Fisheries management based on ecosystem dynamics and feedback control HIROYUKI MATSUDA* AND TOSHIO KATSUKAWA Ocean Research Institute, University of Tokyo, 1-15-1, Minamidai, Nakano-ku, Tokyo, 164–8639, Tokyo 164–8639, Japan ABSTRACT Recently, ecosystem management has become popular for forestry, agriculture and fisheries management. Carrying capacity and maximum sustainable yield for a particular species definitely depend on population sizes of other species in the same ecosystem. Natural stock fluctuations of sardine, anchovy and chub mackerel are well known examples of large, natural fluctuations. There is a negative correlation among their fluctua- tions. In accordance with the cyclic advantage hypo- thesis for replacement of pelagic fish species (Matsuda et al., 1992), we can predict the next dominant spe- cies, despite an uncertainty in the year of the next replacement. We recommend that commercial fisher- ies should switch their target to the next dominant species before the stock of the present dominant spe- cies collapses. Whilst total allowable catch (TAC) of the present dominant species can be as large as we can consume, TAC after the species collapses should be much smaller than the present catch level. Key words: ecosystem management, species replace- ment, target switching WHAT IS ECOSYSTEM MANAGEMENT? Recently, ecosystem management is becoming popular for forestry, agriculture and fisheries management (Christensen et al., 1996). Fish stock dynamics are related to adjacent stocks of the same species. Adja- cent regional stocks make a single Ômeta-populationÕ (Hanski, 1991). Fish stock dynamics are not described as deterministic. The future trend of a stock is unpredictable. We have to incorporate uncertainty and unpredictability in management into the decision- making process. Because of uncertainty, risk manage- ment is indispensable in fishery management. Risk is usually assessed under assumptions that are not yet verified (Rodricks, 1992). Carrying capacity and maximum sustainable yield for a particular species depend on the population sizes of other species in the same ecosystem. In addition, ecosystems, especially marine ecosystems, are charac- terized by (i) temporal fluctuations, (ii) uncertainty in assessment and ecosystem processes, and (iii) an open system concerned with straddling stocks and interna- tional fisheries management. Adaptive management is one of the key concepts for ecosystem management (Holling, 1978; Walters, 1986). Adaptive management incorporates uncer- tainties in stock assessment, ecosystem processes and human impacts into management procedures. Adap- tive management is characterized by adaptability and accountability. Adaptability means a change of fishing pressure in response to stock fluctuation (feedback control). Accountability means incorporating new knowledge into management policy as soon as it is found (adaptive learning). We often recognize man- agement as an experiment. How to change manage- ment policy with successive monitoring data should be considered before the management starts. We call this Ôsystematic feedback controlÕ. In addition, manage- ment policy can be designed to investigate a hypo- thesis for an ecosystem (active adaptive learning, Walters, 1986). Successive monitoring is indispensable for adaptive management. Tanaka (1980) proposed Ôfeedback managementÕ for sustainable fisheries with uncertainty in stock abundance using differential equation models. This concept is close to adaptive management and has been applied to the Revised Management Procedure (RMP) by the International Whaling Commission (IWC). It is easy to find many historical examples of fish- eries mismanagement (Hannesson, 1996). The best approach to fisheries management is considered to be to manage the fishing ground. To protect a part of fish habitat is robust against uncertainty in ecosystem processes. Recently, Katsukawa and Matsuda (2002) *Correspondence. e-mail: matsuda@ori.u-tokyo.ac.jp Received 30 June 2001 Revised version accepted 22 July 2002 FISHERIES OCEANOGRAPHY Fish. Oceanogr. 11:6, 366–370, 2002 366 Ó 2002 Blackwell Science Ltd.