Effect of Harvesting Alternatives on the Quality of U.S. North Atlantic Bluefin Tuna

Abstract

The impact of harvest practice choice on four quality attributes (i.e., freshness, fat content, color, and shape) and weight of U.S. captures of bluefin tuna Thunnus thynnus in the North Atlantic is studied. Using ordered-probit and two-limit tobit estimation methods, grades for the four quality attributes of individual fish were regressed against harvest time, area, and gear. Weight of fish was also regressed against the same exogenous variables, this time using ordinary least-squares estimation. The results indicate that the four quality attributes and weight are, in fact, influenced by harvest practices. In particular, the analyses highlight the importance of harvest timing on the grade of all attributes considered. Implications of the results for public management of the U.S. North Atlantic bluefin tuna fishery are discussed.

Notes

¹ A different interpretation of the attribute grades was also explored. A value between 1 and 5 was given to each attribute grade based on experience accumulated in the Tsukiji Market in Tokyo. Equation (1) above was reestimated assuming that the attribute grades are continuous variables. However, the attribute grades, A_i , are left- and right-censored variables (i.e., 1 ≤ A ≤ 5). Therefore, the use of OLS to compute equation (1) would entail the assumption that it is possible to observe attributes graded below and above those limits. Under these conditions, it is statistically more appropriate to use two-limit censored regression methods. The regression parameters for the four attributes were derived by maximum likelihood estimation using a two-limit Tobit model (Maddala 1983). The results can be found in Table A1.1 in the Appendix 1. The two-limit Tobit model leads to the same conclusions as the ordered-probit approach.