Oligosoma hoparatea, Whitaker & Chapple & Hitchmough & Lettink & Patterson, 2018

Whitaker, Tony, Chapple, David G., Hitchmough, Rodney A., Lettink, Marieke & Patterson, Geoff B., 2018, A new species of scincid lizard in the genus Oligosoma (Reptilia: Scincidae) from the mid-Canterbury high country, New Zealand, Zootaxa 4377 (2), pp. 269-279 : 270-278

Oligosoma hoparatea sp. nov.

( Figures 4, 5 , 6 )

Holotype. NMNZ RE008536 (adult female) Mt Harper , Harper Range, Rangitata River, Canterbury, New Zealand; 43° 38' 16.5"S, 171° 03' 50.1"E; 1050 m; collected A.H. Whitaker, 26 February 2004.

Paratypes (2 specimens). NMNZ RE008537; adult male/female; same collection data as holotype. NMNZ RE008538 (sub-adult); same locality data as holotype; collected T.R. Jewell, 8 October 2007 .

Note: the three type specimens are the only specimens currently known in research collections. Other live specimens were examined by the late AHW in the field.

Diagnosis. Oligosoma hoparatea sp. nov. can be distinguished from other Oligosoma species by the following combination of characters: (a) interrupted subocular scale row; (b) mid-body scale rows>38; (c) fourth toe lamellae>25; (d) a body colouration consisting of prominent brown longitudinal stripes; (e) uniformly white ventral surface; and (f) white lateral stripe passing through the ear. It is most similar to Oligosoma longipes , which is its closest relative ( Chapple et al. 2009) and with which it occurs syntopically ( Chapple and Hitchmough 2016). These two species can be further distinguished by: (a) the presence of a darker mid-dorsal stripe in most O. hoparatea sp. nov. versus usually absent in O. longipes , (b) the dorsal surface without flecks or blotches in O. hoparatea sp. nov. versus well marked with darker flecks in O. longipes , (c) lateral stripes and bands smooth edged in O. hoparatea sp. nov. versus notched in O. longipes , (d) always two or more nuchal scales in O. hoparatea sp. nov. compared with usually one or fewer in O. longipes ; (e) ventral surface uniformly white or cream in O. hoparatea sp. nov. versus white or greyish, often with some dark flecking in O. longipes , (f) series of enlarged scales on top of the front foot in line with the toes in O. hoparatea sp. nov. versus scales of diminishing size in O. longipes , and (g) enlarged precloacal scales wider than deep in O. hoparatea sp. nov. versus enlarged precloacal scales much deeper than wide in O. longipes .

Description of Holotype. (where bilateral counts differ, the count for the right side is in parenthesis)

Body elongate, oval in cross-section; forehead rounded, snout blunt; ear opening large, rounded, slightly higher than wide; limbs well-developed, pentadactyl; hind limbs 37% of SVL and 1.4× length of fore limbs, adpressed limbs meet; 4th front toe longer than 3rd.

Rostral 1.6× wider than high, contacts 1 st supralabials , nasals and frontonasal; nasals widely separated, undivided; nostril in lower half and angled back and up; supranasals absent; frontonasal 1.6× wider than deep, contacts rostral, anterior loreals, prefrontals and frontal, narrow contact with frontal; prefrontals separated, contact with frontonasal, frontal, 1 st supraocular , 1 st supraciliary and anterior and posterior loreals; frontal 1.3× longer than wide, shorter (0.9×) than frontoparietals + interparietal, contacts 1 st and 2nd supraoculars; 4 supraoculars, 2nd largest, 2nd supraocular in broad contact with frontal; frontoparietals distinct, 1.1× longer than interparietal; parietals meet on left of midline, left parietal fragmented (small scale towards midline) and overlaps right parietal, parietals bordered by interparietal, frontoparietals, 4th supraocular, 2 postoculars, 2 temporals and nuchals; 3 pairs of nuchals, twice the width of adjacent dorsals; 2 loreals; anterior loreal largest, rhomboidal, contacts nasal, frontonasal, prefrontal, posterior loreal, 1 st and 2nd supralabial; posterior loreal subtriangular, contacts anterior loreal, prefrontal, 1 st supraciliary , lower preocular, 1 st subocular and 2nd supralabial (narrow contact); 2 preoculars, lower the largest; 9 supraciliaries, 1 st largest ; 8 upper ciliaries, 7th largest, forming prominent eyelid; 12 lower ciliaries; lower eyelid with clear palpebral disc surrounded either side and below with granules; 5(6) suboculars, anterior largest, contacting subocular 2nd and 3rd supralabials (2nd subocular on right side is fragmented); subocular row interrupted by 6th supralabial; 2 postoculars; 2(1) primary temporals; 2 secondary temporals; 3 tertiary temporals; 8 supralabials, 6th supralabial under centre of eye; 2 postlabials; 8(7) infralabials, gradually increasing in size, 5th and 6th largest; mental 2.5× wider than deep; post-mental quadrangular, larger than mental, contacts 1 st and 2nd infralabials; 3 pairs of chinshields; anterior chinshields in broad contact, contact 3rd and 4th infralabials; second chinshields separated by one scale width (i.e three scales wide between infralabials) but two scales deep, contact 4th and 5th infralabials; posterior chinshields separated by 4 gular scales, contact 5th and 6th (5th) infralabials; one scale between 3 temporals and ear opening; ear with 3 prominent triangular lobules on the anterior margin, uppermost the largest, lower 2 small; granules on lower posterior margin of ear opening; dorsal body scales with 3–5 weak striations, lateral and ventral scales smooth; ventral scales largest (1.1× dorsals), laterals smallest (0.8× dorsals); 14 enlarged precloacal scales, 1.2× wider than deep, largest on midline, 1.3× larger than ventrals; 20 postcloacal scales, not enlarged; palmar granules domed; top of front feet with series of three enlarged scales in line with toes, scales on top of hind feet small. Premaxillary teeth enlarged (14) (28 maxillary teeth), curved.

Measurements (given as percentage of snout–vent length). SVL 79.6 mm; ITL unknown (tail is regenerated); SF 32.3%, AG 61.7%, S-Eye 7.7%, D-Eye 3.8%, S-Ear 17.3%, D-Ear 2.4%, HL 15.1%, HW 9.2%, FL 25.9%, FFL 7.4%, HLL 36.7%, HTL 12.3%.

Measurements (in mm; holotype with the variation shown in the type series in parentheses). SVL 79.6 (mean 67.2, range 54.0–79.6); SF 25.7 (mean 24.4, range 21.3-25.7); AG 49.1 (mean 38.9, range 28.0-49.1); S-Eye 6.1 (mean 5.1, range 4.5-6.1); S-Ear 13.8 (mean 12.6, range 11.8-13.8); EF 11.0 (mean 9.8, range 9.1-11.0); D-Eye 3.0 (mean 2.6, range 2.5-3.0); D-Ear 1.9 (mean 1.5, range 1.1 -1.9); HL 12.0 (mean 10.9, range 9.9–12.0); HW 7.3 (mean 6.2, range 5.6-7.3); FL 20.6 (mean 19.5, range 18.8-20.6); FFL 6.0 (mean 6.0, range 6.0-6.1); HLL 29.2 (mean 27.6, range 26.5-29.2); HTL 10.7 (mean 9.5, range 7.8-10.7); NS 3 (mean 3, range 2 -4).

Variation (holotype with the variation shown in the type series in parentheses). Midbody scale rows 40 (mean 41, range 40-43); dorsal scale rows 96 (mean 90, range 80-96); ventral scale rows 103 (mean 97, range 89- 104); third finger scales 13 (mean 13, range 11-14); third finger lamellae 18/19 (mean 18, range 17-19); fourth toe scales 16 (mean 16, range 15-17); fourth toe lamellae 28 (mean 27, range 26-28).

Colouration. This is quite similar among specimens, and is as follows: Mid-dorsal stripe continuous, becoming indistinct anteriorly. Stripe continues down tail, gradually becoming indistinct towards tip. Dorsal surface mid to dark brown, 2 scale rows wide, grading into darker stripe 2 half-scale rows wide. This is bordered by a pale dorsolateral stripe extending from behind head to base of tail, becoming indistinct thereafter. This pale strip extends into brown lateral band 3 scale rows wide, running from behind nostril through eye to base of tail, becoming indistinct thereafter, and bordered on upper and lower edges by a narrow dark brown stripe. The lower stripe is bordered below by a pale stripe, 2 half-scale rows wide running from below the eye, through the ear, above the limbs to stop just anterior to hind limbs. This band is bordered below by a darker band ½ scale rows wide. Soles of feet brown/dark brown, belly white or cream, unmarked. Throat white, unmarked. Outer surface of forelimbs brown, with indistinct pale stripe. There do not appear to be sexually dimorphic colour patterns. Juvenile colouration similar to that in adults.

Etymology. The specific name is from the Maori hōpara, meaning belly or underside, and tea, meaning white, thus hōparatea or ‘white-bellied’ in reference to the uniquely uniform white ventral surface of many specimens of this skink. After consultation with the wider herpetological community, the Society for Research on Amphibians and Reptiles in New Zealand (SRARNZ) and the Department of Conservation revised the common name nomenclature for New Zealand reptiles in 2014. Here O. hoparatea sp. nov. was called the “Pukuma” skink. However in this paper the recommended common name is White-bellied Skink. This aligns with the wishes of the deceased author, AHW.

Phylogeny and relationships. Using Neighbour-Joining, Maximum Parsimony, Maximum Likelihood and Bayesian phylogeny analyses, Chapple et al. (2009) recognised eight major and well-supported clades within the New Zealand skink fauna. O. hoparatea sp. nov. was included with nine South Island taxa in Clade 1; within this clade it grouped closely with O. longipes (southern and northern taxa), O. nigriplantare (Peters) and the O. polychroma (Patterson & Daugherty) species complex.We include a cut-down version of Chapple et al.’s (2009) phylogenetic tree here ( Figure 7 ). We found the Mt Somers O. hoparatea sp. nov. population to be 1.4% divergent for ND2 from the Mt Harper samples (holotype and one paratype) included in Chapple et al. (2009). Overall, O. hoparatea sp. nov. is most closely related to sympatric populations of O. aff. longipes ‘southern’, but the level of divergence (7.1%; uncorrected genetic distance, based on the ND2 mtDNA genes) clearly shows it is a distinct taxon with a divergence time estimated at approximately 5.4 Million years before present (MYBP) ( Chapple et al. 2009). The southern (Canterbury) and northern (Marlborough) populations of O. longipes showed a similar level of divergence (8.2%) for ND2 indicating further diversity within this group of skinks.

Distribution and abundance. O. hoparatea sp. nov. is currently known from four localities, with two on the northern slopes of Mt Harper, above Balmacaan Stream. At this site, O. hoparatea sp. nov. abundance is very low (0– 8 individuals encountered per site visit; observers typically spent several hours searching for emerged skinks using binoculars and the naked eye) and, judging by comparison of encounter rates during searches, appears to have declined since the species’ discovery. For example, no skinks were seen or caught during the most recent attempt to locate the species by visual searching and live trapping undertaken from 3–5 February 2017 (Dylan van Winkel, pers. comm.). During the original (February 2004) survey, O. hoparatea sp. nov. (one individual) was also observed in a nearby scree gully located <1 km to the west of the type locality. Visual searching undertaken in that gully on two occasions in January 2010 did not yield any further sightings of the species.

In 2012, O. hoparatea sp. nov. was found in a scree gully on the Mt Somers Range, c. 20 km north-east of Mt Harper. Abundance at this site is also very low (0–4 individuals encountered per site visit). A fourth site was located c. 2 km north in 2014, consisting of a scree gully and scree slope connected by continuous scree habitat at higher (> 1400 m a.s.l.) elevations. Although the number of O. hoparatea sp. nov. observed per site visit was low (range: 6–14 individuals seen over 3–4 person-hours of searching: Unpubl. data), this locality appears to support the largest sub-population discovered to date. The distribution of O. hoparatea sp. nov. is patchy at all four sites and skinks appear to be absent from much of the habitat that appears to be suitable for them. However as Jewell & Morris (2011) point out, in general appearance and size O. hoparatea sp. nov. is very similar to O. maccanni (Patterson & Daugherty) . Therefore it is possible that it has been mistaken for this ubiquitous species in other parts of its range, and may be more common than is believed.

Ecology. To date, O. hoparatea sp. nov. has only been found in mobile scree habitat consisting of angular greywacke rocks, located within a narrow elevation range of c. 1000¯ 1250 m a.s.l. and warm (north-to-westfacing) aspects. Skinks are strongly associated with “vegetation islands” growing in more stable areas of the scree habitat, which are made up of indigenous divaricating shrub and vine species (including creeping pōhuehue Muehlenbeckia axillaris , matagouri Discaria toumatou , porcupine shrub Melicytus alpinus , mountain wineberry Aristotelia fruticosa , korokio Corokia cotoneaster , and Coprosma, Rubus , Clematis and Parsonsia species; Figures 2 and 3 ). Like the majority of New Zealand skink species, O. hoparatea sp. nov. is a diurnal heliotherm. Individuals are most easily found by scanning rocks on the periphery of the vegetation islands during the first sunny window of the day, once ambient temperatures that permit emergence are reached (±12¯14 °C). After an initial period of basking, skinks often move meters away from the vegetation islands to forage in adjacent scree.

Based on the diets of its congeners and potential available food sources, O. hoparatea sp. nov. is likely to be omnivorous, supplementing its diet with the small fleshy fruits of divaricating shrubs and vines when these are seasonally available. Most New Zealand lizard species consume a variety of arthropod prey and frugivory is widespread (Chapple 2016, Wotton et al. 2016). Skinks have been seen stalking flies amongst mats of creeping pōhuehue. Although vigilant and agile, most observed attempts by O. hoparatea sp. nov. to capture flies by lunging were unsuccessful. Notably, O. hoparatea sp. nov. also includes skinks in its diet. One of the specimens collected in 2004 regurgitated a sub-adult skink (SVL 45 mm) of a smaller species ( O. maccanni ) following capture. The maximum SVL recorded for O. hoparatea sp. nov. is 91mm: Unpubl. Data.

Syntopic lizard species that occur in the scree habitat of O. hoparatea sp. nov. at the type locality are O. waimatense , O. aff. longipes “southern”, O. maccanni , O. aff. polychroma clade 4, O. lineoocellatum (formerly O. aff lineoocellatum “central Canterbury ”; Melzer et al. 2017), Woodworthia “Southern Alps” and W. “pygmy”. This remarkable diversity (6 skink and 2 gecko taxa) represents the greatest species richness known from any site on the mainland of the South Island ( Chapple & Hitchmough 2016). Four of the above species (the Woodworthia gecko species, O. maccanni and O. aff. polychroma clade 4) are relatively abundant at the type locality. On the Mt Somers Range, O. hoparatea sp. nov. co-exists with five of the above species (O. aff. longipes “southern”, O. maccanni , O. aff. polychroma clade 4, O. lineoocellatum and W. ‘Southern Alps’), and a sixth species ( O. waimatense ) occurs in the vicinity.

Resource partitioning among the eight lizard species living in scree habitat at the type locality occurs via the time of their activity (diurnal for skinks; primarily nocturnal for geckos), micro-habitat use, and possibly diet. Although we have not conducted dietary studies, differences in lizard body sizes (maximum SVL ranges from 44 mm in W. “pygmy” to 114 SLV in O. waimatense: Unpubl. Data ) are expected to influence the size of arthropod prey items taken. Three of the six skink species are saxicolous (restricted to rocky substrates; O. hoparatea sp. nov., O. aff. longipes “southern” and O. waimatense ). These species vary in their micro-habitat use: O. aff. longipes “southern” and O. hoparatea sp. nov. are associated with vegetation islands whereas O. waimatense exhibits a preference for relatively bare areas of comparatively larger-sized scree (ML; pers. obs.).

O. hoparatea sp. nov. shares several characters with other saxicolous Oligosoma species, such as high scale counts and elongated digits and tail. The colouration allows the species to be camouflaged in its rocky habitat, although the lack of speckling in the colour pattern distinguishes this species from most other saxicolous skinks in New Zealand, including the syntopic species O. waimatense and O. longipes .

Potential predators of O. hoparatea sp. nov. include other skink species (particularly the larger O. waimatense ), birds (e.g. black-fronted tern Chlidonias albostriatus (Gray) and New Zealand falcon Falco novaeseelandiae Gmelin ) and introduced mammals (feral house mouse Mus musculus Linnaeus , brushtail possum Trichosurus vulpecula (Kerr), European hedgehog Erinaceus europaeus Linnaeus , ship rat Rattus rattus Linnaeus, Norwegian rat R. norvegicus (Berkenhout) , feral pig Sus scrofa Linnaeus, feral ferret Mustela furo Linnaeus , stoat M. erminea Linnaeus , weasel M. nivalis Linnaeus and feral cat Felis catus Linnaeus ). Feral cats, stoats, hedgehogs, possums, pigs and mice are known to be present at or near the type locality. In the absence of habitat loss, the greatest threat to lizard populations on the mainland of New Zealand (e.g. Reardon et al. 2012) is predation by introduced mammals.

Conservation status. O. hoparatea is extremely rare; since its discovery it has been seen on only four scree areas on two adjacent mountain ranges, and may have disappeared from one of these, and the population on each is only a few tens of animals at most, based informally on the small numbers observed. Numerous apparently suitable screes have been searched, but no additional populations have been detected at these locations. The combination of very small fragmented populations and large areas of apparently unoccupied suitable habitat mean that historical and ongoing declines are highly likely, although population trend has not been documented.

Following its discovery in 2004, O. hoparatea sp. nov. was listed by Hitchmough et al. (2007) under the tag-name “ Oligosoma aff. longipes "Rangitata” as Data Deficient (One Location). The Data Deficient category is for very poorly known taxa, which lack sufficient information for even an educated guess at conservation status. The “One Location” qualifier is self-explanatory—at that time, the skink was known only from the site of the original discovery. Hitchmough et al. (2010) shifted it to Nationally Critical (Data Poor, One Location), on the basis of criterion A(2) (Ĺ2 subpopulations, Ĺ200 mature individuals in the largest subpopulation). Nationally Critical is the NZTCS category for species at highest risk of extinction, facing the risks imposed by extremely small population or area of occupancy, or extremely high rate of decline. The Data Poor qualifier indicates low confidence in the information used to make the assessment (in this instance because it was thought quite likely that additional populations would be discovered). Hitchmough et al. (2013) changed only the qualifiers, to Data Poor, Range Restricted, Sparse, following the discovery of another subpopulation on the western slopes of the Mount Somers Range. The “Range Restricted” and “Sparse” qualifiers replacing the previous “One Location” indicate “taxa confined to specific substrates, habitats or geographic areas of less than 1000 km 2” and “taxa that occur within typically small and widely scattered populations” respectively. There was no change to the listed status in the latest ( Hitchmough et al. 2016b) listing.

O. hoparatea sp. nov. has not been assessed under the IUCN red-list system. We have followed the directive in the "guidelines for using the IUCN categories and criteria, version 13, March 2017 " ( IUCN Standards and Petitions Subcommittee 2017) that "precise information on scarce taxa is usually lacking, and although the criteria are highly quantified and defined, one can use projections, assumptions and inferences (as long as they are explicitly stated and clearly justified) in order to place a taxon in the appropriate category". Although there is no formal scientific population size or trend assessment, we are confident that our observations of small numbers, fragmented populations, and declining encounter rates merit listing as Critically Endangered under criterion C2a(i):

C. Population size estimated to number fewer than 250 mature individuals and: 2. A continuing decline, inferred, in numbers of mature individuals AND a. Population structure in the form of one of the following:

(i) no subpopulation estimated to contain more than 50 mature individuals.

Introduced mammalian predators are the likely agent of decline, based on extremely strong evidence for their impact on congeners in similar habitats (e.g. Reardon et al. 2012). O. hoparatea sp. nov. is larger than most of the sympatric congeners which are more common at the sites where it is found; larger body size has been shown to increase vulnerability to extinction ( Tingley et al. 2013).

Figures

FIGURE 2. North face of Mt Harper viewed from the northern shore of Lake Camp showing the rounded landform with steeply-incised gullies with associated active screes and large outwash fans. Oligosoma hoparatea sp. nov. has been found in screes in the gully at the left (the type locality) and to the left of centre [Photograph: Don Geddes].

FIGURE 7. Relationship of taxa within Oligosoma Clade 1 of Chapple et al. (2009)—phylogenetic tree of Maximum Likelihood (ML) from five mitochondrial genes (ND2, ND4, Cytb, 16SrRNA, 12SrRNA) and one nuclear gene (Rag-1); ML bootstrap values and Bayesian posterior probabilities for all branches are 100 and 1.0, respectively [after Chapple et al. 2009]. The newly described species Oligosoma hoparatea sp.nov. is highlighted.

FIGURES 4 & 5. Wild Oligosoma hoparatea sp. nov. photographed on screes at the type locality, Mt Harper, January 2010 [Photographs: Marieke Lettink].

FIGURE6. Detailed drawings of the holotype of Oligosoma hoparatea sp.nov.—dorsal (upper left), left side (middle left) and ventral (lower left) surfaces of the head showing details of the scalation, and the upper surface of the left front foot (upper right) and lower surface of the left hind foot (lower right). Scale bar = 5 mm and scaleis the same for all drawings.

FIGURE3. Activegreywackescrees at thetypelocality at 1050 mon thenorth face MtHarper showing ‘islands’ ofvegetation amongst large expanses of gravel and boulders that are the habitat of Oligosoma hoparatea sp. nov. [Photograph: Marieke Lettink].