Abstract
A regression analysis, which is based on time changes in the ratio of the number of marked fish in the daily catch to the total daily catch, was used to estimate initial population size. This method eliminates overestimation caused by recruitment.