A New Species of the Genus Gekko (Squamata: Sauria: Gekkonidae) from the Dabie Mountains, China

Simple Summary The genus Gekko Laurenti, 1768, currently comprises approximately 86 species, of which 21 are native to China, that commonly live on walls, rocks, and trees, and are distributed across Southeast Asia, western Oceania, and Melanesia. This article describes a new species of Gekko (Squamata: Gekkonidae) based on its distinct morphological features and molecular evidence, which was identified in the Dabie Mountains on the border of Anhui and Henan provinces of Central China. The analysis of phylogeny based on a mitochondrial DNA fragment (16S, CYTB, and COI) indicated that the new taxon is different from its congeners. Morphologically, the new species can be diagnosed from the other subgenus Japonigekko species by a combination of 34 (14 mensural and 20 meristic) morphological characteristics, and Principal component analysis (PCA) and one-way analysis of variance (ANOVA) showed that the new species can be clearly distinguished from its sister species G. hokouensis. Based on the above multiple lines of evidence, we describe this gecko from the Dabie Mountains as a new species, Gekko kaiyai sp. nov. The discovery of this species implies that there are now 87 identified species in the genus Gekko, 22 of which can be found within China. Abstract This study describes a novel species of Gekko (Squamata: Gekkonidae) based on its distinct morphological features and molecular evidence, which was identified in the Dabie Mountains on the border of Anhui and Henan provinces of Central China. Gekko kaiyai sp. nov. could be distinguished from its congeners owing to its morphological characteristics, such as being a medium body sized gecko species (snout–vent length, 56.98–64.99 mm, n = 4, females; 50.03–61.56 mm, n = 11, males); nostrils scale in contact with rostral scale; tubercles on the dorsal and limb, while the upper forelimb is smooth with no tubercles; 22–33 interorbital scales between the anterior corners of the eyes; 157–209 ventral scales between mental and cloacal slit; 90–121 midbody scale rows; 30–43 ventral scale rows; 7–9 sub-digital lamellae on first fingers, 8–13 fourth fingers, 7–9 first toes, and 7–11 fourth toes; free of webbing in the fingers and toes; 9–12 pre-cloacal pores in males, which are absent in females; post-cloacal unilateral tubercles 1 (few 2); and a dorsum that is greyish white to dark brown, with 6–7 brown markings between the nape and sacrum. The phylogenetic tree based on the mitochondrial DNA sequences (16S, CYTB, and COI) indicated that Gekko kaiyai sp. nov. form an independent clade with strong support (100/1) and are a sister group to G. hokouensis. At the inter-species level, the genetic distances were all large, further confirming that an independent species had been identified. The discovery of this species implies that there are now 87 identified species in the genus Gekko, 22 of which can be found within China.

The Dabie Mountains are located at the junction of the Anhui, Henan, and Hubei Provinces, and this area forms the watershed between the Yangtze and Huai Rivers.It has a warm and humid monsoon climate in the North subtropical zone, and its superior geographical environment and suitable climatic conditions indicate that it is rich in animal and plant resources.According to previous records, two species of the genus Gekko (G.hokouensis and G. japonicus) have been identified in the Dabie Mountains [8].During herpetological surveys in the Dabie Mountains from August to September 2022, some Gekko specimens were collected that were not G. hokouensis or G. japonicus.Based on molecular phylogenetic analyses and morphological comparisons, a new species was identified as an undescribed taxon of the genus Gekko.

Sampling
From August to September 2022, 17 gecko individuals (two subadult, four adult females, and 11 adult males) were collected from the Dabie Mountains in the border region of the Anhui and Henan Provinces, China (Figure 1).For comparison, 25 specimens of G. hokouensis and two specimens of G. japonicus from the Dabie Mountains and surrounding areas of Anhui Province were also collected.After taking photographs, the individuals were euthanized (using isoflurane) and fixed in 10% formaldehyde for 2 d, and then finally washed and preserved in 75% ethanol.The study received ethical review and approval from the Institutional Animal Care and Use Committee of School of Life Sciences, Anhui University (project number IACUC(AHU)-2022-050). Vouchered specimens for this work were deposited at the Anhui University Biology Museum (AHUBM).

Molecular Data and Phylogenetic Analyses
Liver tissue samples were taken from all geckos and preserved in 95% ethanol prior to fixation.Purified DNA was then obtained from the liver tissues of individuals using a standard phenol/chloroform extraction method [9,10].

Molecular Data and Phylogenetic Analyses
Liver tissue samples were taken from all geckos and preserved in 95% ethanol prior to fixation.Purified DNA was then obtained from the liver tissues of individuals using a standard phenol/chloroform extraction method [9,10].
Mitochondrial gene segments encoding the 16S ribosomal RNA gene (16S), cytochrome b (CYTB), and cytochrome oxidase subunit I (COI) were selected, and the homologous regions were amplified using the primers previously described by Lyu et al. (2021) [6] and Kumazawa and Endo (2004) [11].The L3975 (5 -CGCCTGTTTACCAAAAACAT-3 ) and H4551 (5 -CCGGTCTGAACTCAGATCACGT-3 ), primers were used for 16S, L-14731 (5 -GAAAAACTATCGTTGTTATTCAACTA-3 ) and H-Thr-2 (5 -GTTTACAAGGTCAGCGCTTT-3 ) primers were used for CYTB [6], and rCOI-1H (5 -TAGTGGAARTGKGCTACTAC-3 ) and rTrp-1L (5 -TAAACCARGRGCCTTCAAAG-3 ) primers were used for COI [11].The PCR amplification system volume was 25 µL, and this included 2 µL of template DNA, 1 µL of the upper and lower primers, 12.5 µL of Taq polymerase, and 9.5 µL of ddH2O.The PCR cycle parameters for the 16S and CYTB included an initial denaturation step at 95 • C for 4 min, followed by 35 cycles at 95 • C for 40 s, annealing at 53 • C for 34 s, and then expansion at 72 • C for 1 min; this was followed by a final extension step for 10 min of 72 • C [6].The COI amplification was performed under the following conditions: 1 min at 94 • C, followed by 5 cycles at 94 • C for 1 min, 1.5 min at 48 • C, and 1.5 min at 72 • C, followed by 35 cycles of 1 min at 94 • C, 1.5 min at 58 • C, and 1.5 min at 72 • C, and then a final 5 min at 72 • C [2].The PCR amplified products were sent to General Biology (Anhui) Co., Ltd.(Chuzhou, Anhui, China), and the derived sequences were stored in GenBank (for GenBank accession numbers, see Table 1).According to previous studies, the gene sequences for all species in the subgenus Japonigekko from GenBank were downloaded for use in extensive phylogenetic comparisons.Moreover, for phylogenetic analysis, the available sequence data for Gekko (Gekko) gecko (Linnaeus, 1758) were downloaded and used to aid in outgroup comparisons.Overall, 105 sequences were selected for analyses in this study, and these were sourced from 47 gecko individuals and 16 species.(For all GenBank accession numbers, see Table 1).
The 16S, CYTB, and COI sequences were aligned using MAFFT 7.110 with the G-INI-i option [13,14].The aligned sequences were then sheared in BioEdit 7.0.5.3 to remove incorrect base sequences from the head and tail region [15].The datasets for the 16S, CYTB and COI regions were then combined manually.Before multi-gene phylogenetic tree reconstruction, the jModelTest was used to estimate the best-fit evolutionary model for the alignment using the calculation from the corrected Akaike information criterion [16,17].The results showed that the GTR + I + G model was the best partition.The 16S, CYTB, and COI combined, sequenced datasets were analyzed using Bayesian inference (BI) in MrBayes 3.2.4[18], and maximum likelihood (ML) was performed using RaxmlGUI 1.3 [19].Two independent runs were conducted using the BI analysis (each of which was performed for 10,000,000 generations and sampled every 1000 generations with the first 25% of samples discarded as burn-in and the remaining 75% retained), and used to construct a 50% majority consensus tree and Bayesian posterior probabilities (BPPs) were calculated.This process generated a potential scale reduction factor (PSRF) of <0.005, which confirmed that the trees had error rates 0.01.Markov Chain Monte Carlo simulation convergence was assessed using Tracer v1.5 [20], and the results were verified using the ESSs of all parameters that exceeded 200 with PSRFs close to 1000.

Morphological Analyses
In accordance with previous studies, 34 (14 mensural and 20 meristic) commonly -used morphological characteristics were assessed in the adult gecko samples using previously described terminologies and methods [2,5,6].Morphological measurements were generally taken to the nearest 0.01 mm using digital calipers (DEGUQMNT, 0-150 mm); however, the characteristics with small values (<10 mm) were measured using a digital stereoscopic binocular microscope.The distance from the tip of the snout to the posterior edge of the vent was the snout-vent length (SVL), from the posterior margin of the cloaca to the tip was the tail length (Tal), the minimum distance between the axilla and groin on a straightened body was the axilla to groin distance (AG), the tip of the snout to the posterior margin of the ear opening was the head length (HL), the maximum head width was the head width (HW), the maximum head height was the head height (HH), the distance from the snout tip to anterior corner of the eye was the snout-eye distance (SE), the distance between the posterior margin of the eye and the posterior margin of the ear opening was the eye-ear distance (EE), greatest diameter of orbit (ED), maximum diameter of the ear opening (TD), the maximum rostral scale width was rostral scale width (RW), the maximum rostral scale height was the rostral scale height (RH), the maximum mental width was mental width (MW), and the maximum mental length was indicated as mental length (ML).
Considering that sexual dimorphism may exist within geckos, sexes were separated for subsequent comparisons among the samples.At the same time, to account for the possible influence of allometry, subadults were omitted from the specimen's data and then scaled to remove allometric effects of body size using the following equation: Xadj = log(X) − β[log(SVL) − log(SVLmean)], where Xadj = adjusted value; X = measured value; β = unstandardized regression coefficient for each population; and SVLmean = overall average SVL of all populations [21][22][23][24]-accessible in the R package GroupStruct (available at https://github.com/chankinonn/GroupStruct,accessed on 11 November 2023).The morphometrics of each species were adjusted separately and then concatenated prior to analysis so as not to conflate intra-with interspecific variation [25,26].One-way analysis of variance (ANOVA) tests were used to evaluate significant differences in the morphometric characteristics between the different species, with a p < 0.05 in the Levene's test.Using the dplyr package in R to perform Bartlett's test of sphericity on the original variables, the data were checked as suitable for Principal component analysis (PCA), and then Kaiser-Meyer-Olkin Measure (KMO) of Sampling Adequacy analysis to screen the variables (communality < 0.5).PCA implemented by the factoextra package in R was employed to extract PCA analysis result information and draw PCA plots.All statistical analyses were performed with R version 4.2.2.
The meristic characteristics and their abbreviations were as follows: naso-rostrals, supra-nasals, and post-nasals as nasals (N); scales between supra-nasals, in contact with rostral scale as intersupra-nasals (I); the number of scales from the commissure of the jaw to the rostral scale as supra-labials (SPL); number of scales from the commissure of the jaw to the mental scale as infra-labials (IFL); number of scales in a line between the anterior corners of the eyes as interorbitals (IO); number of scales in a line from the nostril to the anterior corner of the eye as preorbitals (PO); post-mentals (PM); gulars bordering the post-mentals (GP); dorsal tubercle rows at the midbody (DTR); granules surrounding dorsal tubercles (GSDT); scales in a line from the mental to the front of the cloacal slit (SMC); scale rows at the midbody (including ventral scales, SR); ventral scale rows at midbody (V); sub-digital lamellae under entire first finger (LF1); sub-digital lamellae under entire fourth finger (LF4); sub-digital lamellae under entire first toe (LT1); sub-digital lamellae under entire fourth toe (LT4); pre-cloacal pores (PP); and post-cloacal tubercles (PAT).Bilateral scale counts are given as left/right.

Phylogenetic Analyses
The aligned dataset contained data from 46 individuals from the subgenera Japonigekko species and one individual from the outgroup Gekko gecko species (Table 1).The combined dataset for the 16S, CYTB, and COI regions included 44 collections that represented 16 taxa and resulted in a concatenated alignment of 2474 characters with GTR + I + G as the best-fit evolutionary model.The ML search stopped after 150 BS replicates.For the BI, all chains converged after 10 million generations with an average standard deviation of split frequencies of 0.004379; the average ESS was 5636.The ML and BI algorithms generated similar topologies in the main lineages; thus, only the topology generated by the ML algorithm is presented along with the BS value and BPPs > 50% and 0.90, respectively, at the nodes.Eight specimens for the undescribed species formed an independent clade with strong support (100/1) and were a sister group to G. hokouensis (Figure 2).
To further explore the species relationships among the subgenus Japonigekko, the alignment with 14 selected CYTB sequences, 12 selected 16S sequences, and 7 selected COI sequences underwent a genetic distance analysis.The uncorrected pairwise divergences within these specimens ranged from 6.8% to 29.4% for the CYTB sequences (Table S1), 2.1% to 19.1% for the 16S sequences (Table S2), and 18.8% to 25.9% for the COI sequences (Table S3).For the CYTB, the genetic distance between the undescribed species and other known species ranged from 18.6% (G.hokouensis) to 27.8% (G.chinensis).According to the statistical analyses, all genetic distances between species were >14.0%except for those between G. adleri and G. palmatus, and G. chinensis and G. similignum, which were 7.9% and 6.8%, respectively.For the 16S rRNA sequences, the genetic distance between the undescribed species and other known species ranged from 9.3% (G.hokouensis) to 19.0% (G.melli), and comparison analysis showed that the undescribed species and G. hokouensis were closely related.Taken together, the genetic distance between the undescribed species and the subgenera Japonigekko was found to be high enough to indicate a new species.

Morphological Analyses
The results of the ANOVA indicated that the new taxon group was significantly different from the closely-related species (G.hokouensis) in many morphometric characters (p-values < 0.05; Table 2); the males' significant differences were characteristics in AG, HW, HH, ED, TD and MW, whereas females' were characteristics in SVL, AG, HL, HW, HH, ED, TD, and RW.The results of the Bartlett's test of sphericity show that there are correlations among variables (p < 0.05) which indicated that PCA can be performed.We found most variables have at least a moderate correlation value (i.e.,: >0.3 or <−0.3), except (SE, RW, ML).So, we omitted the variable with low correlation variables.At the same time, the Kaiser-Meyer-Olkin measure (KMO) of sampling adequacy analysis shows an overall Measures of Sampling Adequacy (MSA) of 0.76 (>0.5).In PCA analysis, the first four principal components explained 79.24% of the total variation in the males, where PC1, PC2, PC3, and PC4 eigenvectors accounted for 39.30%, 20.70%, 9.88%, and 9.35% of the total variance, respectively (Table 3).Similarly, the first four principal components occupied a considerable proportion in the females, 76.89% of the total, whereas the PC1, PC2, PC3, and PC4 eigenvectors accounted for 34.83%, 18.12%, 14.06%, and 9.88% of the total variance, respectively (Table 3).Regardless of the sex of the two aforementioned species, the samples showed intraspecific polymerization; moreover, all samples showed interspecific detachment, as evidenced by the fact that there was no overlap between the two species on the two-dimensional graphs for PC1 and PC2 (Figure 3).The results of the ANOVA and PCA indicated that the unnamed populations were significantly different from the closely-related species.

Taxonomic Accounts
The results of the molecular phylogenetic analyses and morphological comparison all indicated that the new taxon of the genus Gekko (Japonigekko) from the Dabie Mountains is significantly different from other known species of the same genus.Therefore, we describe it here as a new species.

Taxonomic Accounts
The results of the molecular phylogenetic analyses and morphological comparison all indicated that the new taxon of the genus Gekko (Japonigekko) from the Dabie Mountains is significantly different from other known species of the same genus.Therefore, we describe it here as a new species.
Paratype.Sixteen specimens (two subadults, four adult females and 10 adult male), collected from the three locations by C.W. Zhang, A.F. Wu, S.L. Yu, X.N.Li, and L.R. Wang.Five specimens AHULXBH001-005 collected on 19 August 2022 from the Dongzhai National Nature Reserve, Luo County, Xinyang City, Henan Province, China.Nine adult and one subadult specimen of AHUXX001-006 and 008-011, respectively, were collected on the same day and in the same location as that of the holotype.One subadult specimen of AHUJGT001 was collected on 19 August 2022, from the Henan Dabie Mountains National Nature Reserve, Shangcheng County, Xinyang City.
Etymology.The specific, Gekko kaiyai sp.nov., a Latinized adjective, was named after Professor Kaiya Zhou of the School of Life Sciences, Nanjing Normal University, China, who has made great contributions to the classification of the Gekkonidae family species, especially Gekko hokouensis.The suggested common English name is "Dabie Mountains Gecko" and the Chinese name is "Dà Bié Shān Bì H ȗ", both of which indicate the location from where the new species was collected (Figure 1).body were greyish brown, with six dark brown wide irregular patches from the neck to the swollen section of the tail.Each patch diminished gradually to the sides of the body and was outlined by a dark brown border.A light grayish tubercle stripe, bordered above with dark brown, extended from the posterior corner of the eye, passing above the ear opening to the occiput.Ventral surfaces of the head, belly, and limbs were light yellow with sparse small black spots.Ground color of the tail was brown, with nine irregular white stripes, each outlined by an inconspicuous dark brown wavy border.Limbs were light grey with greyish brown bars; dorsal surfaces of the limbs were flesh red.Color of holotype in preservative.(Figure 6) In the preservative, the recently preserved specimens resemble those of the living body, and the pale grey coloration on the dorsal surface of the body and limbs becomes darker with increasing storage time.The light yellow body color fades to white with more prominent black spots.
Variation.Measurements and scale counts of the type series specimens are given in Table S4.Ground color on the dorsal surfaces of the head, body, and tail differ between individuals from yellowish grey to blackish grey in the wild, but the bodies of most individuals became darker after capture.Distribution and ecology.Gekko kaiyai sp.nov. is only known from its type-locality, Liankang Mountains National Nature Reserve, Dongzhai National Nature Reserve, and Henan Dabie Mountains National Nature Reserve in the Dabie Mountains, Henan, China (Figure 1).The new species is nocturnal, inhabits scenic fences and rocky cliffs, and is found on the walls of buildings in the countryside of low mountain and hilly areas (Figure 7).Similar habitats were identified in the survey in other areas of the Dabie Mountains.It is thus reasonable to speculate that there may be distribution in other adjacent areas of the Dabie Mountains, Anhui Province.Variation.Measurements and scale counts of the type series specimens are given in Table S4.Ground color on the dorsal surfaces of the head, body, and tail differ between individuals from yellowish grey to blackish grey in the wild, but the bodies of most individuals became darker after capture.
Distribution and ecology.Gekko kaiyai sp.nov. is only known from its type-locality, Liankang Mountains National Nature Reserve, Dongzhai National Nature Reserve, and Henan Dabie Mountains National Nature Reserve in the Dabie Mountains, Henan, China (Figure 1).The new species is nocturnal, inhabits scenic fences and rocky cliffs, and is found on the walls of buildings in the countryside of low mountain and hilly areas (Figure 7).Similar habitats were identified in the survey in other areas of the Dabie Mountains.It is thus reasonable to speculate that there may be distribution in other adjacent areas of the Dabie Mountains, Anhui Province.

Comparison
Two species of the subgenus Japonicgekko have been reported in the Dabie Mountains (Gekko hokouensis and G. japonicus).However, the new species is the sister species of G. hokouensis, they are morphologically similar, but can be distinguished from G. hokouensis by the following morphological characteristics: fewer sub-digital lamellae under the fourth toe (7-11 vs. 15-18), tubercles on the limbs (present vs. absent), more pre-cloacal pores in the males (mostly 9-12 vs. 5-9), and more tubercles between the eyes and ears (Table 4; Figures 5, 8 and 9).The new species can be easily distinguished from G. japonicus as it has fewer scale rows at the midbody (99-121 vs. 130-144), fewer sub-digital lamellae under the first and fourth toes (first toe, 8-9 vs. 10-12; fourth toe, 7-11 vs. 14-16), tubercles on the thigh (present vs. absent), and fewer dorsal midline tubercle scales (sparse vs. dense).

Discussion
The discovery of Gekko kaiyai sp.nov.has brought the total number of known species in the genus Gekko to 87 and the number identified within China to 22 [1][2][3][4][5][6]46].The definition of a species is one of the basic concerns in biological research and is thus the focus of substantial taxonomic and systematic research [47].Incorporating multiple lines of evidence to reveal the taxonomic and evolutionary relationships among species is now considered a standardized method in evolutionary biology [48].In the field investigation, we had previously mistakenly identified the new species as G. swinhonis, because two species share many common attributes, such as the number of supra-labials, infra-labials, ventral scale rows at midbody, and interorbital scales.In our study, the identification of G. kaiyai sp.nov.as a sister to G. hokouensis was surprising, as they differed greatly in their morphologies, such as tubercles on the limbs, number of pre-cloacal pores, and number of tubercles between the eyes and ears.Nevertheless, the phylogeny and morphological difference revealed distinct divergences for G. kaiyai sp.nov.when compared with G. hokouensis and G. japonicus in the Dabie Mountains.
This investigation shows that the Gekko kaiyai sp.nov.only occurs in the northwest region of the Dabie Mountains (junction of Anhui and Henan provinces), China, while G. hokouensis is distributed in the eastern region of the Dabie Mountains, and G. japonicus is distributed in the plains around the Dabie Mountains.The ecological niches of these three gecko species have clearly been differentiated in the Dabie Mountains, and G. kaiyai sp.nov.was found to be predominantly distributed on buildings, rocks, and trees in the low hills (currently, only know distributed in the northwest of the Dabie Mountains), whereas G. hokouensis was mainly distributed in houses and woodlands (distributed throughout the Dabie Mountains), and G. japonicus is mainly distributed in houses on the plains (plain areas around the Dabie Mountains).In addition, we found form our multiple observation that G. kaiyai sp.nov.had strong aggressive tendencies and its tail did not break easily.The detailed distribution range, population size, and feeding habits of these three species in the

Figure 3 .
Figure 3. Principal component analysis performed for Gekko kaiyai sp.nov.and G. hokouensis based on 10 commonly used morphological traits (except Tal, SE, RW, ML).Numbers inside the brackets indicates the percentages of the total variance explained by each axis.

Figure 3 .
Figure 3. Principal component analysis performed for Gekko kaiyai sp.nov.and G. hokouensis based on 10 commonly used morphological traits (except Tal, SE, RW, ML).Numbers inside the brackets indicates the percentages of the total variance explained by each axis.

Figure 4 .
Figure 4. Holotype AHUXXBH07 of Gekko kaiyai sp.nov. in real life: (A), left lateral view of head, (B), right lateral view of head, (C), dorsal view of middle body, (D), ventral view, (E), dorsal view of head, (E), ventral view of head, (F), ventral view of pre-cloacal region, showing six pre-cloacal pores, (G), lateral view of basal tail, (H), dorsal view of hand, (I), ventral view of hand, (J), dorsal view of foot, (K), ventral view of foot.Paratype.Sixteen specimens (two subadults, four adult females and 10 adult male), collected from the three locations by C.W. Zhang, A.F. Wu, S.L. Yu, X.N.Li, and L.R. Wang.Five specimens AHULXBH001-005 collected on 19 August 2022 from the Dongzhai National Nature Reserve, Luo County, Xinyang City, Henan Province, China.Nine adult

Figure 4 .
Figure 4. Holotype AHUXXBH07 of Gekko kaiyai sp.nov. in real life: (A), left lateral view of head, (B), right lateral view of head, (C), dorsal view of middle body, (D), ventral view, (E), dorsal view of head, (E), ventral view of head, (F), ventral view of pre-cloacal region, showing six pre-cloacal pores, (G), lateral view of basal tail, (H), dorsal view of hand, (I), ventral view of hand, (J), dorsal view of foot, (K), ventral view of foot.

Table 1 .
Samples used in the molecular analyses, including their GenBank numbers., Voucher ID and locality."N/A" means no available data.

Table 2 .
Morphological comparison of Gekko kaiyai sp.nov.and G. hokouensis.All units in mm.p-values < 0.05 significance.Morphometric characters are explained in the methods section.

Table 3 .
Variable loadings with the first four principal component, with morphometric characters corrected.

Table 4 .
Morphological comparisons among the species of the subgenera Japonigekko."-" means data unavailable; Black bold fonts represent difference with the new species.