This paper proposes a method for estimating stock abundance for a fishery e
xploiting two stocks, using a model of stock dynamics incorporating effort
allocation. We assume that the fishery selects a target stock every hour of
operation from the two stocks with a certain probability (rx), and that th
e spatial distribution of stock is either regular or aggregated. The parame
ters are estimated by the maximum likelihood procedure. Our model is applie
d to the data of the diving fishery from 1972 to 1976 which catches abalone
and top shell along the coast of Toji in Shizuoka Prefecture. The minimum
Akaike Information Criterion (AIC) supported the model that the ex values w
ere changed between- and within-year. The coefficients of variation of the
abundance estimate for abalone ranged from 0.01 to 1.37, and for top shell
from 0.01 to 0.03. The abundance estimates were compared with those by the
DeLury's method. The modifications of the model are discussed.