誘蛾燈で誘殺されたガの種類とその数

概要

論文の詳細を見る
Fitting trial of Fisher's logarithmic series (Fisher 1943) and the truncated discrete lognormal proposed by Preston (1948) was made to the light trap record of Lepidoptera taken at the Kagawa Agricultural Experiment Station for the years 1960 to 1966. Table 1 is the original listing of captured moths representing the relation between the number of species and that of individuals. The largest number of individuals captured is Nymphula turbata Butler excepting for the year 1965 in which Plutella maculipennis Curtis was the largest. Chilo suppressalis Walker, the most important insect pest of rice plant in Japan, was the second or third largest number of individuals. But the number of Chilo suppressalis was only about 10 per cent of whole captured moths. The total number of individuals of these three species reached to about 85% of whole captured moths. Compared with the data reported by Williams and other workers from England (Williams 1939), the number of moth species of our record was only about a half or one-third of them. However, the number of individuals was about twenty times as many as the records of Williams and other workers. The big differences between the numbers of species and that between the numbers of individuals were assumed to be much owing to the difference of florae of lepidopterous habitat. That is, the area in which our light trap was exposed is almost paddy-field and the flora as the habitat of Lepidoptera was much simpler than the flora in which Williams and other workers made their trap records. Intermediate steps for applying the logarithmic distribution to the moth captures for each year in Table 1 and to their combined record, the statistics R^^^ and α^^^, and their variances are shown in Table 2. The figures shown in Table 3 are the equivalent terms for computing the statistics of the truncated discrete lognormal distribution for the same data. As shown in Table 4, the captures in each series and the seven-year total highly agreed with their expectations for the truncated discrete lognormal distribution. The three capture records for the year 1960, 61 and 63, however, not agreed with their expectations for the logarithmic distribution and the seven year combined record was also disagreeable with its expectation for the logarithmic series. It is considered that the lognormal distribution provided a much better description of the observed captures than the logarithmic series. The relations between the cumulative frequencies captured in probits and the upper class limits of the grouping intervals in logarithms are shown in Fig. 1. As is seen in Fig. 1 and from the result of X^2-test it can be said that of the two parameters characterized the lognormal distribution, the estimated mean y was much more consistent than the standard deviation σ^^^ over the seven years. The total number of species in the habitat sampled by the trap was estimated yearly statistics and the observed catch from N^^^=61 species in 1962 to N^^^=83 species in 1961. On this lognormal model, 4.71 per cent of the species of Lepidoptera escaped from capture in the year 1965, and the maximum per cent of the trap escaped capture was 12.96 per cent in 1964. Over the seven year period the species missed dropped to 7.43 per cent.
日本昆虫学会の論文
1968-06-29

誘蛾燈で誘殺されたガの種類とその数

スポンサーリンク

概要

著者

関連論文

スポンサーリンク