Kaszab Zoltán (szerk.): A Magyar Természettudományi Múzeum évkönyve 76. (Budapest 1984)

Demeter, A. ; Lázár, P.: Morphometric analysis of field mice Apodemus: character selection for routine identification (Mammalia)

Spsp — Size of pectoral spot: scored from 1 to 13, as judged by comparison with Fig. 2. The scale of size variation was erected by studying the range of variation in the pooled material, and also utilizing the figures given in FULLAGAR (1967) for A. sylvaticus and LARINA (1958) for A. flavicollis and A. sylvaticus. Colour of the dorsal and ventral pelage was measured along the mid-dorsal and mid-ventral lines with help of a MOMCOLOR tristimulus colorimeter set up according to the CIE 1931 recommendations. CIE C illumination of 6774 K correlated colour temperature was used to measure the tristimulus values, from which the chromacity coordinates (i. e. the colour parameters) and the luminance factor (a quantity which correlates the brightness of the measured object relative to that of the primary standard light, with a basic photometric standard) were calculated (JTJDD & WYSZECKI 1975). Venx — X chromacity coordinate of ventral pelage Veny — y chromacity coordinate of ventral pelage Venv — Luminance factor of ventral pelage Dorx — x chromacity coordinate of dorsal pelage Dory — y chromacity coordinate of dorsal pelage Dorv — Luminance factor of dorsal pelage All specimens were allocated to one of the four age classes shown in Fig. 3. based on the eruption and enamel attrition of the molars, a relative age determination method suggested for Apodemus flavicollis by ADAMCZEWSKA-ANDRZEJEWSKA (1967). Data analysis In addition to the geographic variation mentioned above, sexual dimorphism (ADAMCZEWSKA 1959; AMTMANN & AMTMANN 1965; BOTHSCHAFTER 1965; FELTEN 1952)| and of course agedependent variation are important sources of variation in field mice. These must be accounted for if the main emphasis lies on interspecific differences; this objective may be achieved by standardization. However, since specifically for identification from owl pellets the sex of an individual is unknown, we decided not to compensate for intersexual differences, but to leave this source of bias in the data. To standardize jointly for age-specific and geographical variation (the latter one restricted to that due to altitude), dummy regression was employed (KIM & KOHOTJT 1975a). Each of the 65 variables were regressed upon four dummy variables using the multiple regression routines of subprogram REGRESSION of SPSS (NIE et al. 1975), with animals of age group IV from the lowland localities serving as reference category. Thus, e. g. an age-class II animal from a highland locality had scores of 0, 1,0, 1 for the dummy variables, the first three standing for the age category, the fourth for the altitude category. In all of these analyses the multiple correlation coefficient between the regressed variable and the dummy variables was significant at 0.01 probability level. The multiple regression coefficients were then used to calculate the constants that had to be added to (or substracted from) each variable depending on which combination of age class — locality the specimen belonged to. Fig. 3. Examples of molar attrition showing the four age classes to which the specimens were allocated