Differentiation in cranial variables among six species of Hylopetes(Sciurinae: Pteromyini)

Song LI, Fa-Hong YU

1. Kunming Natural History Museum of Zoology, Kunming Institute of Zoology, Chinese Academy of Sciences, Kunming, Yunnan 650223, China

2. State Key Laboratory of Genetic Resources and Evolution, Kunming Institute of Zoology, Chinese Academy of Sciences, Kunming, Yunnan 650223, China

3. ICBR, University of Florida, Gainesville, FL 32610, USA

The genus Hylopetes (Thomas, 1908) mainly inhabits mountain forests in Asia, ranging from the Himalayas to the Greater Sunda Islands, and including areas of Nepal, China, Indochina, Philippines, Malaysia,and Indonesia (Corbet & Hill, 1992; Hoffmann et al,1993; Nowak, 1999; Thorington & Hoffmann, 2005).Initially, Hylopetes was regarded as a subgenus of Sciuropterus (Thomas, 1908), but, based on Pocock’s(1923) studies on the morphology of baculum, it was elevated to a full genus. This view, that the arrow-tailed flying squirrel Hylopetes in the eastern extreme of the Himalayas and southeastern Asia is a valid genus from the small Kashmir flying squirrel Eoglaucomys (A. H. Howell,1915) in the western Himalayas, has been sustained by subsequent morphological and molecular studies (Nowak,1999; Oshida et al, 2004; Thorington et al, 1996).

Based on several dental and cranial characteristics as well as external structures, different taxonomists have described varying numbers of Hylopetes species (Allen,1940; Corbet & Hill, 1992; Ellerman, 1940; Ellerman &Morrison-Scott, 1950; Hoffmann et al, 1993; Nowak,1999; Thorington & Hoffmann, 2005). To date, the maximal number of Hylopetes species was put at 13(Ellerman, 1940), though the latest accepted result is 9(Thorington & Hoffmann, 2005).1

Allen (1940) described two subspecies of Pteromys(Hylopetes) alboniger (Allen, 1925), and listed the Hainan flying squirrel as Pteromys electilis (Allen, 1925).Ellerman (1940) listed 13 Hylopetes species, including H.platyurus (Jentink, 1890), H. spadiceus (Blyth, 1847), H.phayrei (Blyth, 1859), H. alboniger (Hodgson, 1836), and H. nigripes (Thomas, 1893) etc., but the Hainan flying squirrel from Namfong, Hainan Island, China was regarded as Petinomys electilis (Allen, 1925). Ellerman &Morrison-Scott (1950) accepted the same results of Petinomys electilis as Ellerman (1940), however, they listed only five Hylopetes species, including H. spadiceus,H. phayrei, H. alboniger, H. fimbriatus (Gray, 1837), and H. sagitta (Linnaeus, 1766). The last two species were emended in later studies: H. fimbriatus was listed as Eoglaucomys fimbriatus and H. sagitta was replaced by H.lepidus (Horsfield, 1822) in the results of Thorington &Hoffmann (2005). Corbet & Hill (1992) regarded H. p.electilis as subspecies of H. phayrei, and Sciuropterus platyurus was accepted as the synonym of H. lepidus(Horsfield, 1822), resulting in their listing of ten Hylopetes species, except for H. fimbriatus and H. baberi (Blyth,1847), which were arranged as Eoglaucomys fimbriatus,and H. platyurus (Thorington & Hoffmann, 2005). The rest of the eight Hylopetes species were the same as the results of Thorington & Hoffmann (2005).

Despite lengthy and protracted debate, there remains several disagreements and uncertainties as to the taxonomy of this polymorphic genus. Among the most contentious include H. electilis in Hainan Island, China and H. alboniger, H. phayrei, H. platyurus, H. spadiceus,and H. nigripes, which have had a complicated taxonomic history due to intraspecific geographic variation in their distributions across Asia (Corbet & Hill, 1992;Hoffmann et al, 1993; Rasmussen & Thorington, 2008;Roberts, 1997; Thorington & Hoffmann, 2005). Despite the controversies, to date no attempt has been made to investigate the differences in measurable skull variables among these groups.

Morphometric data has been used to evaluate cranial, dental, and body measurements of various wild populations of mammals, and the resulting statistic analysis of variations on the morphological level has been proved useful in detecting patterns of geographic variations and delimiting intra- or inter-specific evolutionary units (Li et al, 2008; Li et al, 2012; Munoz& Perpinan, 2010; Slabova & Frynta, 2007; Zelditch et al,2004). Because the specimens of Hylopetes are undoubtedly exiguous, we based our analysis on 89 specimens of 6 Hylopetes species. Principal components analysis (PCA) was performed to compare skull measurable variables among them, and furthermore,special attention was paid to the taxonomic status of Hainan flying squirrel from Namfong, Hainan Island,China, which has been listed as either a synonym or subspecies of H. phayrei (Corbet & Hill, 1992;Hoffmann et al, 1993; Thorington & Hoffmann, 2005).

MATERIALS AND METHODS

Samples

A total of 89 specimens of Hylopetes were examined and used for statistical analysis, including 39 females, 44 males, and 6 specimens of unknown sex (Appendix A).Adults were classified based on phalange ossification and dental patterns. All specimens used in this study were deposited in the American Museum of Natural History(New York, AMNH); the National Museum of Natural History (Washington DC, USNM); the Institute of Zoology,Chinese Academy of Sciences (IOZ, CAS) (Beijing, China);and the Kunming Natural History Museum of Zoology,Kunming Institute of Zoology (KNHMZ, KIZ, CAS)(Kunming, China). In total, 14 measurable cranial variables (Figure 1) taken with a digital caliper to the nearest 0.01 mm were used in statistic analysis.

Figure 1 Measurements of 14 cranial variables used in the study

Data analysis

Statistical analysis of the morphometric study, which included sexual dimorphism, multiple comparisons, and principal component analysis (PCA) were performed using SPSS 11.0 (SPSS Inc., Chicago, USA). First, all variables were transformed into natural logarithms to eliminate the bias effect on large measurements in statistical analysis (D’elia & Pardinas, 2004). Then,Kaiser-Meyer-Olkin Measure of Sampling Adequacy(KMO) and Bartlett’s Test were performed in order to test if the data were fit for analysis. Statistical differences were considered significant where P＜0.05.

The Tests of Equality of Group Means (by Wilks’Lambda) was used to assess the sexual dimorphism of each group. In order to evaluate variations between samples, multiple comparisons (by Least-Significant Difference) between taxa were performed for all 14 cranial measurements. The associations between cranial measurable characters and species were assessed by multivariate analysis of the PCA. PCA based on the variancecovariance matrix of the log-transformed variables was performed to identify variables that account for maximum variation in the data set, and to represent distances between major groups in order to assess the specific relationships among the individuals. The eigenvector scores describing the relative significance of each variable to principal components were used to compare the cranial morphological similarities and differences. The PCA scatter-plot visually represents the variation among different individuals of the samples (Figure 2).

Figure 2 Plots of six Hylopetes species samples on the principal components factors 1 and 2

RESULTS

The means and standard deviation (SD) of 14 measurable cranial variables of Hylopetes are presented in Table 1. The results of KMO is 0.946 and that of P of Bartlett’s Test of Sphericity is 0.000, together confirming that the data garnered in the morphometric study were fit for PCA analysis. The Tests of Equality of Group Means for 83 samples showed no sexual dimorphism in all 14 cranial variables in the 6 Hylopetes species (Table 2). Table 3 presents the results of multiple comparisons on all 14 cranial variables between the species of Hylopetes.

Multivariate analysis was based on principal components analysis. In PCA, eigenvalues for the first three principal components were 12.161, 0.740 and 0.541 respectively, accounting for 96.014% of the total variance (Table 4). The first principal component factor(PC1) accounted for 86.867% of the total original variance and most variables had high positive loadings,suggesting the size variation in the samples. The second principal component (PC2) accounted for 5.284% of the total variance and is strongly correlated with TBL, with factor loadings＞0.9. Along the third principal component factor (PC3), which accounts for 3.863% of the total variance, the main morphological variable contributing to this association is PORCL, with factor loadings＞0.9.

Because the eigenvalue of the first principal components was 12.161, and the third was much lower(0.541), we used PC1 and PC2 to make the PCA scatterplot (Figure 2).

（2）材料加工区布设：对于材料加工区域，主要包含钢筋加工区、成型钢材焊接区、模板加工区、预制构件制作区、施工工具放置区等。在安全控制管理过程中：①将已经加工好的材料放置在容易取用的区域；②对加工区域做好防水工作，避免材料受潮、受淋，产生损坏；③对于易燃的施工材料，设置了消防去，配备了齐全的消防设施，并张贴安全指示牌，提醒施工人员注意。

Table 1 Cranial variable measurements of Hylopetes (mean±SD)/range. Variable codes explained in Figure 1

Figure 2 indicates PC1-PC2 scatter plots of principal component analysis. Inspection of Figure 2,samples of H. electilis, H. platyurus, H. spadiceus, and H.phayrei are distinguished from each other to form their distinct groups, showing different morphological characteristics of their skulls. However, results showed that samples of H. alboniger and H. nigripes are quite close to one another with considerable overlap,suggesting that H. alboniger shares similar measurable skull characteristics with H. nigripes. An interesting result was that the three samples named H. electilis from Chiang Mai, Thailand (labeled as “±” in Figure 2) are separated from those from Namfong, Hainan Island, but scatter in the samples of H. phayrei, hinting that the samples of H. electilis from Chiang Mai, Thailand should be more similar with those of H. phayrei in measurable skull characteristics.

DISCUSSION

Based on the specimens from Namfong, Hainan Island, China, Allen (1925) first described a valid species named Pteromys (Petinomys) electilis, but later on review classified it as Pteromys electilis (Allen, 1940).Later studies by Ellerman (1940) and Elleuman &Morrison-Scott (1950) continued to use Petinomys electilis as the scientific name for this species, and emphasized the species distribution on Hainan Island,China. However, based on their understanding of its two septa in each bulla, Corbet & Hill (1992) listed the group in Hylopetes. While Corbet & Hill (1992) narrated significant differences on pelage characteristics between H. electilis and H. phayrei, they still opted to list it as belonging to H. phayrei, while placing H. p. electilis as its subspecies. More succinctly put, the validity of the Hainan flying squirrel, H. electilis, has been controversial for quite a long time (Corbet & Hill, 1991,1992; Ellerman, 1940; Hoffmann et al, 1993; Nowak,1999; Thorington & Hoffmann, 2005; Wang, 2003;Zhang et al., 1997).

The validity of various measurable cranial variables in distinguishing between species and species groups in Hylopetes has been critical in several studies. For example, Rasmussen & Thorington (2008) discussed the distributions for H. platyurus, H. spadiceus, and H.Lepidus using similar methods. In the present study, we performed statistical analyses on 14 measurable cranial variables from 22 samples of H. electilis, 19 of which came from Namfong, Hainan Island, China, in order to compare them with the samples of other five other validated Hylopetes species. The results showed significant difference between H. electilis and the other five Hylopetes species (Table 3), indicating it could be distinguished from other five species by using measurable skull variables. Similarly, though Figure 2 indicates that samples of H. electilis from Namfong,Hainan Island could be clearly distinguished from those of other five Hylopetes species, we found that they were closest to samples of H. phayrei, as opposed to the other four Hylopetes species. This result not only sustains the idea that H. electilis on Hainan Island should be a valid species, but it also may explain the reasons why there were so many systematic disagreements between H.electilis and H. phayrei.

Table 2 Tests of Equality of Group Means (by Wilks' Lambda) for Sexual Dimorphism

Table 3 Multiple Comparisons (by Least-Significant Difference) of all 14 cranial variables between the six Hylopetes species

Table 4 Factor loadings of the principal component analysis

Both a variety of previous studies as all as the present study have established that the Hainan flying squirrel, H. electilis, is restricted to Hainan Island, China(Ellerman, 1940; Ellerman & Morrison-Scott, 1950), and we thought that the channel isolation may have accelerated its absolute evolution. Based on molecular data, Jorgensen & Demarais (1999) and Oshida et al(2000) discussed the close genetic relationship and the recent divergence between H. electilis and H. phayrei.Their results suggest that these populations rapidly extended the ranges to their present distributions in a relatively short time (Jorgensen & Demarais, 1999;Oshida et al, 2000). Based on our current findings and the previous data, we recommend further correlative ecology studies on H. electilis and H. phayrei should be performed in order to gather more data that would help explain their adaptation process, and in doing so explain their phenotypic differentiation.

Additionally, three samples of H. electilis (labeled as “±” in Figure 2) from Chiang Mai, Thailand, were found to be not only clearly different from those of H.electilis from Namfong, Hainan Island, China, but they also scatter within samples of H. phayrei to form a close group (Figure 2). Based on the results of pelage comparison between H. electilis and H. phayrei, while Corbet & Hill (1992) listed the former as H. p. electilis while also emphasizing some significant differences in pelage characteristics between the two (Corbet & Hill,1992). Our results also indicate that samples of H.electilis from Hainan Island could differ from those of H.phayrei samples from mainland in terms of skull measurable characteristics (Figure 2), and likewise support their earlier results that H. electilis is distributed across Hainan Island, China. (Ellerman, 1940; Ellerman& Morrision-Scott, 1950).

The ultimate results from our statistical analysis of the skull characteristics sustains the argument that H.electilis is neither a synonym nor subspecies of H.phayrei, but a valid species in its own right. As for H.platyurus, although it was accepted as the synonyms of H. lepidus (Corbet & Hill, 1992), Figure 2 indicates that specimens of H. platyurus could be completely separated from five other Hylopetes species, similarly illustrating that it likely must also belong to a valid Hylopetes species. Unfortunately, without specimens of H. lepidus to serve as a comparison, the data collected in the present paper is suggestive but not conclusive, therefore we cannot definitively state whether H. platyurus is a synonym of H. lepidus or not. Further studies with more specimens may help further settle the ongoing debates regarding the taxonomy.

For a final note regarding geography, it is worth noting that Hylopetes are distributed in different zoogeographic subregions of the Oriental region, each with significant geographical variations (Corbet & Hill, 1992;Lekagul & Mcneely, 1988; Zhang et al, 1997). Accordingly,the samples of the six Hylopetes species used in this study can be morphologically distinguished as different, distinct groups (Figure 2). H. alboniger and H. phayrei belong to the Indochinese subregion and overlap extensively in Thailand, Indochina, Burma, and southern China (Corbet& Hill, 1992; Hoffmann et al, 1993; Thorington &Hoffmann, 2005), where the habitat is tropical, subtropical,or evergreen forest, with low elevation (500-1500 m) and within reach of southwest and south monsoons.Additionally, the population of H. alboniger occurs in the eastern extreme of the Himalayas and occupies temperate coniferous forest at mid elevations (1 500-3 000 m) or deciduous and subtropical forests at low elevations, while H. nigripes is endemic to the Philippines and occurs in subtropical or tropical dry forests (Nowak, 1999). Similar morphological characteristics (Figure 2) between H.alboniger and H. nigripes may potentially be adaptations to similar living habitats, though again, further studies are needed to explore this possibility.

Appendix Ⅰ

Specimens examined in this study

H. alboniger n=15

Yunnan, China: IOZ 5507 ♀, 5508 ♂, 15046 ♀; KIZ 00009 ♀,76311 ♂, 84756, 89011, 90004, 206746, 206745,640222 ♂, 640227 ♂; AMNH 114884 ♂, 114885 ♂,114886 ♀;H. electilis n=22

Hainan, China: AMNH 58138 ♀, 58158 ♀, 58159 ♀, 58160 ♂,58161 ♀, 58162 ♂, 58163 ♂, 58164 ♀, 58166 ♀,58167 ♂, 58171 ♂, 58172 ♂, 58176 ♂, 58179 ♀,58180 ♂, 58181 ♂, 58182 ♀, 58186 ♀, 58198 ♂;Chiang Mail, Thailand: AMNH 167891 ♀, 167892, 167893 ♂;H. nigripes n=13

Palawan, Philippines: AMNH 242098 ♀, 203314 ♀, 203313 ♂,203311 ♀, 203309 ♂, 203310 ♂;

USNM 477991 ♂, 477993 ♀, 477996 ♂, 478005 ♀, 478006 ♀,478007 ♀, 478009 ♂;H. platyurus n=10

Selamgger, W Malaysia: USNM 488617 ♀, 488619 ♀, 488620♂, 488623 ♂, 488624 ♀, 488625 ♀, 488626 ♀,488630 ♂, 488631 ♂, 488633 ♂;H. spadiceus n=10

Johore, W Malaysia: USNM 481109 ♂, 481110 ♀, 481112 ♀,481114 ♂, 481116 ♂, 488648 ♂, 488638 ♂, 488641♀, 488645 ♀, 488646 ♀;H. phayrei n=19

Banlad, Siam: USNM 294888 ♂, 294890 ♂, 294891 ♂, 294892♀, 297089 ♀;

Chiang Mai, Siam: USNM 260621 ♀, 260622 ♂, 260624 ♂,153580 ♀;

Mandalaypopa, Burma: AMNH 163552 ♀, 163553 ♀, 163554♂, 163555 ♂, 163557 ♀, 163558 ♀, 163561 ♂,163559 ♂, 163560 ♂, 163563 ♂.

Allen GM. 1940. The Mammals of China and Mongolia. Natural History of Central Asia. Vol. XI, Part 2. New York: American Museum of Natural History.

Corbet GB, Hill JE. 1991. A World List of Mammalian Species. 3rd ed.New York: Oxford University Press.

Corbet GB, Hill JE. 1992. The Mammals of the Indomalayan Region:A Systematic Review. New York: Oxford University Press.

D’ella G, Pardinas UFJ. 2004. Systematics of Argentinean, Paraguayan,and Uruguayan swamp rats of the genus Scapteromys (Rodentia,Cricetidae, Sigmondontinae). Journal of Mammalogy, 85(5): 897-910.Ellerman JR. 1940. The Families and Genera of Living Rodents with a List of Named Forms (1758 to 1936). Vol. I. London: British Museum(Natural History).

Ellerman JR, Morrison-Scott TCS. 1950. Checklist of Palaearctic and Indian mammals, 1758 to 1946. London: British Museum (Natural History).

Hoffmann RS, Anderson CG, Thorington RW Jr, Heaney LR. 1993.Family Sciuridae. In: Wilson D E, Reeder D M. Mammal Species of the World: Taxonomic and Geographic Reference. 2nd ed. Washington,D. C: Smithsonian Institution Press.

Jorgensen EE, Demarais S. 1999. Spatial scale dependence of rodent habitat use. Journal of Mammalogy, 80(2): 421-429.

Lekagul B, Meneely JA. 1988. Mammals of Thailand. 2nd ed. Bangkok:Darnsutha Press.

Li S, Wang YX, Yang JX. 2008. Geographic variation of the Perny’s Long-nosed squirrels (Dremomys pernyi) (Milne-Edwards,1867) (Rodentia: Sciuridae) from southwestern China based on cranial morphometric variables. Belgian Journal of Zoology,138(1): 95-100.

Li S, Yu FH, Lü XF. 2012. Cranial morphometric study of four giant flying squirrels (Petaurista) (Rodentia: Sciuridae) from China.Zoological Research, 33(2): 119-126.

Munoz FM, Perpinan D. 2010. Measurement error in morphometric studies: comparison between manual and computerized methods.Annales Zoologici Fennici, 47(1): 46-56.

Nowak RM. 1999. Walker’s Mammals of the World. 6th ed. Baltimore:Johns Hopkins University Press.

Oshida T, Lin LK, Yanagawa H, Endo H, Masud R. 2000. Phylogenetic relationships among six flying squirrel genera, inferred from mitochondrial cytochrome b gene sequences. Zoological Science, 17(4): 485-489.

Oshida T, Shafique CM, Barkati S, Yasuda M, Hussein NA, Endo H,Yanagawa H, Masud R. 2004. Phylogenetic position of the small Kashmir flying squirrel, Hylopetes fimbriatus (= Eoglaucomys fimbriatus), in the subfamily Pteromyinae. Canadian Journal of Zoology, 82(8): 1336-1342.

Kunming Institute of Zoology (CAS), China Zoological Society Volume 34 Issues E4-5

Pocock RI. 1923. The classification of Sciuridae. Proceedings of the Zoological Society of London, 1923(1): 209-246.

Rasmussen NL, Thorington RW. 2008. Morphological differentiation among three species of flying squirrels (Genus Hylopetes) from Southeast Asia. Journal of Mammalogy, 89(5): 1296-1305.

Roberts TJ. 1997. The mammals of Pakistan (revised edition). New York: Oxford University Press.

Slabova M, Frynta D. 2007. Morphometric variation in nearly unstudied populations of the most studied mammal: the noncommensal house mouse (Mus musculus domesticus) in the near East and Northern Africa. Zoologischer Anzeiger: A Journal of Comparative Zoology, 246(2): 91-101.

Thorington RW Jr, Hoffmann RS. 2005. Family Sciuridae. In: Wilson E,Reeder DM. Mammal Species of the World: Taxonomic and Geographic Reference. 3nd ed. Washington, D.C.: The Johns Hopkins niversity Press.

Thorington RW Jr, Musante AL, Anderson CG, Darrow K. 1996.Validity of three genera of flying squirrels: Eoglaucomys, Glaucomys,and Hylopetes. Journal of Mammalogy, 77(1): 69-83.

Thomas O. 1908. The nomenclature of the flying lemurs. Annals and Magazine of Natural History, 8(1): 252-255.

Wang YX. 2003. A Complete Checklist of Mammal Species and Subspecies in China, A Taxonomic and Geographic Reference. Beijing:China Forestry Publishing House.

Zhang YZ, Jing SK, Quan GQ, Li SH, Ye ZY, Wang FG, Zhang ML.1997. Muridae. In: Li W. Distribution of Mammalian Species in China.Beijing: China Forestry Publishing House.

Zelditch ML, Swiderski DL, Sheets HD, Fink WL. 2004. Geometric Morphometrics for Biologists: A Primer. New York: Elsevier Academic Press.