Ventral (subdigital) surface of digits is slightly darker than ground colour, and lamellae are white. Subcaudal scales borders are slightly darker than the ground colour. The distal part of the tail (ca one-third of the entire length) has alternated dark brownish and ground colour bands, and the tip is black. Supralabials and infralabials with alternated whitish and dark brownish vertical bands, with more intense dark colouration on supralabials (Fig. 6 ). Variation. Variation in morphological characters of ZSM paratypes is reported in Table 1 and Suppl. material 2 (see also Fig. 3 ). After six years in ethanol, colouration of the paratypes ZSM 10/2023 ( ACZCV 600 ) and ZSM 12/2023 ( ACZCV 809 ) is slightly paler to that in life and overall similar to the holotype . Paratype ZSM 10/2023 ( ACZCV 600 ) differs from the holotype in the following characters: nasal scales are in contact, 11 enlarged left supralabials ( SL ), six left I, eight 2 PM , scalation of chin scales identical between right and left sides with lateral chin scales extending forward excluded from contact with first line of infralabials and first postmentals but in contact with second line of infralabials (C), six plates on ventral side of distal (tibial) segments of hind limbs ( VET ), and two scales compose the intermediate longitudinal rows between enlarged basal scales and distal pad of digit I of pes ( SS ). Refer to Table 1 and Suppl. material 2 for the variation in morphological measurements and in SSC and SLMP . The head dorsal surface is slightly darker than in the holotype , with a characteristic pale W-shaped blotch posterior to the eyes and two central pale spots anterior to the eyes. The specimen does not have the tail, and the third digit of the right pes was clipped as tissue sample. Paratype ZSM 12/2023 ( ACZCV 809 ), an unsexed juvenile, differs from the holotype in the following characters: nasal scales are in contact ( IN ), 10 enlarged left SL , eight right I, modified bell-shaped ME , eight second 2 PM , scalation of chin scales identical between right and left sides with lateral chin scales extending forward excluded from contact with first line of infralabials and first postmentals but in contact with second line of infralabials (C), four plates on ventral side of distal (tibial) segments of hind limbs ( VET ), six scales on preaxial-ventral portion of right pes between end of tibia and base of digit I ( SPP ), and two scales make the intermediate longitudinal rows between enlarged basal scales and distal pad of digit I of manus ( SS ). Refer to Table 1 and Suppl. material 2 for the variation in morphological measurements and in the number of subdigital scales and subdigital lamellae on claw-bearing segments of digits II-V ( SSC and SLMP , respectively). Morphological variation and measurements are not available for the paratypes hosted in the UADBA collection: UADBA uncatalogued ( ACZCV 810 ), UADBA uncatalogued ( ACZCV 771 ), UADBA uncatalogued ( ACZCV 770 ), and UADBA uncatalogued ( ACZCV 779 ). Etymology. The specific epithet derives from the type locality Tsaranoro. The name is used as an invariable noun in apposition to the generic name. Distribution. Paragehyra tsaranoro sp. nov. is currently known from south-eastern Madagascar , restricted to three localities within the western part of the region surrounding the Andringitra Massif (Haute Matsiatra administrative region, Fianarantsoa province ): Tsaranoro, Ambatomainty, and Iantaranomby (Suppl. material 1 ; Figs 1 , 2 ). Iantaranomby is on the western slope of the Andringitra Massif, while Tsaranoro and Ambatomainty forests are in the plateau located west to the Massif, laying at the foot of two granitic domes rising a few hundred meters relative to ground elevation. The species has been sampled within an elevation range between 897 m (Ambatomainty) and 1,610 m a. s. l. (Iantaranomby). Habitat and behaviour. The holotype ZSM 11/2023 ( ACZCV 765 ) was found at night (~ 8 p. m.) in the semi-deciduous forest fragment of Tsaranoro under clear weather. The animal was resting on a boulder near the forest edge. Nearby, we sampled an individual of Paroedura rennerae Miralles, Bruy, Crottini, Rakotoarison, Ratsoavina, Scherz, Schmidt, Köhler, Glaw & Vences, 2021 ( Belluardo et al. 2021 a ). The remaining individuals of P. tsaranoro sp. nov. from Tsaranoro forest were found on boulders. Paratype UADBA uncatalogued ( ACZCV 771 ) was collected outside the forest patch along a small stream amidst arboreal riverine vegetation. The other two paratypes , UADBA uncatalogued ( ACZCV 770 and ACZCV 779 ), were found within the forest interior, along with additional individuals not collected as vouchers (Suppl. material 1 ). All Tsaranoro animals were found at night, except for ACZC 10973, collected in the morning (~ 9 a. m.). Paratype ZSM 10/2023 ( ACZCV 600 ) was the only individual collected within Andringitra NP (Iantaranomby, ca 400 m from the park limit; Fig. 2 C ). This animal was found at night in clear weather under a large boulder on a sloping trail surrounded by scattered palm trees. The three individuals from Ambatomainty Tsiomba (including the paratypes UADBA uncatalogued ( ACZCV 810 ) and ZSM 12/2023 ( ACZCV 809 )) were all found outside the small forest patch (Fig. 2 A ). These animals were sampled at night in clear weather on large boulders in open areas near the forest fragment surrounded by scattered trees. Overall, P. tsaranoro sp. nov. seems to be associated with semi-deciduous forest. The species is found both within the forest interior and along its edges, as well as outside the forest or along water streams, but always on boulders associated with arboreal vegetation. The isolated forest fragments in the study area are characterised by numerous boulders, many of which are ancient Betsileo tombs ( Gould and Andrianomena 2015 ). This cultural significance has led to these areas being referred to as “ Forêts sacrées ”, i. e., “ sacred forests ”. The boulders vary in size, sometimes forming large agglomerates, with some reaching considerable dimensions. The higher number of individuals found in the interior of Tsaranoro forest compared to other sites suggests that relatively larger and more mature forests may support higher population densities of this species. Proposed conservation status. The extent of occurrence ( EOO ) and area of occupancy ( AOO ) total 38.8 km 2 and 16 km 2 , respectively (computed with GeoCAT ( Bachman et al. 2011 ) using squared grid cells of 2 × 2 km for AOO ). Following IUCN Red List guidelines ( IUCN Standards and Petitions Committee 2024 ), we propose to consider P. tsaranoro sp. nov. as Critically Endangered ( CR ) under criterium B 1 ab (iii). The proposed evaluation of the species conservation status is justified by the narrow distribution ( EOO < 100 km 2 ), its occurrence at three severely isolated localities, and the continuing decline in the extent and quality of its habitat due to specific threats. The habitat of this species is severely fragmented, with the three known localities isolated by unsuitable landscapes of villages, rice fields, and pastures. In two of these localities, forests areas and arboreal vegetation are heavily degraded. The Ambatomainty forest is extremely small (ca 2 ha) with almost no canopy cover. In Iantaranomby (Andringitra NP), the single specimen encountered was found in an area with only a few scattered palm trees. Unlike Ambatomainty and Iantaranomby, Tsaranoro forest provides particularly favourable conditions for the species, featuring well-structured vegetation and abundant boulders. Tsaranoro is also the largest forest fragment in the region (46 ha; Gould and Andrianomena 2015 ; Gould and Gabriel 2015 ), likely contributing to the higher abundance of the species observed there. Iantaranomby is the only locality with formal legal protection, located within Andringitra NP borders. No legal protection or local management is known for Ambatomainty. Tsaranoro forest is managed by the local community Association Tantely since 2002. Although in the past the forest has been subject to selective logging ( Gabriel et al. 2018 ), reforestation projects consisting in the establishment of tree nurseries have been recently funded ( Gould and Andrianomena 2015 ). However, and apart from Tsaranoro, deforestation in the region has been particularly intense and seems to be a recent and still ongoing process. Most deforestation occurred in the last 60 years and until 20 years ago there were still remnant parts of continuous forests ( Gould and Andrianomena 2015 ; Gould and Gabriel 2015 ). The process of deforestation and forest degradation in the region is continuing through direct tree logging (normally used as firewood) and fires that are normally used to clean areas for cattle grazing and that sometimes go out of control and affect the remaining forest fragments ( Crottini et al. 2015 ; Gould and Andrianomena 2015 ; Gould and Gabriel 2015 ). We could directly observe these activities in some of these fragments, even within the legally protected area of Andringitra NP. Ongoing deforestation, forest degradation, and fragmentation seem to be the main threats to the conservation of the small forest fragments where P. tsaranoro sp. nov. lives and their persistence is expected to have an impact on this newly described gecko species. Updated distribution and proposed conservation status of P. felicitae . When P. felicitae was formally described, the species was only known from its type locality within Anja reserve and a few boulders located ca 1 km away, on the opposite side of the national road RN 7 relative to the reserve (Fig. 2 F ; Suppl. material 1 ; Crottini et al. 2015 ). Due to the limited knowledge of the species distribution within the areas surrounding Anja and following IUCN Red List guidelines, Crottini et al. (2015) proposed to consider the species as Data Deficient. During the herpetological survey conducted by Belluardo et al. (2021 a ) in several localities of the Andringitra region, including several areas surrounding Anja, the species was confirmed within Anja reserve and a few individuals were found in the forest of Sakaviro reserve (Fig. 2 E ; Suppl. material 1 ), located ca 8 km north of Anja (Fig. 1 ). The semi-deciduous forest in Sakaviro is smaller than in Anja (14 vs 36 ha, Gould and Andrianomena 2015 ), and host similar environmental characteristics, being located at the base of a small granitic peak and with the presence of several large granitic boulders. Species elevation ranges between 950 m (Anja) and 1,089 m a. s. l. (Sakaviro). With the improved knowledge on the species distribution in the region surrounding the type locality, it is now possible to propose a new conservation status for P. felicitae . The updated EOO and AOO total 5 km 2 and 16 km 2 , respectively. The EOO has been adjusted to 16 km 2 as it cannot be smaller than AOO ( IUCN Standards and Petitions Committee 2024 ). Although smaller grid cells might produce lower and more realistic AOO extents, we followed IUCN Red List guidelines in using 2 × 2 km cells. We propose to evaluate this species as CR under criterium B 1 ab (iii) based on the narrow distribution ( EOO < 100 km 2 ), the occurrence at three isolated threat-defined locations (Anja reserve, the few boulders located at ca 1 km away from Anja and Sakaviro reserve), and the continuing decline in the extent and quality of its habitat due to persistent deforestation and forest degradation in the surrounding areas. Similar to P. tsaranoro sp. nov. , P. felicitae seems to be associated with granitic boulders within or close to semi-deciduous forest ( Crottini et al. 2015 ). Despite their limited extensions, both Anja and Sakaviro forests seem to host suitable habitats and the species was found to be relatively abundant in both localities. Anja and Sakaviro forests are managed by local community associations: Anja Miray (since 2000) and Sakaviro Miray (since 2012), respectively ( Gould and Andrianomena 2015 ). Anja forest has been subject to selective logging and introduction of non-native trees before the community-managed reserve became established ( Gabriel et al. 2018 ). Some human disturbance at the forest edge (i. e., tree logging and cattle grazing) has been observed ( Crottini et al. 2015 ), suggesting that some level of forest use for livelihood resource remains, as expected in community-managed forests ( Raik 2007 ; Nopper et al. 2017 ). The combination between severe habitat fragmentation and continued widespread deforestation within the region surrounding Anja and Sakaviro and some level of human disturbance in these forests ( Crottini et al. 2015 ; Gould and Andrianomena 2015 ; Gould and Gabriel 2015 ) represent the main threats to P. felicitae and might seriously impact the viability of local populations causing the long-term decline of the species.