Fishery Science: How ’Scientific’ is it?

Menakhem Ben-Yami, in the keynote address delivered at the St. John’s Conference of Harvesters in the North Atlantic fishery, raised some vital issues about fishery science. These are excerpted below. This discussion assumes special relevance in the context of the collapse and depletion facing some of the world’s most productive fishing grounds.

To discuss fishery management debacles, we must ask, among other questions, how stocks are affected by fishing effort and by environment? How scientific is the science used to determine the desirable yields and/or effort, and how reliable are the assessment methods responsible for the answers we’re getting? The question whether the `best available science’ of today provides an adequate basis for rational management is hotly argued among fisheries scientists.

The traditional science is based on mathematical models of exploited fish populations developed decades ago. These models, along with acoustic surveys, are still the main pillars of wisdom on which fishery management is based.

Most recent models used by fishery biologists do not express environmental changes and fluctuations and their real-time effect on the abundance, natural mortality, availability, and vulnerability of fish populations. They use guessed and `guesstimated’ inputs, and mostly ignore climatic and hydrographic variables and inter-relations with other species in the system.

They may be relied on for some stocks existing under ecologically stable conditions and, with qualifications, for long-living species. Even more recent models designed to deal with multi-species fisheries are not the answer, because both, intensive fishing and environmental factors, such as, for example, water temperature, often cause fast changes in the species composition, and, hence, in the whole ecosystem of such fisheries.

At the other side of the argument stands the so-called `holistic’ school of thought. It contends that a science that handles populations in separation from the system’s physical parameters, on the one hand, and the given population’s prey, predators and competitors, on the other, can not be relied on. Unfortunately, while the traditional models are relatively simple and do not involve too many variables and actual data, and can be solved using simple hardware and software, holistic models would be very difficult both to design and operate, and very complex in structure.

Such models would require an assessment of natural mortality and recruitment in real time, for each different environment. This would need plenty of knowledge on the effect of the fishing effort and environmental conditions on the various stocks, and a better understanding of the working inter-relations of the cogs of the ecological clockwork.

New supercomputers and software capabilities signal the evolving possibilities of integration of population models with environmental data and causation correlations. Until, however, enough time, money and research effort have been spent, applicable holistic models are not on the cards, and we are left with the unsolved issue of the inadequacy of conventional methodology.

The problem of the adequacy of the science is topped with another one. There are time lags between several stages: collection of data, analysis, preparation of reports, discussions with decision makers, the process of having the management measures agreed on and implemented.

All this removes the reaction of the management system, often by a couple of years, from the processes occurring in real time in the fishery system. Additional distortions may occur if scientists in charge of stock assessment, due to exerted or perceived pressure, are not telling the truth, only the truth, and nothing but the truth.