Zoogeographical barriers causing discontinuous osteometrical variations in the northern treeshrew skulls

Abstract

Morphological variation of the skull was examined in the northern treeshrew (Tupaia belangeri) from various localities across Southeast Asia. Through a multivariate analysis, the treeshrews from South Vietnam exhibited distinct morphological characteristics compared to other populations from Thailand and Laos, and Malaysia. The plots of the specimens of North Vietnam are not randomly mixed with Thailand plots segregation in the scatteregrams of canonical discriminant analysis. Since the skulls of the population from North Vietnam were morphologically similar to those form central Laos and northern and northeastern Thailand, the zoogeographical barrier effect of Mekong River was not clearly confirmed. The population of the Kanchanaburi in western Thailand is clearly smaller in size compared to the other populations. The southern border of the distribution of this species is determined by the Isthmus of Kra or Kangar-Pattani Line. In the northern treeshrew, which is distributed from southern China to Bangladesh and southern Thailand, we have detected osteometrical geographical variation driven by geography. These results indicate that the skull morphology in the Tupaia glis-belangeri complex distinctively differs in South Vietnam, western Thailand, and southern Thailand. The zoogeographical barrier and factor separating these districts are expected to clarify in the future.

Keywords: Isthmus of Kra; Mekong River; geographical variation; northern treeshrew; skull.

Fig. 1.

Distributions (grey areas) of the northern treeshrew populations in Indochinese and Malayan regions. I, Isthmus of Kra, K, Kangar-Pattani Line, M (arrow), present Mekong River. Locality symbols are explained in Table 1.

Fig. 2.

Scattergram of the plots in the principal component analysis. The raw measurement values of male are applied to the analysis. Plot symbols indicate the locality of the specimen as shown in Table 1. Horizontal axis, the first principal component. Vertical axis, the second principal component.

Fig. 3.

Scattergram of the plots in the principal component analysis. The raw measurement values of female are applied to the analysis. Plot symbols indicate the locality of the specimen as shown in Table 1. Horizontal axis, the first principal component. Vertical axis, the second principal component.

Fig. 4.

Scattergram of the plots in the canonical discriminant analysis. The raw measurement values of male are applied to the analysis. The individual scores are shown on the discriminant axes 1 (horizontal) and 2 (vertical). Plot symbols indicate the locality of the specimen as shown in Table 1.

Fig. 5.

Scattergram of the plots in the canonical discriminant analysis. The raw measurement values of female are applied to the analysis. The individual scores are shown on the discriminant axes 1 (horizontal) and 2 (vertical). Plot symbols indicate the locality of the specimen as shown in Table 1.

Fig. 6.

Scattergram of the plots in the canonical discriminant analysis. The proportion indices of male are applied to the analysis. The individual scores are shown on the discriminant axes 1 (horizontal) and 2 (vertical). Plot symbols indicate the locality of the specimen as shown in Table 1.

Fig. 7.

Scattergram of the plots in the canonical discriminant analysis. The proportion indices of female are applied to the analysis. The individual scores are shown on the discriminant axes 1 (horizontal) and 2 (vertical). Plot symbols indicate the locality of the specimen as shown in Table 1.