Fishing about and about fishing
FAO, TUNA AND FUZZY LOGIC
“Not all that can be measured is important
and not all that is important can be measured” - Albert Einstein
I have on my desk “Management of tuna fishing capacity: conservation and socio-economics” (FAO Fisheries Proceedings 2 – publication-sales@fao.org. 2005, 336 p). The book presents results of the studies carried out under a Japan-funded project under the same name. It deals with (i) the development of tuna fisheries since their inception, including the trends in tuna fishing technology and tuna catches; (ii) the status of the tuna stocks; (iii) the tuna catch data; (iv) purse-seine and longline fishing capacity; (v) non-industrial tuna fisheries; (vi) the tuna market influence on catches; (vii) management of purse-seine and longline tuna fishing capacities, past and future. This book, edited and partly written by 3 experts on tuna resources, W.H.Bayliff of Inter-American Tropical Tuna Commission, and J.I. de Leiva Moreno and J.Majkowski, both of FAO Fisheries, comprises contributions of several more, among them two outstanding specialists, P.M. Miyake and James Joseph.
The way the editors handle the section “Status of the tuna stocks in the world” (p.58-114), drew my particular attention. They don’t produce stock figures and don’t tell the world how many fish of the various tuna species swim in the ocean (some call it, for some reason “standing stock”), and, with only one or two exceptions, they don’t tell how many fish can be safely captured. What they tell the readers is how the various fisheries should behave in the future. Their recommendations are based on assessments how far the size of a given species’ population is from certain optimum reference point. These reference points are estimated on the basis of past catches, various models, and of estimates by regional tuna commissions. Stocks are described as being “above”, “above-near”, “near”, “near-below”, “below”, or “unknown”. Accordingly, fisheries are advised to “reduce” their catches, “not to increase”, “increase, but to an unknown extent”, or “increase, but only after the 0 and 1-class mortality could be reduced”.
Outlook. Thus, for example: Indian Ocean, East and West-Central Pacific skipjack, and South Pacific albacore are the only stocks with potential for some increases in their catches. All stocks of bigeye, yellowfin, Atlantic and Pacific bluefin tunas could possibly also produce increased yield, but only if catches of their younger, smaller individuals could be reduced. Fishing for all others should be reduced.
Their assessments and other chapters make “Management of tuna fishing capacity” an exceptionally important book that’s authoritatively summing up the present knowledge of tuna resources and their fisheries, and most useful to all who’re involved in the capture operations, management and research, or otherwise interested in any local, regional, or global tuna fishing industry.
The manner in which de Leiva Moreno and Majkowski presented their assessment of tuna stocks reminded me of the fuzzy logic approach to some realities of our world. The fact is that, apart for exact sciences and technologies, most of our science, from medicine to ocean ecology and climatology, is dealing with dynamic, unstable, and not fully known systems. As a rule, we’re trying to handle such scientific matters using various statistical methods. In some cases, however, statistical-mathematical models are plainly inadequate to handle complex situations, especially, where the data they’re fed with are insufficient, unreliable, or otherwise lacking. One example: fisheries science and management.
Fuzzy logic. Some 40 years ago, Dr. Lotfi Zadeh of University of California at Berkeley introduced the concept of fuzzy logic as a means to model the uncertainty of natural language. It is a tool, which handles realities that cannot be truly expressed in absolute figures. This concept assumes that the known values represent partial truths, somewhere between "completely true" and "completely false". Formally, it is a branch of mathematics that allows a computer to model the real world by imitating human reasoning, and provides a simple way to reason with vague, ambiguous, imprecise and noisy input or knowledge. Most important, it provides a flexible approach to solving problems.
Introduction of fuzzy logic approach to fisheries science should help fisheries scientists and managers to stop representing their discipline as an exact or a quasi-exact science, and the presently dominant fish population dynamics models as a reliable tool for objective assessment and forecasts. Quantification and "mathematisation" of fishery science is not the only way to describe and explain the ecosystem in which people encounter fish, and both - the environment. A growing number of fishery scientists think that not quantifiable information, ignored in the present models, must be taken into account as explained in qualitative terms, that's in words. Any calculated or assessed fish stock values and catch quotas should be accompanied, or where necessary, replaced by relevant verbal or “fuzzy” descriptions, reservations, additions, corrections, etc.
-------------------------------------------------------------------------------------------------------------------
The following is a layman’s example how fuzzy logic can represent a given stock assessment:
The biomass of a stock is estimated at between 100,000 and 200,000MT.
There's about little chance, say 10%, that it is between 100,000 and 120,000 and between 180,000 and 200,000MT.
There's more chance, say, 20%, that it is between 120,000 and 130,000 and between 170,000 and 180,000MT.
There's more chance (about 30%) that it is between 130,000 and 140,000, and between 160,000 and 170,000MT.
And, finally, the best chance is, say 40%, that the biomass is between 140,000 and 160,000MT.
------------------------------------------------------------------------------------------------------------------
The above is a simplified example, but in many cases such approach would be both reasonable and reliable, and its integration with fishermen’s information and with catch and environment time series, should enable sensible selection of management steps. To my fisherman’s mind it makes more sense than an assessment telling that, for example, the stock size is 143,500MT, and the TAC should be 67,530 MT. With such virtual counting of fish in the sea, one wouldn't know whether to laugh or cry…
Which is why I’m looking for a partner to write a guide on application of some sort of fuzzy logic to fisheries management.