Though many taxonomic studies on the japanese
Lepidotrigla had been made by several authors, most of them were not sufficient in the treatment of variation of the specific characteristics. Therefore, the resultant uncertainties in the identification embarrassed the population studies of the fishes belonging to the genus.
Accordingly, in order to make clear the specific characters, I engaged in an observation of a fresh body and in the measurement of several parts of body in the present paper, which was based on the large numberrt of specimens collected from the commercial landings of the Eastern Sea trawlers from 1948 to 1950.
In the pattern of the innerside of pectorals, six types are distinctly recognizable as shown in Plate I, though they are somewhat divergent. In order to make clear the specification of each types, from the viewpoint of morphometry, I chose the measured value of the individuals falling in the interval of 15±1cm in their standard length. (Thus, I unified the size approxi. mately, and avoided the error caused by a difference in their size.) As the frequency distribution of these values could be regarded as nearly normal type (see Fig. 1), I computed the rejection limit for each item under 5 per cent level of signlificance, to indicate the thresholds of their variation. If these limit separated from each other perfectly, I regarded them as distinct species morphometrically. According to the above mentioned standpoint, the types A, B, C, E and first group of F are separated clearly, as illustrated in Fig. 2. (Here, type F was divided into two different groups by the measured value, for instance, that of rostrum as I showed in this figure.) But the types B, C and second group of F overlapped each other in variation, and I cannot assert the distinctness of these groups in the natures dealt with in Fig. 2. While, in Fig. 3, the separation is perfect between the second group of F and the types B, C. Therefore, it may be concluded that the groups of type A, B, C, E, F (I): and F (II) respectively are distinct ones, not only from their pectorals pattern, but also from the morpho. metrical data. Type D is excluded from the present treatment, because it is quite a rare species, and is not important economically.
Next, I examined the relation between the above mentioned groups and the species hitherto described. And obtained the following result: namely, type A group correspond to
L. japonica, type B group to
L. guentheri, C to
L. kishinouyei, D to
L. punctopectoralis, E to
L. microptera, F (I) to
L. alata, and F (II) to
L. abyssalis. When we identify the samples, though the method is somewhat troublesome, it is possible to do it by the measured values of several parts of body using Fig. 2 and 3, which were made for every section of the size of the fish, as a chart for the key of species. On the other hand, the patterns of the innerside of pectorals are available as a guide for species, and these patterns can be recognize in the individuals of smaller size, even in 50mm length. According to the latter method, it is rather difficult to divide the two groups of type F. However, one group of type F,
L. alata, can be easily and unrnistakablly distingushed from the other group of
L. abyssalis, by its large triangular rostral process. The latter method of identification can be made by inspection irrespective of the size of fish.
抄録全体を表示