Lost and Found: Taxonomic revision of the speckled skink (Oligosoma infrapunctatum; Reptilia; Scincidae) species complex from New Zealand reveals a potential cryptic extinction, resurrection of two species, and description of three new species Author Melzer, Sabine Author Hitchmough, Rod A. Author Bell, Trent Author Chapple, David G. Author Patterson, Geoff B. text Zootaxa 2019 2019-06-26 4623 3 441 484 journal article 26376 10.11646/zootaxa.4623.3.2 b753c71a-aad2-4e6e-9e62-c80348c2cff2 1175-5326 3258580 402E9C29-3A36-4A0A-AE8D-5D7B0C335701 Oligosoma newmani Wells & Wellington 1985 Figures 5a, b Synonyms Lygosoma infrapunctatum WERNER 1895 ; ( Leiolopisma : Boulenger ): SMITH 1937 ; McCANN 1955 Leiolopisma infrapunctata (Boulenger): MITTLEMAN 1952 Leiolopisma infrapunctatum (Boulenger): McCANN 1955 ; (Boulenger): McCANN 1956 ; (Boulenger): SHARELL 1966 ; (Boulenger): WHITAKER 1968 ; (Boulenger): MEADS 1970 ; (Boulenger): SOPER 1970 ; (Boulenger): WHITAKER 1970; (Boulenger): TOWNS 1971 ; (Boulenger): SCHIPPER 1972 ; (Boulenger): TOWNS & HAYWARD 1973 ; (Boulenger): ANDREWS 1974 ; (Boulenger): GREER 1974 ; (Boulenger): ROBB 1974 ; (Boulenger): BULL & WHITAKER 1975 ; (Boulenger): GUNDY & WURST 1976 ; (Boulenger): WHITAKER 1976 ; (Boulenger): HARDY 1977 ; (Boulenger): ROBB 1980 & 1986 ; (Boulenger): SCHMIDT 1980 ; (Boulenger): ALLISON 1982 ; (Boulenger): AINSWORTH 1985 ; (Boulenger): ALLISON & BLAIR 1987 ; (Boulenger): GOFF et al. 1987 ; (Boulenger): WORTHY 1991 ; (Boulenger): EAST et al. 1995 ; (Boulenger): MARKWELL 1995 ; (Boulenger): EFFORD et al. 1997 . Oligosoma newmani WELLS & WELLINGTON 1985 Oligosoma infrapunctatum DUNCAN 1999 ; EFFORD et al. 2001 ; HEAP et al. 2003 ; STEPHENS 2004 ; GREAVES et al. 2008 ; CHAPPLE et al 2009; JEWELL & MORRIS 2008 & 2011 ; TOWNS et al. 2002 ; BELL 2014 ; DUMONT 2015 ; MOCKETT et al. 2016; NELSON et al. 2016 ; VAN WINKEL et al. 2018 . Holotype . Stephens Island ( Takapourewa ) ( 40º 40’S , 174º 00’E ), NMNZ RE004481 (coll. G. Woodward , Jan 1963 ). Paratypes ( 7 specimens ). Stephens Island ( Takapourewa ) ( 40º 40’S , 174º 00’E ), NMNZ RE005407, female (coll. B. Bell , 28 Jan 1984 ) ; Stephens Island ( Takapourewa ) ( 40º 40’S , 174º 00’E ), 3 specimens ( NMNZ RE003697 [S481], male; NMNZ RE003697 [S52], female; NMNZ RE003697 [S612], female) (coll. B. Parker , 04 Oct 1963 ) ; Stephens Island ( Takapourewa ) ( 40º 40’S , 174º 00’E ), NMNZ RE001066 [S148], male) (coll. R. Tillyard , Jan 1921 ) ; Mt Bruce (originally Stephens Island ( Takapourewa )) ( 40º 40’S , 174º 00’E ), NMNZ RE005242, female) (coll. Unknown , unknown date) ; Stephens Island ( Takapourewa ) ( 40º 40’S , 174º 00’E ), NMNZ RE001693 [2/2/4.1], male) (coll. G. Walls , unknown date) . Other specimens examined ( 43 specimens ). St Arnaud ( 41º 49’S , 172º 51’E ), NMNZ RE005410 (organs removed) (coll. R. Hitchmough, 31 Mar 1991 ); Hokitika ( 42º 43’S , 171º 00’E ), NMNZ RE005440, male (coll. Laing May 1996 ) ; Hokitika ( 42º 43’S , 171º 00’E ), NMNZ RE005416, female (coll. J. Lyall , 01 Jan 1996 ) ; Cobden ( 42º 26’S , 171º 12’E ), 2 specimens : NMNZ RE006087, male ; NMNZ RE006088 (organs removed) (coll. J. Lyall , 10 Sep 1995 ) ; Hokitika Cemetery ( 40º 43’S , 171º 00’E ), 3 specimens : NMNZ RE005334, female ; NMNZ RE005335, male ; NMNZ RE005336, female (coll. R. van Mierlo , P. van Klink , 30 Oct 1997 ) ; Birchfield Beach ( 41º 40’S , 171º 49’E ), 2 specimens : NMNZ RE005352, male ; NMNZ RE005355, male (coll. R. van Mierlo , P. van Klink , 06 Oct 1997 ) ; Karamea , Oparara Road ( 41º 14’S , 172º 08’E ), 3 specimens : NMNZ RE005361, female ; NMNZ RE005362, male ; NMNZ RE005363, male (coll. R. van Mierlo , P. van Klink , 12 Jan 1998 ) ; Costello Hill , Pakihi Charleston ( 41º 53’S , 171º 29’E ), NMNZ RE005389, female (coll. R. van Mierlo , P. van Klink , 05 Mar 1998 ) ; Kaniere Rd , Hokitika ( 42º 44’S , 171º 00’E ), 5 specimens : NMNZ RE005393, male ; NMNZ RE005394, female NMNZ RE005395, female ; NMNZ RE005396, female ; NMNZ RE005397, female (coll. R. van Mierlo , P. van Klink , 31 Mar 1998 ) ; Orowaiti Lagoon ( 41º 45’S , 171º 37’E ), 3 specimens : NMNZ RE005503, female ; NMNZ RE005504, male ; NMNZ RE005505, male (coll. R. van Mierlo , P. van Klink , 08 Apr 1998 ) ; St Arnaud ( 41º 46’S , 172º 44’E ), 5 specimens : NMNZ RE003891 [S249], male ; NMNZ RE003892 [S250], female ; NMNZ RE003895 [S253], female ; NMNZ RE003899 [S257], male ; NMNZ RE003893 [S251], male (coll. B. Thomas , 02 Apr 1968 ) ; Paparoa ( 42º 06’S , 171º 27’E ) , NMNZ RE005258, male (coll. Unknown , unknown date); Denniston ( 41º 44’S , 171º 48’E ) , NMNZ RE006210, female (coll. Unknown , 09 Oct 2001 ) ; Cobden ( 42º 26’S , 171º 12’E ), 9 specimens : NMNZ RE005340, female ; NMNZ RE005342, female ; NMNZ RE005343, female ; NMNZ RE005344, male ; NMNZ RE005346, female ; NMNZ RE005347, male ; NMNZ RE005348, male ; NMNZ RE005349, female ; NMNZ RE005351, female (coll. R. van Mierlo , L. Adams , 19 Oct 1997 ) ; Granity , West Coast ( 41º 38’S , 171º 51’E ), 5 specimens : NMNZ RE007401, male ; NMNZ RE007402, female ; NMNZ RE007403, male ; NMNZ RE007404, male ; NMNZ RE007405, female (coll. T. Bell , January 2015 ) . FIGURE 5. a) O. newmani holotype, Stephens Island (Takapourewa) (photo: Jean-Claude Stahl). b) Live specimen of O. newmani , Serpentine Creek (photo: Marieke Lettink). Diagnosis. O. newmani can be distinguished from other species in the O. infrapunctatum species-complex by a combination of characters ( Figure 4 a–j). There are statistical differences between O. newmani and O. salmo sp. nov. (S-Ear/EF, VS, upper ciliaries). Compared with O. salmo sp. nov. MS is usually 33 or below whereas O. salmo is mostly 33 or above. There are statistical differences from O. robinsoni (AG/SF). Compared with O. robinsoni subdigital lamellae are usually 20 or below whereas usually 20 or above in O. robinsoni . O. newmani has a shorter tail relative to SVL than O. robinsoni . Compared with O. albornense sp. nov. nuchal pairs are usually 3 or below versus 3 or above ( O. albornense ); usually 69 or more VS ( O. newmani ) versus 69 or fewer VS ( O. albornense sp. nov. ). HL/HW is always 1.7 or below in O. albornense sp. nov. whereas it is usually 1.7 or above in O. newmani . O. auroraensis sp. nov. usually has more than 20 subdigital lamellae on fourth hind toe versus 20 or fewer for O. newmani . The mean TL/SVL in O. auroraensis sp. nov. is 1.38 compared with 1.22. There is also a significantly higher VS count in O. auroraensis sp. nov. compared with O. newmani . Description of Holotype . Habit lacertiform, body elongate, oval in cross-section; limbs well developed, pentadactyl. Lower eyelid with a transparent palpebral disc, bordered on sides and below by small, oblong granules. Snout moderately blunt. Nostril centred in lower middle of nasal, not touching bottom edge of nasal, pointing up and back. Supranasals absent. Rostral broader than deep. Frontonasal broader than long, not separated from frontal by prefrontals meeting in midline. Frontal longer than broad, shorter than frontoparietal and interparietal together, in contact with 2 anteriormost supraoculars. Supraoculars 4, 2 nd largest. Preoculars, 2, lower one larger. Frontoparietals distinct, larger than the interparietal. A pair of parietals meeting behind interparietal and bordered posteriorly by a pair of nuchals and temporals, also in contact with interparietal, frontoparietal, 4 th supraocular, and 2 postoculars. Loreals 2, similar size; anterior loreal in contact with 1 st and 2 nd supralabial, posterior loreal, prefrontal, frontonasal, and nasal; posterior loreal in contact with 2 nd and 3 rd supralabial, 1 st subocular, upper and lower preocular, prefrontal, and anterior loreal. Supralabials 7, 6 th largest. Infralabials 6, several equally largest. Fifth supralabial below centre of the eye. Temporals: 1 primary; 2 secondary. Ear opening round, moderately large (1.7% as percentage of SVL), with several small projecting granules on anterior margin. Suboculars 8, 3 rd and 4 th separated by 5 th supralabial. Mental broader but shallower than rostral. Postmental larger than mental. Chinshields 3 pairs. Dorsal scales similar in size to ventral scales, weakly striate. Ventral scales and subdigital lamellae smooth. Adpressed limbs meeting. Digits moderately long, subcylindrical. Third front digit similar in length to 4 th . Measurements (in mm; holotype with the variation shown in the specimens examined in parentheses). SVL 100.3 (mean 71.4, range 47.9–101.1), HL 14.1 (mean 10.9, range 6.0–18.0), HW 9.3 (mean 6.4, range 4.5– 10.3), AG 50.9 (mean 37.6, range 24.4–53.5), SF 33.2 (mean 27.1, range 18.4–42.9), S-Ear 19.0 (mean 13.8, range 9.8–23.5), EF 12.6 (mean 14.1, range 9.3–21.1), HLL 33.7 (mean 22.9, range 14.4–36.5); D-Ear 1.7 (mean 1.5, range 0.7–3.4). Variation ( holotype with the variation shown in the paratypes /specimens examined in parentheses). Upper ciliaries 7 (mean 6, range 4–9); lower ciliaries 12 (mean 10, range 7–13); nuchals 3 pair (mean 3 pairs, range 1–6 pairs); midbody scale rows 34 (mean 32, range 29–36); ventral scale rows 87 (mean 71, range 62–87); subdigital lamellae 25 (mean 20, range 15–27); supraciliaries 5 (mean 5, range 4–6); suboculars 8 (mean 6, range 4–8). Frontonasal not usually separated from frontal by prefrontals meeting in midline. Anterior loreal usually in contact with first and second supralabial, posterior loreal usually in contact with second and third supralabial. Supralabials 7 (usual), 8 or 9, the sixth or seventh the largest. Infralabials 5, 6 (usual), or 7. Projecting scales usually present in ear opening. Maximum SVL 101.1 mm . Only 7 of the specimens examined had an intact tail, tail length of intact specimens ranging between 81–124 mm . Mean TL/ SVL = 1.22. Ratios for morphological measurements (+ SD): AG/SF 1.39 + 0.11; S-Ear/EF 0.98 + 0.11; HL/HW 1.70 + 0.11 (N=51). Colouration. This is very variable among specimens, but the most common colouration is as follows: Middorsal stripe where present not usually continuous. Dorsal surface mid to dark brown, usually with light and dark flecking, 6 scale rows wide, grading into pale dorsolateral stripe extending from behind head to base of tail, becoming indistinct thereafter. This pale stripe extends into brown lateral stripe two or more scale rows wide, notched on upper and lower edges, running from behind nostril through eye towards tip of tail. This brown band is bordered on each edge by a dark brown band 1 to 2 half-scale rows wide running above the limbs and becoming indistinct after hindlimb. The lower dark brown band is bordered below by a pale stripe, 1 to 2 half-scale rows wide running from behind or below the eye, above or through the ear, above the limbs to become indistinct after the hindlimbs. This band is bounded below by a darker band breaking up into a ventral pale colour. Soles of feet grey/black. Belly yellow, cream, or pinkish orange, usually speckled with darker flecks. Throat grey. Outer surface of forelimbs bro- wn, speckled with light and dark. Speckling often present on throat and underside of tail. There do not appear to be sexually dimorphic colour patterns. Regarding juvenile colouration, on Stephens Island McCann (1955) noted “in young specimens there is a reddish or coppery flush behind the head as far as the insertion of the forelimbs and occasionally a line of dark dots extending from behind the head to the base of the tail”. A distinctive looking O. newmani (represented by a single individual) was recorded from the Paparoa Ranges ( Miller 1999 , Whitaker & Lyall 2004 ). This specimen has three thin notched longitudinal stripes dorsally, with prominent dorso-lateral whitish stripe bordered by thin black stripes ( Jewell & Morris 2008 ). Etymology. Named for Donald G. Newman of New Zealand , in recognition of his contributions to herpetology ( Wells & Wellington 1985 ). The current accepted vernacular name is ‘speckled skink’ which we suggest changing to ‘Newman’s speckled skink’. Distribution. South Island only, on Stephens Island (Takapourewa), the West Coast and Nelson Lakes across a range of ecological districts (39.03 Sounds, 46.08 Karamea, 48.01 Ngakawau, 48.02 Foulwind, 48.06 Maimai, 48.10 Greymouth, 49.01 Rotoroa and 50.01 Hokitikia Ecological Districts). This species is present in a very wide range of warm-to-cool and low-to-very high rainfall environments from the coast to mountains (up to 1400 m ASL), occupying various habitats from coastal scrub/forest on islands, cobble and boulder beaches, dune vegetation and pakihi wetlands on coastal flats, densely vegetated or shrubland habitats, open grasslands, fernland, and cool temperature tussock-scrub-low forest mosaic forest on the rocky plateaux in alpine regions; includes glacial river terraces in cool inland beech country. Natural History. Diurnal, strongly heliothermic, terrestrial. This is a species that can be highly variable in size and colour; it is largest on Stephens Island (Takapourewa) (SVL range: 31–115 mm ; weight range: 1.5–35.5 g , Stephens 2004 ), but smaller elsewhere on the mainland ( 86 mm SVL, 10 g ). A study on spatial distribution, abundance and relative densities of O. newmani and congeners in various habitat types on Stephens Island (Takapourewa) were undertaken by East et al. (1995) and Markwell (1995) , and followed up by Stephens (2004) , with a focus on the effects of plant succession and ecological restoration of that island on the lizard fauna. This species was found in nearly all habitat types , but unlike other species in the complex, it is also abundant in replanted and natural mature open forest ( Whitaker 1970 b , East et al. 1995, Stephens 2004 ), preferring shady, damp dense vegetation ( Efford et al. 1997 ); it is also increasing in capture rates on the island which may be influenced by increasing habitat availability ( Stephens 2004 ). In 2002–2003 , O. newmani consisted 5–6% of all skink captures on Stephens Island (Takapourewa). On that island, tuatara ( Sphenodon punctatus ) are a natural predator ( Sharell 1966 ), and East et al. (1995) noted tunnelling behaviour in soft soil and leaf litter near seabird burrows which may contain resident tuatara, such fossorial behaviour is likely to help avoid predation. Surprisingly, Sharell (1966) also recorded O. newmani using seabird (fairy prions, Pachyptila turtur ) and tuatara burrows, despite the risk of predation by tuatara. Tail loss rates is high on the island (86–97%), potentially attributable to tuatara and avian predators, however aggression from the northern spotted skink ( O. kokowai ) may also be a cause ( Stephens 2004 ). This species appears to be displaced by the similarly-sized O. kokowai in sympatric habitats on Stephens Island (Takapourewa) and at the Upper Buller Valley, St Arnaud, and this is assumed to be due to ineffective niche partitioning and strong competition ( East et al. 1995 , Stephens 2004 , Efford et al. 1997,1998). Duncan (1999) investigated activity, behaviour and trappability of the same two skink species at St Arnaud and found significant ecological differences and variation in behaviour in the two. Miller & Daugherty (2001) recognised genetic diversity in West Coast populations of O. infrapunctatum . Translocated from Stephens Island (Takapourewa) to Mana Island and Maud Island ( n =40 to each island, both during 2004) ( Sherley et al . 2010 ). Prior to release on Mana Island, a cage trial was undertaken by Heap et al. (2003) to study potential sympatric interactions between O. newmani and the island-resident McGregor’s skink ( O. macgregori ). Since release, it appears that this translocation has failed, as few have been seen since, however it was later realised that this species was not ecologically appropriate for Mana Island, as this was outside the species’ known distributional range. Efford et al. (1997 , 2001 ), Duncan (1999) and Dumont (2015) studied sympatric populations of O. newmani , O. polychroma and O. kokowai at St Arnaud. Efford et al. (2001) found density to be stable between 1994–2000 , with high survival rates (~0.89 – 0.98 / 6 mo), however Dumont (2015) and Nelson et al. (2016) later demonstrated serious decline in the same population by 2012. Capture rates dropped from 0.116 skinks / trap night in 1995 to 0.001 by 2010–2012 , and very few young skinks were found present in this population despite the female-dominated sex ratio, indicating that population recruitment is now very low ( Dumont 2015 ). Dumont (2015) also estimated size at sexual maturity, birthing season and habitat preferences at St. Arnaud: the mean is 72 mm SVL (range 35–98 mm ), females are slightly smaller than males; sexual maturity is estimated to be> 55 mm SVL; pregnant skinks were caught between November to January; sex ratio is 1 ♀ : 0.3 ♂ : 0.1 juvenile; the skinks were more trappable during January than in November, but recapture rates are low at 9.2%; larger, older skinks may be more trappable; and skinks were more often trapped in rocky shrubland habitats. Meads (1970) indicates that mating occurs from late September to November, gestation is 12–14 weeks, and the young (3–4) are born between January and March (all the species described here are viviparous); such information is probably derived from captive O. newmani populations from Takapourewa /Stephens Island. The species is a dietary generalist, diet including shade-dwelling insects such as tiger beetles ( Neocicindela sp.) and fruit at St Arnaud ( Efford et al. 1997 , Efford et al. 1998), but it is also known to devour its own young and other lizards on Stephens Island (Takapourewa) ( Sharell 1966 ). The acari mite Neotrombicula sphenodonti ( Ainsworth 1985 , Goff et al. 1987 ), the nematodes Capillaria sp. ( Allison 1982 ), Hedruris minuta ( Andrews 1974 ) , Parathelandros sp. ( Allison 1982 , Ainsworth 1985 ) and Skrjabinodon trimorphi (Ainsworth 1992, see Mockett et al. 2016); the trematode Dolichosaccus ( Lecithopyge ) leiolopismae (with heavy infestations of>200 flukes/host, Ainsworth 1985 ; Allison & Blair 1987 ) and the protozoan Haemogregarina sp. ( Schmidt 1980 ) have been recorded on O. infrapunctatum from Stephens Island (Takapourewa). Discussion. Until recently descriptions of O. newmani emphasized the relatively large, heavy-bodied skink found at the type locality of Stephens Island (Takapourewa). For example, Hardy (1977) described 32 out of 53 specimens of what is now newmani solely from this one locality. Unfortunately, it appears that this island form differs significantly from mainland (i.e. South Island) populations. In our paper we have tried to redress this imbalance by describing 8 out of 51 specimens from Stephens Island (Takapourewa). The differences between mainland and Stephens Island (Takapourewa) populations are obvious from the measurements in our paper. For example, maximum SVL for mainland animals is 87.7 mm compared with 101.3 for Stephens Island (Takapourewa). Ventral scales—maximum 87 Stephens Island (Takapourewa), compared with 81 for mainland populations; lamellae—max 25 Stephens Island (Takapourewa) compared with 22 mainland; higher mean midbody scale count for Stephens Island (Takapourewa) animals, etc. There is some argument that this population should be classed a subspecies on morphology alone, but this is outside the scope of this paper.