ABSTRACT

Department of Fisheries and Oceans, Bedford Institute of Oceanography, Ocean Science Division, Box 1006, Dartmouth, Nova Scotia, B2Y 4A2, Canada

Department of Biology, Queen’s University at Kingston, 74 University Avenue, Kingston, Ontario, K7L 3N6, Canada

E-mail: choij@mar.dfo-mpo.gc.ca, frankk@mar.dfo-mpo.gc.ca, petrieb@mar.dfo-mpo.gc.ca, wleggett@post.queensu.ca

Abstract

This review examines a large marine continental shelf ecosystem (the Eastern Scotian Shelf of Canada (ESS)), that has undergone dramatic hysteresis-like changes in the recent past, using a pragmatic approach that combines empirical, reductionist and holistic methods based on the integrated analysis of 55 primary and secondary biotic, abiotic, and human variables over a 43-year period. The integrated analysis reveals that the ESS ecosystem has changed states, i.e., a ‘regime shift’ from a community dominated by large-bodied demersal fish to one dominated by small demersal and pelagic fish species and benthic macroinvertebrates. A dynamic interplay between ocean physics, biology and exploitation presents a more realistic casual scenario than any single component hypothesis such as fishing pressure alone. The forces contributing to the stability of the alternate state include both top-down processes involving strong trophic interactions initiated at the apex predator level and bottom-up processes involving energy flow and nutrient cycling that have fundamentally altered the matter and energy flow patterns in the ESS ecosystem. It is suggested that the ESS has been literally ‘devolving’ when viewed from the perspective of the theory of ecological succession. Further, higher-order variables are identified as early warning indicators, sensitive to the underlying structural and functional changes that occurred on the ESS ecosystem. They have been determined for an adjacent system where systemic changes have not yet been observed and predict

a potential collapse within a decade. Integrated assessment of ecosystems is a great challenge and their management requires

com-

prehension

of ecological systems. Description alone is not sufficient to allow comprehension, especially if ‘information overload’ (the disconnect between system

description

and system

com-

prehension

) is to be avoided and meaningful insights and strategies are to emerge. Reductionistic analysis involves the dissection and identification of key processes or feedback mechanisms likely to be operative in an ecological system. However, value cannot be ascertained from variations in processes, as a directionality of time does not exist in such a perspective. In fact, the approach has accelerated the information overload experienced by all stakeholders. Holistic approaches are being increasingly used to aid in the valuation of ecological systems as the directionality of time is made explicit in this perspective. It is suggested that integrated assessment requires not only the integration of descriptive information, but also the integration of our perception of ecological systems as being both a whole and a part.