ボンゴネットで同時に得られる2つの採集数の違いについて

概要

論文の詳細を見る
The present paper treats differences between a pair of catches C1 and C2 of eggs and pre-larvae of Englaulis japonicus taken with a bongo net haul. lf the differences are caused by a random process, the catches may follow a binomial distribution. Then the fiducial limits of a parameter D=(C1-C2)/(C1+C2)can be expressed by means of the formula, |D|≤tP/√2N, where 2N=(C1+C2)and tP is the value for p % in the standard normal distribution. Moreover, a normalized catch γi=(C1-C2)/√C1+C2 should follow the standard normal distribution. Through the examinatioa of N-D diagram and the frequency distribution of γ1, it followed that the mean and variance were not independent of each other in a pair of catches taken with a bongo net haul. Moreover, the distribution of the catches are slightly over-dispersed. This over-dispersed distribution may have resulted from some physical accumulative processes, for example small scale horizontal convergence, etc., because eggs and pre-larvae cannot aggregate by themselves, they have more similar differences than expected in a random model.
日本水産學會の論文