Bull. Jpn. Soc. Fish. Oceanogr. 69(1), Page 27-36, 2005   Estimation of the Supply Source of Anchovy (Engraulis japonicus) Larvae Caught in Ise Bay and the Western Enshu-Nada Using a Spawner-Recruit Model Motohiko Nakamura†1 and Kouichi Fujita2 1 Marine Resources Reserch Center, Aichi Fisheries Research Institute, Toyoura, Toyohama, Minamichita-cho, Chita-gun, Aichi 470-3412, Japan2 Mie Prefectural Science and Technology Promotion Center, Fisheries Research Division, Suzuka Fisheries Laboratory, Shiroko, Suzuka, Mie 510-0243, Japan†e-mail: motohiko_nakamura@pref.aichi.lg.jp We estimated the supply source of anchovy (Engraulis japonicus) larvae caught by the coastal fishery in Ise Bay and the western Enshu-Nada using a spawner-recruit model. In this model, the Ricker curve, which was evaluated to be the most appropriate among some functions, was adopted to show relationships between the monthly catch of larvae and the number of eggs caught in Ise Bay and the western Enshu-Nada. The determination coefficients of the model were 0.188-0.285 (p<0.01) from April to September and 0.052-0.178 (p>0.01) from October to December. Fluctuation of the number of eggs caught in the western Enshu-Nada was similar to that in the other parts of the coastal sea off the Tokai districts. Using the model, we estimated the monthly larval recruits from Ise Bay, the coastal sea off the Tokai districts and the other distant areas, respectively. The estimated value of the mean monthly larval recruits from Ise Bay was over 200 tons from June to September, that from the coastal sea off the Tokai districts was over 200 tons from May to August, excluding July，and that from the other distant areas was over 200 tons in May and July. The seasonal changes of the estimated values of the Ricker parameters and the monthly larval recruits were consistent with the seasonal changes of the anchovy spawning grounds and the rainfall, both of which affect survival and transport of larvae. Key words: anchovy larvae, spawner-recruit model, supply source of larvae