Geographical Variation of the Skull of the Lesser Mouse Deer

FULL PAPER Anatomy

Hideki ENDO1), Katsuhiro FUKUTA2), Junpei KIMURA3), Motoki SASAKI4) and Brian J. STAFFORD5,6)

1)Department of Zoology, National Science Museum, Tokyo, 3–23–1 Hyakunin-cho, Shinjuku-ku, Tokyo 169–0073, 2)Laboratory of Animal Morphology and Function, Graduate School of Bioagricutural Sciences, Nagoya University, Nagoya, Aichi 464–8601, 3)Department of Veterinary Anatomy, College of Bioresource Sciences, Nihon University, Fujisawa, Kanagawa 252–8610, 4)Department of Veterinary Anatomy, Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido 080–8555, Japan, 5)Mammal Division, National Museum of Natural History, Smithsonian Institution, Washington DC. and 6)Department of Anatomy, Howard University College of Medicine, Washington DC, U.S.A.

(Received 21 January 2004/Accepted 1 June 2004)

ABSTRACT. We examined the geographical variation of the skull size and shape of the lesser mouse deer (Tragulus javanicus) from Laos, Thailand, Peninsular Malaysia, Sumatra, Java, Borneo, Langkawi and some Islands of Tenasserim in Myanmar. Although the influence of the climatic condition on skull size was not confirmed in the mainland populations, the skull became rostro-caudally longer in the populations of Tenasserim and Sumatra because of island isolation effect. The skull size was classified into the following three clusters of localities from the matrix of Q-mode correlation coefficients: 1) Langkawi and Tenasserim, 2) Laos and Thailand, 3) Sumatra and Borneo. The skulls in the population of Java belong to the cluster of Langkawi and Tenasserim in male, however were morphologically similar to those in the cluster of Borneo and Sumatra. The canonical discriminant analysis pointed out that the Laos and Tenasserim populations were separated from the other ones and that the populations of Sumatra, Java and Borneo were intermingled each other. KEY WORDS: geographical variation, lesser mouse deer, osteometry, skull, Tragulus javanicus. J. Vet. Med. Sci. 66(10): 1229–1235, 2004

The lesser mouse deer (Tragulus javanicus) is largely dis- ness of the skull suture was not examined in this study. tributed in the Indomalayan region including Thailand, Skull measurement was carried out utilizing vernier calli- Cambodia, Laos, Vietnam, Myanmar, Peninsular Malaysia, pers until the nearest 0.05 mm. The measurements, which Sumatra, Java, Borneo and some islands [14]. Some races were based on Driesch [5] and Duerst [6], are defined in have been described within Tragulus javanicus and the tax- Table 2. Since the incisive bone had been easily broken and onomic discussion was confusing about this species [1, 2, lost in this species, we used the most cranial point of the 12, 13, 15–17, 21]. Although the conflicting studies have maxilla instead of the Prosthion of the incisive bone in the dealt with skin color, external measurements and skull mor- four length measurements (PL2, BL2, SL2 and MPL2) phology in a few individuals at the subspecies level, the sta- (Tables 1 and 2). tistical analysis of the skeleton [23] and the taxonomical The statistical analysis was undertaken in each sex. To reviews [3, 24] have pointed out that only two species of tra- clarify the geographical tendency in size among the popula- gulids, Tragulus javanicus and Tragulus napu, should be tions of Laos, Tenasserim, Peninsular Malaysia, Sumatra, recognized in the Indomalayan region at the species level. Java and Borneo including many specimens, statistical dif- We reported the growth pattern of the skull in the lesser ference among mean values of each measurement was ana- mouse deer [20], however the geographical variation has lyzed by Student’s t-test. Using the measurement data the remained unclear in the skull morphology in Tragulus jav- clustering analysis was carried out to evaluate the osteomet- anicus. rical similarities between populations. Since we also applied the specimens without incisive bone for clustering MATERIALS AND METHODS analysis, the 21 measurements except for PL1, BL1, SL1, and MPL1 were used. The clustering was performed by We examined totally 221 skulls of the lesser mouse deer UPGMA method for the distance matrix converted from the (Tragulus javanicus). These specimens are housed in the matrix of the Q-mode correlation coefficients to avoid loss Smithsonian Institution (Washington DC, U.S.A.), Ameri- of information [19]. The distance matrix and the dendro- can Museum of Natural History (New York, U.S.A.), gram between localities were obtained by using the software Nationaal Natuurhistorische Museum (Leiden, The Nether- of Statistica (StatSoft, Inc, Tokyo, Japan). The canonical lands), and National Science Museum, Tokyo (Tokyo, discriminant analysis was also undertaken using the 7 local- Japan) (Appendices 1 and 2). Sex determination was depen- ity groups except for Thailand population, since the male dent on the description of biological data and the canine size population from Thailand including only one specimen of the specimens. We selected specimens that possessed the could be applied to the analysis. The first two canonical dis- fully erupted lower third molar and considered them adult. criminant functions were calculated and plotted by using the The composition of the origin and the sex is shown in Table software of Statistica. 1. Each locality is indicated in the map (Fig. 1). The close- 1230 H. ENDO ET AL.

Appendix 1. Specimen No. used in this study (male) Symbol L Vientiane National Science Museum, Tokyo 31072 31085 31086 31087 31088 31092 31099 31109 31110 31114 31125 31127 T Surathani Smithsonian Institution 255718 E Tenasserim (James Id., Kisseraing Id., Mergui Id., St. Matthews Id., and St. Luke's Id.) Smithsonian Institution 104419 104423 104424 104426 104427 104428 124029 124038 124145 124178 M Peninsular Malaysia Smithsonian Institution 115371 115372 115376 283560 489437 489440 489442 489445 SSumatra Smithsonian Institution 114328 114329 114331 114339 114340 114419 114420 114426 114427 143318 143321 143488 143489 143497 143500 143501 143502 143503 143504 143514 143515 143516 143517 143518 143523 143524 144140 144145 144146 144150 144153 144155 144157 144313 144356 144357 144367 144368 144370 145325 145326 American Mueum of Natural History 106551 106553 Nationaal Natuurhisorische Museum, Leiden 4600 11811 33768 JJava Smithsonian Institution 194246 American Mueum of Natural History 102090 Nationaal Natuurhisorische Museum, Leiden* 172 4936 33767 33778 194246 T-102 T-105 T-106 BBorneo Smithsonian Institution 19191 145354 145355 145357 145358 145359 145360 145361 153740 153742 153744 153746 153747 153748 153752 153754 154350 197687 197688 197689 American Mueum of Natural History 103696 103696 106070 106297 107110 107111 Nationaal Natuurhisorische Museum, Leiden* T-18 T-20 T-27 T-28 A Langkawi Id. Smithsonian Institution 104404 104405 104406 104407 * The numbers with the symbol T are temporary for this study.

RESULTS pared between the M mainland population and the E, S, and B island ones. The size-related measurements such as PL1, Mean values and standard deviations in 25 measurements PL2, BL1 and BL2 were significantly larger in E and S pop- are shown in Table 3. The data on the significant differ- ulations and females of J population than in M population ences between localities are arranged in Table 4. (Tables 3 and 4). In contrast, the skull size of Borneo popu- We compared the L mainland population with M to con- lation was not significantly different from that of Peninsular firm the influence of the climatic conditions on the skull size Malaysia population. OZW and AZW were not signifi- by the following reasons: the distribution of the two popula- cantly different between the M population and the island tions is obviously separated into the north and south districts populations unlike the measurements on the entire length. of the Indochinese and Malayan Peninsulae; the conditions As for the length measurements of the splanchnocranium of the temperature and precipitation in Vientiane District are the mean values were significantly smaller in SL1, SL2, different from those in Peninsular Malaysia. The mean val- MPL1 and MPL2 in M population than those in E and S ues of PL1, PL2, BL1 and BL2 in both sexes were larger in populations. However, the other comparisons between the the Laos population than those in the Peninsular Malaysia M population and the island populations did not obviously one, though the significant difference was not found in some show statistical differences. The values of LBO and LBS cases (Tables 3 and 4). were significantly larger in S population than in M one, The island-isolation effects on skull size could be com- which indicated the development in the width of the SKULL VARIATION OF LESSER MOUSE DEER 1231

Appendix 2.Specimen No. used in this study (female) Symbol L Vientiane National Science Museum, Tokyo 31076 31077 31078 31081 31103 31104 31105 31106 31107 31108 31122 31124 T East Thailand Smithsonian Institution 236613 253408 258875 E Tenasserim (James Id., Kisseraing Id., Mergui Id., St. Matthews Id., and St. Luke’s Id.) Smithsonian Institution 104431 104420 124025 124026 124027 124030 124035 124037 124040 124085 124086 124179 124245 M Peninsular Malaysia Smithsonian Institution 115368 115369 115373 489439 489444 SSumatra Smithsonian Institution 114330 114332 114333 114338 114421 114422 114424 114425 121697 121698 121838 123108 143319 143493 143494 143495 143496 143498 143499 143506 143507 143508 143509 143511 143512 143520 143521 143522 143528 144143 144148 144149 144151 144154 144156 144158 144316 144317 144369 Nationaal Natuurhisorische Museum, Leiden 3777 4661 37770 JJava Smithsonian Institution 156474 156484 268293 268296 Nationaal Natuurhisorische Museum, Leiden* 4933 4935 14523 T-104 B Borneo Smithsonian Institution 142348 145356 145362 153739 153741 153745 153749 153751 197686 American Mueum of Natural History 32645 103700 106299 106300 106302 106303 A Langkawi Id. Smithsonian Institution 104408 104409 104411 * The numbers with the symbol T are temporary for this study.

Table 1. Origin and sex composition of the specimens in both sexes Origin of Specimens Symbol Male Male* Female Female* Laos Vientian L 8 4 7 5 Thailand East part and Surathani T 1 0 3 0 Burma Tenasserim** E 10 0 12 1 Malaysia Peninsular Malaysia M 8 0 5 0 Sumatra Id. S 43 3 38 4 Java Id. J 9 0 8 0 Borneo Id. B 24 6 11 4 Lankawi Id. A 4 0 3 0 Total 107 13 87 14 * Specimens lacking incisive bone. ** The locality means St. Luke’s, James’, St. Matthews, Mergui, and Kisseraing Islands located in Anda- man Sea along the coast of the South Myanmar. splanchnocranium and in the length of the neurocranium in tally similar to that of female, although the important differ- the skulls from Sumatra. The size-tendency was not con- ence was demonstrated only in the J population. J firmed in the width measurement of the neurocranium such population was connected to the M-B-S cluster in female as GNB and GMB between the M population and the island dendrogram, whereas to E-A cluster in male. ones. The results of the discriminant analysis were visualized The dendrogram and the distance matrix are shown in (Fig. 3). In both sexes, the L and E populations were sepa- Fig. 2 and Table 5. The dendrogram of male was fundamen- rated from the plots of the other populations. The plots of 1232 H. ENDO ET AL.

the M, S, and B populations were intermingled each other. The obvious tendencies were not found in the plot distribution of the J and A populations.

DISCUSSION

The localities of Tenasserim represent the small islands named St. Luke’s, James’, Mergui, and Kisseraing located in Andaman Sea along the coast of South Myanmar. We could not confirm the tendency such as Bergmann’ s rule between the L population and the S one from the mainland. However, the skull length became larger in the populations of Tenasserim and Sumatra in comparison with the M population because of the island-isolation effect. This result is not consistent with the suggestions [22] that the dwarfism is confirmed in some island populations of Tragulus napu, and that the isolation effect is not obviously seen in Tragulus javanicus. From these results, we suggest that Tragulus javanicus becomes larger in islands like smaller rodents, whereas the Tragulus napu is smaller like other large artio- dactyls [7, 8, 11, 22]. In our clustering analysis, the shape factor could be effi- ciently extracted by the multivariate statistical analysis using the Q-mode correlation coefficients, since the method evaluates the true morphometric similarity within each form Fig. 1. Map showing the specimen localities in the Indomalayan [18, 19]. We emphasized the shape factor for its importance Region. The locality symbols are explained in Table 1. The two in the clustering studies, since phylogenetical relationships symbols of T indicate the East Thailand District and Surathani among close-related forms are well reflected in the affinities Area in Peninsular Thailand. Arrow, the Isthmus of Kra. of overall shape morphology [4]. Table 2. List of measurements and their abbreviations 1.Cranium *Profile length PL1 *Basal length BL1 *Short lateral facial length SL1 From the most rostral point of maxilla to Akrokranion PL2 From the most rostral point of maxilla to Basion BL2 Short lateral facial length from the most rostral point of the maxilla SL2 Oral zygomatic width OZW Aboral zygomtic width AZW Least breadth between the orbits LBO Greatest neurocranium breadth GNB *Median palatal length MPL1 Median palatal length from the most rostral point of the maxilla MPL2 Length from Basion to Staphylion LBS Dental length DL Length of the molar row LMR Greatest palatal breadth GPB Greatest mastoid breadth GMB Greatest breadth of the occipital condyle GBO Height from the most caudal point of the occipital bone to Basion HOB

2.Mandible Length from the condyle LC Length from the angle LA Length of the horizontal ramus LHR Length of the molar row LMRM Oral hight of the vertical ramus OHR

**Height of the mandible at M1 HM The measurements are fundamentally based on Driesch [5]. * The items cannot be measured in specimens lacking incisive bone. ** The measurement is based on Duerst [6]. SKULL VARIATION OF LESSER MOUSE DEER 1233

Table 3. Mean values (mm) and standard deviations for craniometric measurement in various populations

Symbol PL1 BL1 SL1 PL2 BL2 SL2 OZW AZW LBO GNB MPL1 MPL2 LBS DL LMR GPB GMB GBO HOB LC LA LHR LMRM OHR HM male L 82.09 94.56 44.34 76.00 87.80 37.15 41.70 40.68 25.68 31.82 58.49 52.29 23.62 47.48 34.31 27.44 27.29 20.50 20.15 73.77 72.65 53.37 36.83 27.36 8.72 1.97 2.30 1.27 2.65 3.05 1.58 1.72 1.63 0.97 0.90 1.55 1.85 1.07 1.44 1.33 1.04 0.94 0.72 2.58 2.79 2.58 1.78 1.17 1.22 0.69 T 84.70 95.45 45.20 77.70 87.90 36.95 42.45 41.25 25.60 31.70 60.50 52.95 24.85 48.25 33.45 27.90 28.70 19.60 20.90 75.75 76.00 55.40 37.00 30.45 9.55 E 83.56 94.11 44.49 77.68 88.24 37.92 44.11 42.21 26.19 30.60 58.96 53.43 24.53 47.76 33.54 28.71 28.98 19.60 18.63 75.33 74.97 54.87 38.25 31.75 9.67 2.07 2.52 1.21 1.75 2.26 1.19 1.81 1.75 1.42 1.36 1.62 1.48 1.16 0.71 1.21 0.99 0.79 0.44 0.88 2.32 2.37 1.14 1.33 1.81 0.76 M 81.21 91.73 43.51 75.68 85.87 37.64 43.07 42.18 25.71 31.62 57.73 52.58 23.31 46.32 32.52 28.01 29.94 19.81 18.34 73.88 74.30 54.03 36.13 29.58 8.86 1.16 1.71 1.34 0.96 1.68 1.67 0.96 1.38 0.83 0.63 1.18 1.18 1.23 0.68 2.12 0.50 0.89 0.84 1.53 1.24 1.13 0.59 1.37 1.18 0.77 S 84.14 95.50 45.28 77.93 89.38 38.55 42.63 41.86 26.70 31.69 59.25 53.54 24.66 47.43 33.11 28.13 29.46 20.19 18.79 75.84 76.21 54.97 37.03 29.56 9.39 3.84 3.57 2.10 3.66 3.45 1.94 1.61 1.53 1.16 0.75 2.94 2.78 1.56 2.04 1.53 1.32 0.89 0.80 0.99 3.54 3.78 2.38 1.56 1.93 0.73 J 82.26 94.09 43.31 75.72 87.33 35.66 42.53 42.11 25.69 31.12 58.57 52.24 23.69 47.62 33.09 28.21 29.14 20.33 17.63 73.82 73.86 54.14 37.13 30.04 9.47 3.00 2.80 1.72 2.56 2.62 1.08 1.70 1.59 0.96 0.97 2.24 1.99 0.97 1.37 0.90 1.19 1.38 0.83 1.01 2.27 2.31 1.68 1.77 1.48 0.77 B 82.69 94.05 44.20 76.32 87.36 37.52 42.02 41.03 26.17 31.24 58.76 52.60 23.74 46.38 31.90 27.73 29.06 20.40 18.74 73.89 73.93 54.04 35.85 28.52 8.73 4.55 5.22 3.46 3.79 4.38 2.44 2.11 2.16 1.09 1.38 3.98 3.15 1.24 2.49 1.84 1.43 1.27 1.08 1.42 3.94 4.32 2.99 1.91 2.28 0.73 A 82.36 90.19 44.51 76.53 86.56 37.64 43.56 42.03 26.74 29.79 59.14 53.30 23.16 47.71 34.76 29.03 28.34 18.83 18.60 75.46 76.60 56.04 39.18 30.56 9.73 1.38 5.26 0.27 1.40 0.19 0.26 0.65 1.02 0.52 0.64 0.49 0.67 0.90 0.64 0.67 0.45 0.63 0.20 1.35 1.59 1.30 1.94 1.03 1.29 0.35 female L 82.14 94.84 44.81 76.34 88.81 37.46 42.00 40.10 25.80 31.97 58.69 53.17 23.68 47.17 34.15 27.98 26.99 20.40 19.28 74.08 73.12 53.47 37.16 27.71 8.87 1.04 1.30 1.06 2.22 3.19 1.50 1.46 2.81 1.05 0.76 0.89 2.03 0.94 1.62 1.51 1.08 0.90 0.91 0.84 2.79 2.66 1.72 1.30 1.54 0.63 T 78.60 91.53 42.38 73.50 85.63 35.80 42.40 41.00 25.53 31.32 56.95 51.72 22.27 45.62 33.92 27.72 28.33 20.00 17.78 70.85 70.95 52.40 38.48 29.50 8.78 1.98 2.72 1.24 1.78 2.09 0.98 1.55 1.17 1.02 1.03 1.98 1.50 1.03 1.16 1.70 0.60 0.50 0.41 0.63 1.13 1.03 0.74 2.15 1.69 0.83 E 84.36 94.82 45.18 78.35 88.71 38.46 45.44 43.16 27.12 31.16 60.49 54.46 24.10 46.49 33.47 29.78 28.99 19.74 18.23 76.63 76.76 54.59 37.97 32.69 9.69 1.82 1.76 1.03 2.08 2.00 1.33 1.39 1.19 0.82 1.04 1.42 1.43 1.31 1.34 1.42 1.09 1.11 0.50 1.34 1.79 1.79 1.47 1.76 1.84 0.61 M 80.92 91.53 42.82 75.18 85.74 36.29 42.25 41.45 25.37 31.81 57.46 51.86 23.36 43.81 31.07 27.31 29.00 20.06 17.47 72.13 72.37 51.80 34.65 28.53 8.84 1.84 1.93 1.79 1.60 1.67 1.70 0.59 1.09 0.46 0.69 2.18 2.31 1.01 1.21 1.31 0.92 0.68 0.60 0.96 1.03 0.88 1.17 1.79 1.84 0.66 S 84.35 95.96 45.35 78.31 89.75 38.46 43.06 42.37 26.76 31.66 59.68 53.83 24.66 46.45 33.18 28.83 29.43 20.31 18.65 76.49 76.62 54.80 37.07 30.18 9.24 3.01 3.60 2.08 2.81 3.14 1.89 1.92 1.32 1.22 1.10 2.52 2.22 1.54 1.56 1.33 1.27 1.01 0.85 1.13 2.66 2.81 1.82 1.61 1.45 0.55 J 84.14 95.89 43.67 77.26 88.61 36.15 43.49 43.13 26.43 31.40 60.06 53.52 24.04 47.93 33.57 29.43 29.66 20.30 17.82 75.03 75.59 55.00 38.46 32.45 9.31 1.54 2.45 1.12 1.67 2.27 1.18 1.28 1.36 1.41 0.51 1.87 1.86 0.68 1.19 1.18 0.60 1.31 0.78 1.07 1.98 1.77 1.53 1.07 2.09 0.88 B 81.55 92.39 42.91 74.48 85.78 36.33 41.57 40.45 25.94 31.23 57.08 51.35 23.26 44.50 32.44 27.42 28.89 20.29 18.89 72.04 72.26 52.38 36.01 28.71 8.75 3.45 2.40 1.68 2.09 2.49 1.44 2.22 1.88 1.31 1.05 1.49 1.44 1.42 1.52 1.75 1.10 0.75 0.79 1.04 2.35 2.69 1.71 1.69 1.49 0.62 A 83.67 93.33 44.33 78.02 87.50 38.73 43.05 41.27 26.08 29.62 58.87 53.03 24.95 45.72 34.18 28.98 28.15 18.62 17.98 76.07 77.32 55.60 38.18 31.17 9.42 2.58 2.83 1.73 2.57 2.70 0.76 1.40 1.23 1.01 0.23 1.67 1.78 1.05 0.65 1.11 0.83 0.95 0.30 0.61 1.77 1.46 1.15 1.66 1.23 0.33 The mean values are shown in the upper rows, and the standard deviations in the lower rows. Only the mean values are arranged in males of the T population including one specimen.

Table 4. Significant differences of measuremants among various population

Populations PL1 BL1 SL1 PL2 BL2 SL2 OZW AZW LBO GNB MPL1MPL2 LBS DL LMR GPB GMB GBO HOB LC LA LHR LMRMOHR HM male LXE 0.074 0.349 0.399 0.051 0.355 0.110 0.002 0.023 0.166 0.010 0.272 0.066 0.035 0.297 0.086 0.004 0.000 0.001 0.045 0.088 0.021 0.016 0.008 0.000 0.003 LXM 0.146 0.007 0.112 0.373 0.061 0.254 0.028 0.024 0.469 0.299 0.145 0.348 0.283 0.024 0.016 0.083 0.000 0.031 0.046 0.461 0.054 0.162 0.117 0.000 0.332 LXS 0.075 0.241 0.114 0.047 0.077 0.012 0.041 0.011 0.003 0.310 0.240 0.073 0.017 0.466 0.008 0.048 0.000 0.113 0.003 0.033 0.002 0.017 0.340 0.000 0.003 LXJ 0.448 0.357 0.092 0.405 0.360 0.013 0.143 0.030 0.491 0.053 0.465 0.480 0.438 0.416 0.015 0.065 0.001 0.303 0.006 0.482 0.142 0.163 0.322 0.000 0.015 LXB 0.362 0.396 0.456 0.397 0.377 0.315 0.320 0.310 0.090 0.095 0.426 0.376 0.380 0.080 0.000 0.260 0.000 0.377 0.014 0.463 0.173 0.236 0.053 0.052 0.478 EXM 0.006 0.018 0.061 0.005 0.013 0.347 0.081 0.482 0.209 0.035 0.046 0.104 0.024 0.000 0.108 0.045 0.014 0.249 0.311 0.066 0.238 0.040 0.002 0.005 0.021 EXS 0.323 0.126 0.130 0.417 0.162 0.164 0.006 0.261 0.113 0.000 0.381 0.450 0.402 0.311 0.207 0.099 0.059 0.013 0.310 0.332 0.162 0.444 0.013 0.001 0.143 EXJ 0.143 0.495 0.049 0.033 0.216 0.000 0.033 0.447 0.195 0.180 0.336 0.079 0.053 0.391 0.190 0.167 0.379 0.013 0.018 0.086 0.159 0.141 0.068 0.020 0.295 EXB 0.285 0.487 0.398 0.141 0.275 0.312 0.004 0.063 0.486 0.105 0.442 0.215 0.042 0.048 0.006 0.026 0.426 0.015 0.405 0.142 0.239 0.201 0.000 0.000 0.001 MXS 0.019 0.003 0.013 0.046 0.004 0.110 0.231 0.292 0.013 0.400 0.080 0.172 0.012 0.068 0.172 0.401 0.084 0.107 0.137 0.065 0.082 0.137 0.065 0.491 0.033 MXJ 0.183 0.028 0.397 0.483 0.099 0.005 0.220 0.462 0.479 0.117 0.179 0.341 0.246 0.014 0.234 0.334 0.092 0.108 0.138 0.477 0.315 0.433 0.108 0.242 0.062 MXB 0.187 0.115 0.295 0.321 0.178 0.446 0.092 0.082 0.138 0.229 0.240 0.494 0.194 0.473 0.209 0.297 0.038 0.081 0.243 0.497 0.407 0.497 0.349 0.109 0.327 SXJ 0.087 0.137 0.006 0.045 0.050 0.000 0.430 0.330 0.009 0.026 0.259 0.095 0.039 0.397 0.486 0.434 0.183 0.324 0.001 0.054 0.039 0.161 0.432 0.239 0.378 SXB 0.085 0.092 0.057 0.034 0.014 0.022 0.079 0.027 0.025 0.035 0.285 0.088 0.004 0.024 0.001 0.108 0.053 0.175 0.424 0.014 0.009 0.067 0.002 0.018 0.000 JXB 0.398 0.491 0.235 0.331 0.494 0.017 0.259 0.087 0.120 0.402 0.447 0.376 0.454 0.082 0.035 0.184 0.437 0.431 0.018 0.482 0.480 0.462 0.040 0.034 0.006 female LXE 0.005 0.490 0.237 0.014 0.462 0.046 0.000 0.001 0.001 0.019 0.004 0.039 0.187 0.132 0.129 0.000 0.000 0.016 0.015 0.006 0.000 0.046 0.104 0.000 0.002 LXM 0.087 0.003 0.018 0.155 0.031 0.089 0.362 0.161 0.197 0.348 0.101 0.131 0.268 0.000 0.001 0.122 0.000 0.227 0.001 0.078 0.278 0.034 0.003 0.178 0.464 LXS 0.032 0.212 0.253 0.015 0.182 0.049 0.042 0.000 0.009 0.187 0.157 0.181 0.022 0.085 0.018 0.020 0.000 0.371 0.042 0.004 0.000 0.014 0.426 0.000 0.028 LXJ 0.006 0.165 0.033 0.166 0.440 0.026 0.015 0.006 0.136 0.041 0.050 0.352 0.186 0.136 0.187 0.001 0.000 0.397 0.002 0.209 0.017 0.028 0.016 0.000 0.103 LXB 0.336 0.013 0.009 0.017 0.005 0.029 0.284 0.354 0.386 0.027 0.010 0.006 0.190 0.000 0.007 0.099 0.000 0.362 0.157 0.025 0.208 0.056 0.031 0.049 0.306 EXM 0.002 0.002 0.002 0.004 0.005 0.006 0.000 0.007 0.000 0.109 0.002 0.005 0.137 0.001 0.002 0.000 0.494 0.134 0.134 0.000 0.000 0.001 0.001 0.000 0.010 EXS 0.495 0.159 0.395 0.484 0.132 0.496 0.000 0.030 0.160 0.074 0.157 0.169 0.123 0.470 0.258 0.009 0.093 0.013 0.131 0.432 0.434 0.353 0.045 0.000 0.007 EXJ 0.393 0.141 0.004 0.114 0.458 0.000 0.002 0.479 0.082 0.274 0.288 0.103 0.451 0.011 0.433 0.205 0.112 0.030 0.237 0.036 0.079 0.273 0.247 0.392 0.131 EXB 0.013 0.007 0.001 0.000 0.001 0.000 0.000 0.000 0.005 0.425 0.000 0.000 0.058 0.001 0.052 0.000 0.391 0.021 0.075 0.000 0.000 0.001 0.003 0.000 0.000 MXS 0.009 0.005 0.007 0.010 0.004 0.009 0.179 0.070 0.008 0.386 0.034 0.034 0.037 0.000 0.001 0.006 0.179 0.263 0.015 0.000 0.001 0.000 0.002 0.012 0.072 MXJ 0.003 0.003 0.155 0.024 0.017 0.432 0.034 0.020 0.069 0.120 0.021 0.090 0.087 0.000 0.002 0.000 0.161 0.285 0.282 0.006 0.002 0.001 0.000 0.003 0.164 MXB 0.355 0.248 0.462 0.252 0.486 0.478 0.258 0.139 0.180 0.135 0.345 0.279 0.442 0.184 0.064 0.421 0.391 0.284 0.007 0.468 0.465 0.247 0.071 0.412 0.389 SXJ 0.426 0.479 0.016 0.157 0.165 0.001 0.270 0.073 0.245 0.256 0.344 0.357 0.137 0.007 0.225 0.101 0.286 0.487 0.030 0.074 0.161 0.385 0.012 0.000 0.373 SXB 0.006 0.002 0.000 0.000 0.000 0.000 0.008 0.000 0.016 0.097 0.001 0.000 0.002 0.000 0.047 0.000 0.032 0.462 0.237 0.000 0.000 0.000 0.018 0.001 0.003 JXB 0.032 0.003 0.141 0.002 0.007 0.381 0.018 0.001 0.209 0.340 0.001 0.003 0.080 0.000 0.059 0.000 0.042 0.485 0.015 0.003 0.002 0.001 0.001 0.000 0.044 Each value indicates the limit percentage in which the significant differences are confirmed. 1234 H. ENDO ET AL.

Kra at least to Hat-Yai District including Surathani. Thus, we think that the Surathani population is more closely related to that of the East Thailand population than that of the Peninsular Malaysia population in the phylogenetic status. In fact, all specimens of T population used in this study have been identified as a single subspecies Tragulus javanicus affinis based on the color variation at the subspecies level from the biological data of the specimens. The specimens from Thailand are rare, so we will need the further examination of the geographical variations in some districts of Thailand. The population of Peninsular Malaysia is obviously originated from the genetic groups in the south part of the Isthmus of Kra. We suggest that the similarities repre- sented in the dendrograms may indicate the phylogeneti- cally close relationships or the presence of common ancestors within the cluster of M-S-B populations. The L and E populations were also separated from the other ones by evaluating the size factor from the results of the canonical discriminant analysis. Although E and A population were connected in the dendrogram, we could not Fig. 2. Dendrograms of morphological similarities among 8 point out that the A plots were closely-related to E ones in populations based on the skull shape calculated using Q- the scattergram. In the J population, which was obviously mode correlation coefficients. Clustering was carried out by different between the sexes in the dendrogram, the distribu- UPGMA method. A) Male. B) Female. The locality symbols are explained in Table 1. The distance matrices are shown in tion tendencies remain unclear in comparison with the plots Table 5. of M, S and B populations in the scattergram. We believe that the size factor has an influence on the status of the A, E and J populations in the scattergram unlike in the clustering The dendrograms showed the clusters of E-A, L-T and analysis. M-B-S (Fig. 2). We suggest that the similar island-isolation effect may be established in some islands of Tenasserim ACKNOWLEDGEMENTS. We thank Dr. James Mead District and Langkawi Island. We also suggest that the phy- and Ms. Gordon Linda and staff of Smithsonian Institution, logenetical relationships are demonstrated in the similarities Dr. Darrin Lunde and Ms. Jean Spence of American of the two mainland populations of L and T. The locality of Museum of Natural History (New York, U.S.A.) and Dr. the T population included both East Thailand District and Chris Smeenk of Nationaal Natuurhistorische Museum Surathani Area in Peninsular Thailand. However as sug- (Leiden, The Netherlands), for their kind management in gested in zoogeography of Tupaia [9, 10], the population of our examinations of Museum specimens. This study was the northern variation immigrated through the Isthmus of financially supported by Grant-in-Aids for Scientific

Table 5. Lower half distance matrix transformed from Q-mode correlation coefficients maleELMBTAJS E0.000 L 1.666 0.000 M 1.154 1.447 0.000 B 1.584 0.868 0.551 0.000 T 1.321 0.563 1.328 1.156 0.000 A 0.287 1.380 1.437 1.733 1.235 0.000 J 0.775 1.433 0.731 1.023 1.569 1.056 0.000 S 1.429 1.107 0.509 0.257 1.257 1.569 0.935 0.000 femaleELMBTAJS E0.000 L 1.685 0.000 M 1.194 1.184 0.000 B 1.680 0.739 0.461 0.000 T 1.329 0.788 1.234 0.822 0.000 A 0.501 1.217 1.637 1.643 1.404 0.000 J 0.572 1.589 1.225 1.475 0.846 0.984 0.000 S 1.094 1.095 0.541 0.870 1.783 0.969 1.211 0.00 SKULL VARIATION OF LESSER MOUSE DEER 1235

based on metric characters. Amer. J. Phys. Anthropol. 38: 743– 753. 5. Driesch, A. 1976. A Guide to the Measurement of Animal Bones from Archaeological Sites. Harvard Univ., Cambridge. 6. Duerst J. U. 1926. Handbuch der Biologischen Arbeitsmeth- oden, Abt. 7. Urban & Schwarzenberg, Berlin. 7. Endo, H., Maeda, S., Yamagiwa, D., Kurohmaru, M., Hayashi, Y., Hattori, S., Kurosawa, Y. and Tanaka, K. 1998. Geographi- cal variation of mandible size and shape in the Ryukyu wild pig (Sus scrofa riukiuanus). J. Vet. Med. Sci. 60: 57–61. 8. Endo, H., Kurohmaru, M., Hayashi, Y., Ohsako, Y., Matsu- moto, M., Nishinakagawa, H., Yamamoto, M., Kurosawa, Y. and Tanaka, K. 1998. Multivariate analysis of mandible in the Ryukyu wild pig (Sus scrofa riukiuanus). J. Vet. Med. Sci. 60: 731–733. 9. Endo, H., Nishiumi, I., Hayashi, Y., Rashdi, A. B. M., Nadee, N., Nabhitabhata, J., Kawamoto, Y., Kimura, J., Nishida, T. and Yamada, J. 2000. Osteometrical analysis in skull of the common tree shrew from both sides of the Isthmus of Kra. J. Vet. Med. Sci. 62: 375–378. 10. Endo, H., Hayashi, Y., Rerkamnuaychoke, R., Nadee, N., Nab- hitabhata, J., Kawamoto, Y., Hirai, H., Kimura, J., Nishida, T. and Yamada, J. 2000. Sympatric distribution of the two morphological types of the common tree shrew in Hat-Yai Districts (South Thailand). J. Vet. Med. Sci. 62: 759–761. 11. Foster, J. B. 1964. Evolution of mammals on island. Nature (Lond.) 202: 234–235. 12. Kloss, C. B. 1916. On a new mouse deer from upper Siam. J. Nat. Hist. Soc. Siam 2: 88–89. 13. Kloss, C. B. 1918. Notes on Malayan and other mouse-deer. J. Fed. Malay St. Mus. 7: 245–256. 14. Lekagul, B. and McNeely, J. A. 1977. Mammals of Thailand, 2nd ed., Saha Karn Bhaet, Co., Bangkok. 15. Miller, G. S. 1900. A new mouse deer from lower Siam. Proc. Biol. Soc. Wash. 13: 185–186. 16. Miller, G. S. 1902. Two new Malayan mouse deer. Proc. Biol. Fig. 3. The visualized results of the canonical discrimi- Soc. Wash. 15: 173–175. nant analysis among 7 populations. Scattergrams show 17. Miller, G. S. 1903. Description of eleven new Malayan mouse the individual scores on the discriminant axes 1 (horizon- deer. Proc. Biol. Soc. Wash. 16: 31–44. tal) and 2 (vertical). The locality symbols are explained 18. Rohlf, F. J. and Sokal, R. R. 1965. Coefficients of correlation in Table 1. A) Male. The first function comprises 37.2% and distance in numerical taxonomy. Univ. Kansas Sci. Bull. of the variance and the second function 30.7%. Eigenval- 45: 109–126. ues for the first and second functions are 2.60 and 2.15, 19. Sneath, P. H. A. and Sokal, R. R. 1973. Numerical Taxonomy: respectively. B) Female. The first function comprises The Principles and Practice of Numerical classification. Free- 45.0% of the variance and the second function 27.7%. man, San Francisco. Eigenvalues for the first and second functions are 3.89 20. Terai, S., Endo, H., Rerkamnuaychoke, W., Hondo, E., S. and 2.39, respectively. Agungpriyono, S., Kitamura, N., Kurohmaru, M., Kimura, J., Hayashi, Y., Nishida, T. and Yamada, J. 1998. An osteometri- Research nos. 13640705, 13575027, 14405030 and cal study of the cranium and mandible of the lesser mouse deer (Chevrotain), Tragulus javanicus. J. Vet. Med. Sci. 60: 1097– 15570092 from the Ministry of Education, Science and Cul- 1105. ture, Japan. 21. Thomas, O. and Wroughton, R. C. 1909. On mammals from Rhio Archipelago and Malay peninsula. J. Fed. Malay St. Mus. REFERENCES 4: 99–129. 22. Van Dort, M. 1988. Note on the skull size in the two sympatric 1. Chasen, F. N. 1935. Three new Malaysian mammals. Bull. Raf- mouse deer species, Tragulus javanicus (Osbeck, 1765) and Tra- fles Mus. 9: 98–100. gulus napu (F. Cuvier, 1822). Z. Säugetierkunde 53: 124–125. 2. Chasen, F. N. 1940. A handlist of Malayan mammals. Bull. 23. Van Dort, M. 1989. Skin, skull and skeleton characters of Raffles Mus. 15: 1–209. mouse deer (Mammalia, Tragulidae), with keys to the species. 3. Corbet, G. B. and Hill, J. E. 1992. The Mammals of the Bull. Zoöl. Mus. Univ. Amsterdam 12: 89–96. Indomalayan Region: A Systematic Review. Oxford Univ. 24. Wilson, D. E. and Reeder, D. A. 1993. Mammal Species of the Press, Oxford. World. A Taxonomic and Geographic Reference. 2nd ed., 4. Corruccini R. S. 1973. Size and shape in similarity coefficients Smithsonian Institution Press, Washington, DC.