The extent of overestimation in the PETERSEN-type estimate of fish population size, caused by the recruitment occuring throughout the experiment, was analysed. Here, the type of recruitment was considered for two cases, i.e., a constant rate of recruitment which means that the ratio of recruitment to the size of fish population at each time is costant, (Table 1 and Fig. 1), and a constant size of recruitment which means that the size of recruitment in each period is constant, but the ratio of recruitment to the population size at each time is generally not constant (Table 2 and Fig. 2).
By using the tagged ratio method, it was easy to remove the bias in the PETERSEN-type estimate, caused by the constant rate of recruitment (Eqns. 16 and 17). But, the bias in the case of the constant size of recruitment was a little difficult to remove, except the case in which the relationship between time and the reciprocal of sample tagged ratio is approximately linear (Eqns. 22, 23 and 24).
The validity of these methods were ascertained by the bead-drawing experiment (Table 3, Fig. 3 and Fig. 4).
Moreover, the bias of the estimate by PETERSEN method when the tag diminution and the recruitment exist simultaneously will obviously be larger than the bias which occurs when they exist independently (Eqns. 28, 29 and 30).
It was possible to remove the bias of estimate, caused by the constant rate of recruitment and tag diminution occurring simultaneously, by the tagged ratio method (Eqns. 31 and 32). However, it was very difficult to remove the bias caused by the simultaneons occurrence of the constant size of recruitment and the tag diminution (Eqn. 34).
抄録全体を表示