名古屋地方における蚊族個体群の発生動態 : A. 1967 年名古屋市内の 1 共同墓地におけるヒトスジシマカ幼虫個体群

抄録

An ecological study was made on the larval population of Aedes albopictus (Skuse) in a cemetery of Nagoya City, central Japan, during a period from June to December, 1967. The present paper reports the larval population dynamics in natural habitat from the view point of seasonal change of population size and age structure. 1) The number of mosquito larvae and pupae sampled from 10 stone basins by a dipper totalled 2,090 during the study period. Most of them (97%) were A. albopictus, and the rest were Culex pipiens pallens, C. tritaeniorhynchus and C. orientalis in decreasing order. 2) Statistical analyses were made of the relationship between mean and variance of A. albopictus larval (pupal) counts in the 10 stations, and it was revealed that distribution of individuals seemed to be contagious and to fit well to the negative binomial distributions with a common k except a few cases (Figs. 3,4 and Table 2). The log(x+1) transformation was adopted in order to normalize the data (x : individual counts per stone basin) according to Bartlett (1947), and the Williams' means (Williams, 1937) were used to represent the seasonal change of population size and age structure of the larval population in the cemetery as a whole. 3) The larval population was observed from late June to late November, and the population size reached its peaks three time (early July, early August and early September), of which the peak in early August being the largest (Fig. 5). 4) Consecutive analysis of larval age structure indicated that adult emergence at population level seemed to occur about three times (late June-mid July, late July-early September, mid Septembermid October) roughly in parallel with the above-mentioned population peaks. Judging from the largest population size and continuous adult productivity in larval population, emerging adults were estimated to be most abundant at the second emergence period (Fig. 5). 5) There were appreciable differences in seasonal fluctuation pattern of the larval population size between the two representations : one was expressed by arithmetic mean of actual individual counts and the other by the Williams' mean (Fig. 5). So, a comparison was made between the arithmetic mean of actual individual numbers and that of log-transformed ones in regard to frequency distribution and variability of the observed counts. As a result, the log-transformed numbers seemed to be more suitable for analyzing dynamics of mosquito population in a habitat area as a whole, because the arithmetic means of the log-transformed numbers were more centrally located and the coefficients of variation of them were more minimized (Figs. 6,7).