LL1-02 — Ch2 Fisheries: taking stock#

Report: Late lessons from early warnings: the precautionary principle 1896–2000, EEA Environmental Issue Report No 22 (2001) Report pages: 17–30 (chapter text pp. 17–26; Table 2.1 p. 27; references pp. 27–30) PDF pages: 17–30 (printed and PDF page numbers coincide)

Reading record. - Full read. I read the whole text extract, from the first page marker (PDF p. 17) to the last (PDF p. 30). - Visual check. I checked page images of pp. 17, 23, 24, 25, 26 and 27. The extract is clean except for the reference-point subscripts on p. 24 (B_lim, B_pa and so on), which break across lines. - What the chapter contains. Its only non-prose element is Table 2.1 (p. 27). There are no figures, boxes or panels. - Cross-reference only. I also looked at the author’s contributor note (pp. 197–198) and at the passages in the concluding chapter that reuse this case (pp. 171–172, 175–182). The concluding chapter (Ch. 16, “Twelve late lessons”) is credited to the “Editorial team” (p. 168), of which the author is a member.

Quotation note. The report is “All rights reserved” (PDF p. 2). I quote the author’s prose only in short phrases. Quotations of historical and US government sources that appear in the chapter get a little more room. Everything else is compressed paraphrase.

Authors and standpoint#

• Sole author: Malcolm MacGarvin. The chapter itself gives no affiliation.
• The contributors list (pp. 197–198) gives modus vivendi, Glenlivet, Scotland, and a PhD in ecology.
• Since the mid-1980s he has been a consultant “for industry, environmental non-governmental organisations, the European Commission and the European Environment Agency”, recently on marine eutrophication and fisheries.
• He is also the report’s Executive editor (title page and acknowledgements, PDF pp. 1, 6).
• The author of this case is therefore also likely to have been a significant shaper of the report’s framing (an inference from his title; the report does not describe the division of editorial labour) and, as a member of the credited “Editorial team” (p. 168), of its concluding chapter, which reuses this case repeatedly (pp. 171–172, 175–182).
• The chapter is one of the foundations of the report’s thesis, not an independent test of it. (The acknowledgements list external peer reviewers for the report as a whole, p. 6, but do not say who reviewed which chapter.)
• His own work is used as evidence; none of it appears in a peer-reviewed journal (two are consultancy/NGO reports, one a chapter in an edited book):
• the qualitative cost–benefit account of 19th-century fisheries (MacGarvin and Jones 2000, a WWF-UK recovery plan; p. 19);
• the cost of the cod collapse, “several billion Canadian dollars” (MacGarvin 2001a, modus vivendi/WWF-UK; p. 22);
• the claim that there is no single correct precautionary approach (MacGarvin 2001b; p. 25).
• Disciplinary vantage. He is trained in ecology (PhD) and consults on marine fisheries (pp. 197–198); he writes as a critic of stock-assessment science from the ecology side rather than as a practising stock assessor.
• He sides with the theoretical and community ecology critique of single-species stock assessment (pp. 22–23, 25).
• He favours the ecosystem approach, no-take zones, involving fishers, and investing in recovery (pp. 25–26).
• He cites Greenpeace (Earl 1994) as an “early advocate” of ecosystem-level precaution, presenting it as one legitimate value choice among several rather than endorsing it outright (p. 25).
• Stance. He is pro-precaution, but notably self-critical about how precaution has been implemented.
• His centrepiece is northern cod, where a regime that was precautionary in intent still collapsed (pp. 20, 25). He uses it to question precautionary approaches built on calculated limits and heavy modelling (p. 22).
• He calls the collapse a “systems failure” rather than individual failure (p. 23).
• He is sympathetic to most fishers (pp. 19, 26), pointed but analytically cool about mobile processing capital (depletion “could be argued” to have been their optimal strategy; p. 20), and sharply critical of the Canadian Department of Fisheries and Oceans (DFO) (p. 21).
• He is more diplomatic in tone about EU and ICES (International Council for the Exploration of the Sea) implementation, though still substantively critical (“precautionary gloss”; p. 24).
• Panels and commentaries: none. No industry, regulator or scientific response is included.
• Table 2.1 is credited “Source: EEA”, so the editors, not necessarily the author, compiled it. Most of its departures from the text sharpen the narrative of unheeded warnings or overstate progress, though one row is more sceptical of Canada’s precautionary intent than the text (see Table 2.1 below and the bias check).

Section-by-section notes#

Opening (p. 17)#

• Principle versus implementation. The principle is the same as for pollutants, but implementation is “necessarily very different”. This sets up the principle/approach distinction developed in 2.5 and placed first among the lessons (p. 26).
• The arc. The chapter runs from the Middle Ages, through 19th-century Scotland and the Californian sardine, to northern cod. The author reads it as “history repeating itself”, with some positive lessons.
• The 2001 diagnosis. The need for an explicitly precautionary approach has been “increasingly recognised in the last decade”. In the North Atlantic, the main problem is now helping fishers turn theory into practice against short-term economic pressures. The problem is framed as implementation and economics, not as lack of knowledge.

2.1 Early warnings (p. 17)#

• Framing. The section opens by calling the relationships between precautionary attitudes, cultural perspectives, technological ability, and risks and benefits “complex”, and says there is “nothing new” in varying degrees of precautionary management or in warnings of overexploitation.
• Native North American communities. The source is McEvoy 1986, a single secondary source.
• There is “some evidence” that some communities “may have died out” by overexploiting marine resources.
• Others circumscribed their catches by rights and taboo, reinforced by ecological knowledge, and “maintained substantial catches for centuries” (e.g. of salmon). The chapter does not use the word “sustainable”.
• The commercial fishers who displaced them “extinguished the stocks”.
• The evidence is hedged.
• The 1376–77 English petition (via March 1953).
• It complained of a net so fine-meshed that small fish were caught in such numbers they were fed to pigs, “to the great damage of the whole commons of the kingdom”.
• Parliament’s answer was a commission “to inquire and certify on the truth of this allegation”.
• Later regulation. Later English rules used “still familiar” tools. In 1716 these included a minimum mesh size, a ban on nesting one net inside another, and minimum landing sizes. Both the regulatory toolkit and the ways of evading it are old.

2.2 19th-century British fisheries (pp. 17–19)#

Rapid growth, and uncertainty about herring and trawling, led to official inquiries between 1866 and 1893. The author credits a “‘lay’” Wick historian, Ian Sutherland, with recovering this history (p. 17).

2.2.1 Scottish herring (pp. 17–18) - Natural variability. Herring and sardine swing hugely over decades for natural reasons. So it “took a long time” to accept that fishing could make these cycles worse (p. 17). - Bertram’s evidence, 1865 (primary source). - Between 1818 and 1863, the drift-net area per boat rose from 4,500 to 16,800 sq yd, while the catch fell from 125 to 82 crans (barrels) (p. 17). - My arithmetic: catch per unit of net fell by about 82%. - Bertram wrote: “I have always been slow to believe in the inexhaustibility of the shoals”. He feared the Wick fishery “must come some day to an end”. Nets once carried in a creel now needed “a cart and a good strong horse” (pp. 17–18). - Reassurance from on high (p. 18). - Thomas Huxley, President of the Royal Society and Inspector of Fisheries, is called “no fool, yet now infamous” for saying “marine fisheries are inexhaustible” (1883). In 1881 he said there was “nothing to show” that, “taking an average of years”, herring had ever been more or less numerous. The chapter says this came “several decades” after Bertram, though 1881 is only 16 years later. Both quotations are cited second-hand, via EPAP 1998 and Sinclair 1988. - An 1893 parliamentary committee saw “no indication of any falling off”. - The author’s diagnosis, explicitly an inference: “Presumably landings were equated with stock size”, without allowing for increasing effort, area fished and length of season. - Heincke (1890s) established, “contrary to general belief”, that herring exist as isolated races, which implies managing them race by race. By then the innermost shoals of Wick and the Moray Firth “were gone”, and fishing moved further offshore (p. 18). The science came after the local damage, and the same insight reappears for cod a century later (p. 22). - The 20th century (p. 18). - From 1900 to 1950, North Sea landings were “without trend”. The reason was not lessons learned but turmoil in continental markets, which held back the new motorised drifters. Restraint came from markets, not from governance. - Post-war industrial fishing for fishmeal and oil underestimated “both the new technology and the necessary safety margins”. - By the 1970s herring were near collapse and a moratorium followed. Stocks recovered, and the industrial fishery on herring was restricted when fishing reopened. But pressure from “human consumption” fisheries stayed so high that further emergency measures were needed by the mid-1990s, after which stocks recovered again. - The author’s verdict: by the late 1990s “we appear to have learned enough … to stave off total collapse”, but it was “hardly a model of effective management”. - This partial success is told without dates or figures. The chapter never asks why herring recovered when cod did not.

2.2.2 Steam trawlers (pp. 18–19) - Technology removed natural refuges. Steam trawling for demersal (bottom-dwelling) species reached grounds that sail and oar could not. The author calls these “a natural equivalent of modern ‘no-take’ zones” (p. 18). - Complaints against trawling. Critics said it destroyed spawn, caught immature fish, wasted damaged fish and interfered with other fishers. They also said “mobile outsiders” were depleting others’ traditional grounds. Scientists and fishers were both divided (p. 18). - Voluntary restraint failed. In 1883 Yarmouth trawlermen agreed voluntary seasonal closures to protect immature fish. Other trawlermen undermined the agreement and it was abandoned (p. 18). - Bertram was not anti-technology. In 1865 he himself raised these accusations against trawling, yet he also thought trawling, “used wisely”, was the best tool for some fisheries (p. 18). - The 1883–85 inquiry’s research, by McIntosh (p. 18). - It found that trawling’s catch of immature fish was “insignificant” (less than long-lining’s), that most fish were undamaged for sale, and that of the commercial species only herring spawn lay on the seabed, with injury even there questionable. - Yet McIntosh warned that inshore waters “could very easily be trawled out”. - He proposed suspending trawling there as “a very valuable experiment… upon safe lines, and does no harm to anyone”. Scottish officers, other scientists, some trawlermen and foreign evidence backed him. - The inquiry’s conclusion (p. 18). Although fishers tended to blame each other for any decline, natural fluctuations “largely influence the supply of fish”. The inquiry could not justify prohibition unless experimental evidence “decisively settled” the issue, because of the loss of valuable catches (the author’s paraphrase; only “decisively settled” is quoted in the chapter). - My reading: this is the burden-of-proof moment: the cost of acting is weighed, and the cost of not acting is not. The author’s lesson 2 (“appropriate levels of proof”) cites this episode (p. 26). - Scotland acted anyway (pp. 18–19). - In a political decision, Scotland alone banned inshore trawling, including the Moray Firth. The chapter gives no date. - The ban lasted until the 1920s, when fishers, under pressure from the failing herring market, began dragging seine nets as trawls. - The “initially large landings” of codfish and flatfish that followed “suggested” the ban had nurtured stocks. The author’s own verb is hedged, and the evidence is thin: no figures or comparison are given, and a one-off accumulation of fish on closed grounds does not by itself show a lasting gain in productivity.

2.2.3 Costs and benefits (p. 19) - The analysis is “qualitative”, drawing on MacGarvin and Jones 2000. - Distribution. - Herring-dependent towns such as Wick declined. - The centralisation that came with trawling enriched larger ports (Fraserburgh, Peterhead, Aberdeen; Grimsby, Hull and others) “at the expense of smaller communities”. - Scotland, 1898–1998 (Scottish Office annual reports). - Landings were roughly constant at about 333,000 t, and fleet tonnage at about 109,000 t. - Boats fell from 11,536 to 2,661, and fishers from 36,161 to 7,771. Sail and oar boats fell from 11,383 to nil. - Fossil fuel use rose “enormously”, but no figure is given. - Prosperity “fluctuated widely, but overall was not maintained”. The late-1990s average net profit on capital invested for the Scottish demersal fleet was 0.1% a year (EC 2001 Green Paper, vol. 2). - My arithmetic: boats fell by about 77%, fishers by about 79%, and landings per fisher rose about 4.7-fold. - What was gained? The author asks “just what had been gained by this investment in technology”. - He credits safety: deaths at sea fell markedly. - “In the good years” it “may also have” let some individuals build capital in a remote area and invest it elsewhere for higher returns. “In such circumstances there is no particular interest in maintaining a sustainable fishery.” - For others, “probably the majority”, fishing was a way of life they wished to hand on, with “a sense of being forced down a track whose direction they do not control”. - Changing catches. - North Sea catches shifted from herring for human consumption to demersal fish, then to “industrial” catches (initially herring, sprat and mackerel; later sand-eel), which grew “explosively” from the 1950s. - Industrial catches peaked in the 1970s at 2.2 of 3.5 million t. - The author reads this as part of a global pattern of “fishing down the food web” (Pauly et al. 1998). He surmises that the “vast removal of biomass” will have affected other species too, but “comprehensive data are scarce”.

2.3 Californian sardine, 1920s–1942 (pp. 19–20)#

• Setting (p. 19). California was “precocious”. Reduction to fishmeal and oil drove the economics, and there were no catch limits.
• The state scientists’ warnings (via McEvoy 1986; pp. 19–20).
• In the mid-1920s they urged: “Unnecessary drain upon the supply should be avoided until research has shown that it is possible to detect overfishing in time”.
• They saw government as the only agency uniting “all factions and successive generations”.
• The chapter reports the classic signs, as in Scotland decades earlier: catch not keeping pace with effort, falling average age, longer trips. A catch limit was recommended.
• “While there were uncertainties”, the State Fisheries Laboratory judged the growing intensity of the fishery “fact enough to make us sure that we are headed for destruction and a great loss”.
• The federal view (US Bureau of Fisheries; p. 20).
• It accepted that the evidence of depletion was strong, but said there was “no clear-cut or convincing evidence that will satisfy everyone”.
• It declared “to us conservation means wise use. We do not believe in hoarding our fisheries resources”.
• It said unnecessary restrictions “hamper or restrict legitimate business enterprise”.
• Fishers divided by interest (p. 20).
• Those disadvantaged by the expansion saw depletion as inevitable.
• Expansionists saw a plot to “impoverish one of the few successful enterprises in the Depression”. They blamed environmental fluctuation, cited the fish’s fecundity, and called dismissal of their testimony “brutal… medieval scholasticism”.
• Political override (p. 20).
• In 1939 a new Governor replaced the state experts with “emergency” appointees.
• The reports’ language went from “unmistakable” depletion and “imperative” cuts (1938) to “no reason to be concerned” (1942).
• “It was then agreed to increase the catch (to ‘assist the war-effort’), but in that year the sardine stock collapsed.”
• Recovery began only in the mid-1980s, according to a museum field-guide web page.
• The concluding chapter goes further: “critical agency scientists were dismissed” (p. 179), and a “precautionary programme of stock conservation was reversed with a change in government” (p. 180).
• Economics (p. 20).
• Impacts on other species are, as for Britain, “imponderables”.
• Returns were “extraordinary”: many plants recovered their whole investment in one season (McEvoy).
• For processors such as Starkist and Van Camp, “it could be argued” that “mining” the fish and moving to South America was the optimal strategy.
• There they helped open the Peruvian anchoveta fishery, which “followed the Californian model” and collapsed in the early 1970s.
• This is the chapter’s clearest statement that depletion can be rational for mobile capital. It is offered as an argument, not a verdict.

2.4 Newfoundland northern cod (pp. 20–23)#

Set-up (p. 20). - After California, managers turned to “increasingly complex mathematics” and more computing power, and by the 1970s there was optimism that past mistakes could be avoided. - Northern cod was historically the world’s largest cod stock, fished by Europeans since the 16th century. Fishing intensity grew “dramatically” in the 1960s; catches peaked at 800,000 t in 1968, then fell “well below” the levels the international regulator actually authorised. (My reading: the authorised levels had stopped binding, consistent both with depletion and with the regulator’s reputation for ineffectiveness; the chapter does not spell out the implication.) - The international regulator, the International Commission for the Northwest Atlantic Fisheries (ICNAF), was “regarded by many Canadians as ineffective” (O’Reilly Hinds 1995). The author adds that this was “also true” of its successor, the Northwest Atlantic Fisheries Organization (NAFO) (Day 1995). - Canada extended its jurisdiction from 12 to 200 nm in the late 1970s. The chapter says it did so “using the then novel UN Convention on the Law of the Sea”; accuracy flag: UNCLOS dates from 1982. - Canada set “deliberately conservative” limits of about 20% of the stock, to rebuild it. - The author says these were measures “many would even now consider as precautionary”. - DFO’s calculations showed rebuilding, and offshore catches by Canadian trawlers increased. In 1988 DFO claimed a “five fold increase in Northern Cod since 1976”. The stock was held up as a model of cautious, science-driven management.

2.4.1 Warnings and responses (p. 21). - Inshore fishers dissented, and were “the only dissenting group”. Unlike the offshore fleet they had “generally not upgraded their technology”, and they could not square their falling catches with the reported growth. Their protests were disregarded. - They commissioned the Keats Report (1986). The reference list gives its authors as Keats, Steele and Green, as a report to the Newfoundland Inshore Fisheries Association, cited via Finlayson 1994 (p. 28); the chapter does not say who the authors were professionally. - It highlighted DFO’s own “(downplayed)” retrospective analyses, which indicated “consistent and severe” underestimation of fishing pressure since Canadian control began, “with the result that we have consistently taken from 1.5–3 times the (20 % of stock) catch since 1977”. - DFO dismissed it “as biased pseudoscience written to support a political agenda”. - Media attention then forced the minister to commission an official review. - The Alverson Report (1988) had authority, “in official eyes”, “because it was prepared by fisheries scientists” (the qualifier implicitly contrasts it with Keats). - Body: the stock had increased since 1977, although after 1982 “probably only very slowly”. But fishing mortality was “considerably in excess of target mortality” because of “consistent overestimation of the current stock size”. Alverson showed how sensitive stock-size conclusions were to the wide range of fishing-mortality estimates the data allowed. - Technical core (the author’s explanation): a year’s estimates converge on the truth only after about five more years of data. Until then they largely echo the fishing mortality that was assumed. This is “a critical flaw” when a depleted stock depends on the last few years’ breeding success (“now commonplace for many stocks”), and during a sharp decline “the method instils a false sense of assurance”. - Executive summary: in the author’s words, “this was turned on its head”. What was inverted is specifically the sensitivity demonstration: the width of the plausible range was used to say that DFO’s fishing-mortality estimates fell “within the range of estimates supported by the data”, “albeit at the lower end”. The chapter gives no specific citation for this paragraph, and the Alverson Report does not appear in the reference list (pp. 27–30), so its quotations cannot be traced from the chapter. - Managed release: DFO “had responsibility for the management of the release”. Alverson himself said publicly “it’s rather amazing that we are as close to each other as we are”. DFO stressed a difference of only “4 to 5 percent” and said “the credibility of DFO science was not questioned”. Privately, the Alverson team were “less sanguine”, and there was an internal reappraisal of DFO methodology. - DFO’s 1989 reassessment. The stock was not growing, and the offshore catch should be “virtually halved”. This was read as an admission that DFO “had got it wrong all along”. - It damaged an administration that relied on the “‘science’ touchstone as an arbiter to conflicting claims to resources”. - Now it was the offshore fishers who protested that there was no evidence of decline. - The Harris Report (1990), “fully independent of DFO control”. - Before 1989, fishing mortality had “probably” been more than double the intended level, and the stock little more than half its assumed size, so the stock had been fished at levels pointing towards commercial extinction. The media reported this widely. - Harris “cautiously concluded” that DFO’s revised 1989 assessment was “a better approximation of reality”, but set out major, hard-to-resolve problems “at every conceivable level”. - Harris wrote that “we acted in substantial ignorance of the animals… and in almost total ignorance of the dynamics of the ecosystems”. - They had worn “rose tinted glasses”, reading data to confirm “the model of growth upon which our hearts had been set”. - The best available methods “will legitimately support claims of stock status ranging from sustainable growth to dangerous decline”. The author’s framing inserts “(and often still are)”, extending Harris’s point to 2001.

2.4.2 Costs (pp. 21–22). - The catch-limit trade-off. - Meeting the 20% goal needed the total allowable catch (TAC) cut from 235,000 t (1989) to about 125,000 t (1990). Harris judged this “drastic” socially and economically. - He proposed 190,000 t (about 30% removal), warning “this may contribute to further decline”. That figure was adopted. - Even this cut from 235,000 t came “at a loss of” CAD 26 million (about EUR 21 million) in landings, CAD 66.6 million (EUR 53 million) in processed product, and the equivalent of about 1,000 jobs. (These are the costs of the cut actually made; the chapter gives no figure for what the deeper cut to 125,000 t would have cost, only Harris’s “drastic”.) Similar limits followed for 1991–92. - Collapse and aftermath. - During the 1992 season it became apparent there was “little left to catch”, and the situation was “far worse than even the most pessimistic projections”. An emergency moratorium came in July 1992, “initially for two years”. - The stock did not rebound. Only a 9,000 t inshore fishery was allowed, in 1999. - The costs in the 1990s, “including lost sales, unemployment benefit and financial assistance”, were “in excess of several billion Canadian dollars” (MacGarvin 2001a, the author’s own report). - DFO’s 2000 assessment. - The 1999 catch already exceeded the 20% reference level, “unacceptable under a precautionary approach”. Even a monitoring fishery carried risk. - Little is known about “this most studied stock”: why it collapsed or failed to recover, or what proportion is taken by predators and other fishing activities. Its prey, capelin, are in poor condition. - The inshore–offshore link, “the source of the original controversy”, is “now open to question”. There is “growing awareness” that the “stock” is made up of “more or less discrete local populations” (compare Heincke), with “considerable implications for recovery programmes”. - Positives. - Assessments are now freely available, with the uncertainties “clearly set out”. - “Attempts are being made” to incorporate fishers’ “‘lay’ assessments of stock strength” (DFO 2000). - A policy review is under way. - Approaches less dependent on assessment accuracy, such as no-take zones, are being evaluated.

2.4.3 The human dimension (pp. 22–23). - A challenge to model-based precaution. The cod case “challenges the likely success” of precautionary approaches, “still advocated”, built on calculated limits to fishing mortality and heavy modelling (p. 22). This is the author’s most important self-critical claim. - Source. Finlayson (1994), Fishing for truth, a sociological interview study “completed just before the final collapse”. The author frames what follows as “the views quoted there”, i.e. participants’ accounts as reported by Finlayson. - DFO scientists warned of uncertainty early, but “in the event not loudly enough”. - Their objections that the long-term forecasts demanded of them were impossible were overridden. They “apparently took seriously the threat” that otherwise “economists would do it for them”. - Double pressure to downplay uncertainty. - Politicians wanted “constancy and certainty” to settle disputes. - Scientists expected the top of any range to be taken, so they leaned towards lower (in their view precautionary) estimates, stated more precisely than their internal assessments justified. - This “hoist them on their own petard” when even their most pessimistic view rested on underestimates of historic fishing mortality. - Known weaknesses. Scientists also knew of shortcomings in biological and physical parameters, of “fundamental sampling problems”, and of “often dubious” advice produced for many stocks “in a few weeks of intensive meetings”. - The Ptolemaic analogy. Harris compared fisheries science to Ptolemaic astronomy, adding complexity rather than questioning the theory. This is cited to Finlayson (ibid., p. 69), so it is second-hand. It “accorded with strong criticism” from theoretical ecology (Peters 1991). - Verdict. Participants were not “stupid, careless or lacking in intent to restore the stock”. It was “a systems failure” that stopped fishers, scientists and politicians from responding to existing information “or extracting themselves from the situation” (p. 23).

2.5 Precaution becomes explicit (pp. 23–25)#

• Two 1995 instruments. By the early 1990s precaution had become explicit, owing both to disasters such as northern cod and to its “increasing prominence in other fields”. The most significant global developments were the 1995 FAO Code of Conduct for Responsible Fisheries and the UN Agreement on Straddling Fish Stocks and Highly Migratory Fish Stocks (p. 23).
• “Approach” versus “principle” (p. 23).
• The FAO Code’s “precautionary approach” reflected industry and managers’ nervousness that the “principle” had been captured by environmental NGOs and “might unjustifiably be used as a weapon to substantially reduce or even halt fishing”.
• The author reads the wording as “probably” mainly a statement about “ownership of the process”. But he sees real value in separating a universal principle from implementation that differs by field, while noting that this “was not a distinction made by those who negotiated the Code”.
• The Code’s mechanism (p. 23). Precaution in the Code nominally covers uncertainty about individual stocks, other affected species, and environmental and socio-economic conditions, but “the aspect emphasised” is handling uncertainty in stock-assessment models: target and limit reference points, with actions agreed in advance if they are approached or exceeded. These were “novel developments”.
• The UN Agreement (p. 23).
• The fishing rate that yields maximum sustainable yield (F_MSY) should be “a minimum standard for limit reference points”. The author explains that it is a minimum because the old MSY concept “is known to overestimate sustainable yields” (his gloss, not the Agreement’s text).
• “The implications are profound”, because many stocks are depleted “well beyond formal MSY”.
• He quotes the Agreement: for stocks not overfished, fishing mortality should not exceed that corresponding to MSY; for overfished stocks, the biomass producing MSY “can serve as a rebuilding target”. He calls this a “possible contradiction” without spelling it out (my reading: MSY treated as a limit in one clause and a target in the other), and puts it down to negotiated compromise, not “a single consistent mind”.
• Overall, the UN Agreement and, indirectly, the FAO Code “emphasise precaution at the level of maximising long-term yields” (pp. 23–24). This sets up the contrast in 2.6.
• US (p. 23 and n. 3). US federal guidelines (Restrepo et al. 1998) recognise that stocks should be managed consistently with the UN Agreement and FAO Code. The author flags the US position as “of interest beyond fisheries”.
• Restrepo et al. 1998 requires protective measures “even in the absence of scientific certainty”. Previously there was a “perceived… inability” to act without proof of overfishing.
• “Thus, the precautionary approach is essentially a reversal of the ‘burden of proof’.”
• Canada (p. 23). Stock and fishing rate are held to MSY levels, or proxies.
• EU: split responsibility (p. 24). ICES advises, while the Commission and Member States manage.
• The reference points. ICES advice centres on B_lim and a larger B_pa (with matching F_lim and F_pa), “at least according to interpretations arising from within the FAO” (Garcia 2000). B_lim is set not at MSY but at the point where a stock “is in immediate danger of collapse”. B_pa “is set inconsistently”; for stocks in poor condition (North Sea cod, haddock, plaice) it sits at or near the old minimum biologically acceptable level (MBAL).
• Where stocks stand. Many stocks are below B_pa and some near B_lim (e.g. North Sea cod in 2000).
• Peer criticism. Criticism from within the FAO, the US and Canada saw this as conflicting with the FAO/UN approach. The earlier regime “often remained unchanged, although the language has been given a precautionary gloss”.
• ICES’s own view (hedged). The author says ICES “may concur”, citing an ICES-published working-group report (ACFM ICES 2000b, the same CWP meeting report that contains Garcia 2000) calling this use of limit reference points “a needlessly restricted interpretation of a concept”, and noting “obliquely” that precautionary reference points require “discussion with fisheries management agencies”. He takes this to illustrate “the tensions that arise from split responsibility”. ICES calls B_pa a “buffer” or “precautionary reference point”, not a “target”, “presumably” to avoid treading on the managers’ role.
• Crisis management. A 1995 criticism by the EU’s own Scientific, Technical and Economic Committee for Fisheries (STECF), of a system “at ease with crisis management”, “still appears valid”. Member States seem unwilling or unable to fund the short-term investment needed to restore stocks.
• The author’s rebuttal. The Commission’s 2000 communication says ICES does not refer to MSY because, for “a number of stocks”, its conditions are “difficult, if not impossible” to define, and because fishing pressure on many EU stocks is far above it. The author offers “a likely response”: MSY is “no more or less certain” than collapse thresholds, and high pressure “is the reason why the code is necessary, not a reason why it should not be applied”.
• Tone. He warns the EU–FAO/UN relationship “has the potential to become a highly contentious issue”, but praises the 2001 Green Paper’s “frank identification of the problems” and suggests “concentrating on the future rather than dwelling overmuch on the past”. This is gentler than his treatment of DFO.
• Data quality (pp. 24–25).
• Data quality “remains unchanged” as a problem. For northern cod, underestimating fishing mortality “was a key factor”.
• A 1999 retrospective analysis of the best-documented North Sea stocks (cod, haddock, whiting, plaice, sole) found “the same fundamental problems”. But the result was mixed: fishing mortality for cod, haddock and whiting was much higher than originally stated; for plaice there was no correlation, and for sole a possible negative one. So the one-directional bias held for three of the five stocks.
• The analysis also linked underestimated fishing mortality to overestimated stock size. The problem had been “identified as early as 1977”. “The parallels are startling.”
• The conclusion that Northeast Atlantic cod can tolerate more fishing than northern cod “seems to be based on the circular argument that they have not yet crashed”. Northern cod, too, had survived periods of fishing heavier than at the time of its crash.
• Canada’s post-1977 goal was “identical” to today’s precautionary approach. “Yet still it crashed”, so it is “legitimate to ask whether the change — as yet — is sufficient” (p. 25).

2.6 The ecosystem approach (pp. 25–26)#

• Precaution for what? UN/FAO ecosystem precaution is nominal; practice remains single-stock. Precaution might aim to:
• avoid collapse (ICES/EU);
• keep stocks above MSY (UN/FAO);
• protect dependent species (Greenpeace 1994; Fowler 1999, who argues catches should be comparable to those of other predators).
• “There is thus no one ‘correct’ precautionary approach; it depends on the objectives set”, and judgements vary with “activities and interests” (p. 25).
• Interdependence. Stocks prey on each other, and single-species models fall “far short of the evident complexity”.
• The gulf between fisheries science and ecology is “remarkable”.
• An unattributed quote from “prominent ecologists” in the early 1990s calls fisheries management a field “so accustomed to inaccuracy in its basic models that striking differences between model and observation are scarcely noted”, where biologists fit data to models that are “clearly inaccurate” and decide on that basis. Citation gap: no source is given (possibly Peters 1991, cited on p. 23, but the chapter does not say; to verify).
• Signs of change. The US Ecosystems Principles Advisory Panel (EPAP 1998), ICES’s Advisory Committee on the Marine Environment (ACME 2000) and Canada are bringing in ecologists.
• Chaotic dynamics may make systems “fundamentally unpredictable”.
• An earlier Grand Banks study (Gomes 1993) “suggests” that the more realism is built into multispecies models, the less predictable the effect of changing any one species: “in effect the system may work as a gigantic random number generator”.
• What is known (per the US report). Ecosystems have limits beyond which change can be irreversible, diversity matters, systems operate at multiple scales, and boundaries are indistinct.
• EPAP: uncertainties “are not acceptable excuses to delay implementing an ecosystem-based management strategy”. European experts (Daan 1998) reached “similar conclusions”.
• Hjort’s 1914 problem of predicting recruitment from egg numbers is still unsolved. Yet “according to the US approach”, we know enough about ecosystems to manage better than in the past.
• Burden of proof. In 1919 a Californian regulator (Thompson) required that proof “that seeks to change the ways of commerce and sport must be overwhelming”. Now, in the US approach, “the burden of proof is on fisheries to take account of uncertainty regarding ecosystem effects” (p. 25).
• The way forward. “At an operational level stakeholder involvement is essential” (this follows the account of the US approach; it is unclear whether it is EPAP’s view or the author’s).
• A “second generation” precaution, less model-centred and using “error-resilient” tools such as no-take zones, “appears to be producing results” on Georges Bank (Murawski et al. 2000).
• Canadian fisher involvement and northern European ecosystem policy are “at an early stage”, “without overall consensus”, but have “considerable potential if implemented” (p. 26).

2.7 Late lessons (p. 26)#

• The 14 lessons are listed below under “The authors’ own lessons”.
• The closing paragraph makes five points:
• Positive changes are happening, but may not come fast enough to “stave off further collapses”.
• Although “some interest groups” will see a short-term strategy as in their interest, the need for precaution is “generally not something that should need to be laboured within the fishing industry”.
• Not cutting stocks “close to the bone” gives “greater certainty” about maintaining stocks and would also “greatly increase the economic returns” (Whitmarsh et al. 2000).
• Natural capital is so depleted that industry “in many cases cannot absorb the short-term hit” of rebuilding.
• Interest groups should stop arguing with each other and engage with economic ministries. Those ministries hold “a highly jaundiced view” of recovery investment because of past “inappropriate subsidy”.

Table 2.1 (p. 27; “Source: EEA”)#

• Structure. The table has four blocks:
• early UK fisheries (1376–77; 1866–93);
• the Californian sardine (mid-1920s; 1942);
• northern cod (late 1970s; 1986; 1988; 1989; 1990; 1992; 1999);
• general fisheries (1990s; 1995; 2001).
• Where it departs from the text: 1. For 1866–93 it says “no action is taken”, but the text reports Scotland’s inshore trawling ban (p. 18). 2. For 1942 it says “Continuous inaction leads to the collapse”. The text never states the cause so baldly, and it records the environmental argument (p. 20). 3. For 1999 it attributes concern about the catch level to “dissidents”. In the text the concern comes from DFO’s own 2000 assessment (p. 22). 4. For the 1990s it says the ecosystem approach “slowly settles” into management. The text calls it early-stage and without consensus (p. 26). 5. For the late 1970s it says Canada managed the stock “claiming it is doing so precautiously”. The text is less sceptical: it calls the measures ones “many would even now consider as precautionary” and makes their genuine precautionary intent the point of the case (pp. 20, 25). This row cuts against the chapter’s self-critical argument rather than for precaution.
• Direction of the departures. Rows 1–3 sharpen the narrative of unheeded warnings (1 and 2 by overstating inaction; 3 by moving an official admission onto “dissidents”). Row 4 overstates progress. Row 5 is more sceptical of Canada’s intent than the text. The 1988 row (“consistent overestimation of the stock size leads to overfishing”) is accurate to the Alverson body but omits the softened executive summary.

Evidence base (references, pp. 27–30)#

• Primary sources: Bertram 1865; the 1866 and 1885 Royal Commission reports; Scottish Office statistics; Harris 1990; Thompson 1919; the FAO and UN texts; DFO 2000; ICES/ACFM documents; EC documents; Restrepo et al. 1998; EPAP 1998.
• The vivid narratives are single secondary sources:
• California: McEvoy 1986;
• DFO behaviour, Keats and the interviews: Finlayson 1994;
• Huxley: second-hand twice;
• the 1376 petition: March 1953.
• Grey literature is prominent: the author’s WWF-UK reports, Greenpeace, a museum web page, and ICES internal documents (one cited by internal network path).
• Missing references: the Alverson Report (1988), central to 2.4.1, is not in the reference list. The “prominent ecologists” quotation (p. 25) has no citation.

Reuse in the concluding chapter (cross-reference only; Ch. 16 is credited to the “Editorial team”, p. 168)#

• p. 171. Fish stocks “can take a long time to recover from a crash, if ever” (irreversibility).
• p. 172. The Canadian multispecies model (Gomes) as an example of research compounding uncertainty.
• p. 175. Circumvention of fisheries regulation; single-stock models that do “not reflect actual conditions”.
• p. 176. For cod (and BSE), “the pros and cons of taking early action were identified prior to the event”. Limited action was taken because of uncertainty and cost, and far greater costs followed (“many of which were anticipated”). Even without collapse, letting depleted stocks recover brings “major economic, social and environmental benefits”.
• p. 177. Fishers “can be less precautionary” than others, but many are blocked from acting precautiously.
• p. 178. Precautionary objectives depend on the value judgements of interest groups.
• p. 179. Californian “critical agency scientists were dismissed”, as an example of “shooting the messenger”.
• p. 180. The sardine programme was “reversed with a change in government”; the federal Bureau “strongly opposed” the state’s precautionary action as unduly inhibiting commercial activity.
• p. 181. The EPAP “not acceptable excuses” quote is reused.
• p. 182. Fisheries listed among cases showing a differing US regulatory approach.

Case timeline#

A. Scottish herring and trawling | Date | Event | Type | p. | |—|—|—|—| | 1376–77 | Petition against fine-mesh net; commission of inquiry | Lay warning; inquiry | 17 | | 1716 | Mesh, anti-circumvention and landing-size rules | Regulation | 17 | | 1818–63 | Net area ×3.7, catch −34% per boat | Data | 17 | | 1865 | Bertram’s quantitative warning | Early warning (strong, one author) | 17–18 | | 1866–93 | Official inquiries | Inquiry | 17 | | 1881/1883 | Huxley: no change in herring; “inexhaustible” | Expert reassurance | 18 | | 1883 | Yarmouth voluntary closure collapses | Self-regulation fails | 18 | | 1883–85 (chapter: “by 1883”; report 1885) | McIntosh: most complaints unfounded, but inshore could be “trawled out”; suspend as experiment. Inquiry requires evidence that “decisively settled” the issue | Split advice; burden on restriction | 18 | | Undated | Scotland alone bans inshore trawling | Action against official advice | 18 | | 1890s | Heincke: separate races; inner shoals already gone | Science after the damage | 18 | | 1893 | Committee: “no indication of any falling off” | Official reassurance | 18 | | 1900–50 | Landings trendless; market turmoil restrains technology | Non-regulatory restraint | 18 | | 1920s | Ban eroded by seine-as-trawl; large landings “suggest” benefit | Weak evidence | 18–19 | | 1970s | Herring near collapse; moratorium; recovery; industrial herring fishery restricted on reopening | Late action | 18 | | Mid-1990s | Human-consumption fishing pressure forces further emergency measures; recovery | Action | 18 | | 1898–1998 | Flat landings; boats −77%, fishers −79%; 0.1% return | Outcome | 19 |

• Lag. About 25–35 years from Bertram’s warning to the loss of the inner shoals (by the 1890s), with no herring measure reported in the chapter in between. The North Sea moratorium came about 110 years after Bertram, triggered by near-collapse rather than by warning; but it concerned a different, offshore and industrial fishery, so this is a pattern of repetition rather than a lag in the strict sense.

B. Californian sardine | Date | Event | Type | p. | |—|—|—|—| | Mid-1920s | State scientists: classic signs; limit recommended; “headed for destruction” | Early warning (strong, institutional) | 19–20 | | Undated in chapter (1920s–30s) | Federal Bureau: evidence of depletion “strong” but not convincing “to everyone”; “wise use” | Opposition from another level of government | 20 | | 1930s | Plants recover investment in one season; expansionists cry “plot” | Incentives; counter-claims | 20 | | 1938 | “Unmistakable” depletion; cuts “imperative” | Warning | 20 | | 1939 | Governor replaces experts with “emergency” appointees | Political override | 20 | | 1942 | “No reason to be concerned”; catch increase agreed for war effort; collapse “in that year” (the chapter’s date) | Collapse | 20 | | Later | Capital moves to Peru; anchoveta collapses early 1970s | Model exported | 20 | | Mid-1980s | First recovery signs | ~40+ years to recovery | 20 |

• Lag. Roughly 15–18 years from the first warnings (mid-1920s) to the collapse (the chapter’s dating; see caveat), with no catch limit ever mentioned as imposed. Recovery took about four decades.

C. Northern cod | Date | Event | Type | p. | |—|—|—|—| | 1968 | Peak catch of 800,000 t; catches then fall well below authorised levels | Overexploitation | 20 | | 1977 (chapter: “late 1970s”; Keats quote implies 1977) | 200 nm limit; ~20% “deliberately conservative” target | Action precautionary in intent | 20 | | From 1977 | Removals 1.5–3× target, visible in DFO’s own downplayed analyses | Buried internal warning | 21 | | “Early stage” of Canadian management (after 1977; not dated more precisely) | DFO scientists warn of uncertainty “not loudly enough” | Weak internal warning | 22 | | Pre-1986 | Inshore catches fall; protests disregarded | Lay warning | 21 | | 1986 | Keats Report; dismissed as “pseudoscience”; media | Strong warning, contested | 21 | | 1988 | DFO claims 5× increase; Alverson finds stock grew only slowly after 1982 and fishing mortality well above target; summary “turned on its head”; release managed | Review and messaging | 20–21 | | 1989 | DFO: stock not growing; “virtually” halve offshore catch | Internal correction | 21 | | 1990 | Harris: fishing mortality “probably” >2× intended; TAC 190,000 t vs 125,000 t needed | Independent review; partial action | 21–22 | | July 1992 | Moratorium | Action after collapse | 22 | | 1990s | No recovery; >several billion CAD in costs | Harm | 22 | | 1999–2000 | 9,000 t fishery; DFO says even this exceeds reference level | Continuing fragility | 22 |

• Lag. Keats to moratorium: 6 years. DFO’s own reappraisal to moratorium: 3 years. The biased estimates had been in use since 1977, 15 years earlier. Effective action came only once there was “little left to catch”.

D. North Sea assessment bias. - 1977: the problem is identified (the chapter does not say by whom or for which stocks). - 1999: it is confirmed for cod, haddock and whiting, but not for plaice (no correlation) or sole (possible negative correlation). - 2000: North Sea cod is near B_lim (p. 24). - About 22 years passed; the chapter implies, without stating outright, that the bias had not been corrected (“the parallels are startling”).

E. Burden of proof. - 1919: Thompson’s “overwhelming” proof standard (p. 25). - 1995: the FAO Code and UN Agreement (p. 23). - 1998: US guidance reverses the burden of proof (p. 23 n. 3). - About 80 years from one to the other.

What was known, when (my synthesis across the timelines, not the author’s formulation). In each of the three main cases the core signal was available well before collapse: effort outpacing catch, a younger age structure, longer trips, and DFO’s own retrospective analyses. What was missing was: - acceptance of that evidence under the standard of proof being applied; - willingness to bear short-term costs; - in Canada, an agency without a stake in defending its earlier numbers.

The authors’ own lessons and conclusions#

A. The 14 stated lessons (p. 26), with the author’s case attributions 1. Principle versus approaches. A logical distinction, and a political one (FAO). 2. Appropriate levels of proof. 19th-century Scotland; 1920s California; the 1990s US ecosystem approach. 3. Uncertainty versus ignorance. Harris; uncertainties attached to estimates versus ignorance of ecology. 4. Unrealistic expectations, or “incredible claims”, of scientific soundness. Cod, “but a general feature”. 5. Drawing on historical knowledge. Heincke; natural no-take zones; areas protected from trawling. 6. Not brushing blind spots under the carpet. Harris’s rose-tinted spectacles. 7. Avoiding dominance by one discipline. Stock modellers. 8. Accounting for real-world conditions. Underestimated fishing mortality and technology. 9. Weighing the pros and cons of each approach. Stock assessment versus wider approaches. 10. Using lay knowledge. Native Americans; Scottish fishers; Canadian inshore fishers. 11. Wider social perspectives, value judgements and all the options. Different interpretations of the precautionary approach. 12. Not relying on ever more elaborate models to explain away failures. The Ptolemaic analogy. 13. Institutional obstacles and regulatory independence. Reluctance to address fundamental economic issues; blurred adviser/policy-maker independence “from Californian sardines to present”. 14. Due humility. DFO’s “biased pseudoscience” response.

B. Other conclusions in the body of the chapter - Cod was a “systems failure” (p. 23). - Precautionary approaches that rely on calculated limits to fishing mortality and heavy modelling are challenged by the cod case. Since Canada’s precautionary-in-intent regime still crashed, “it is legitimate to ask” whether current changes suffice (pp. 22, 25). - There is no single “correct” precautionary approach (p. 25). - North Sea herring was “hardly a model of effective management” (p. 18). - The EU has a “precautionary gloss” on an often unchanged regime (p. 24). - The underestimation of fishing mortality was identified in 1977 and still present in 1999 for three of five major North Sea stocks (p. 24). - Depleted natural capital blocks recovery, so economic ministries are the key audience (p. 26).

C. Derived lessons versus advocacy - Conceptual or procedural rather than case-derived: lessons 1, 9 and 11 (the principle/approach distinction; weighing the pros and cons of approaches; value judgements and options). They draw on the chapter’s governance discussion (pp. 23–25) rather than on a documented failure. - Derived from documented episodes: lessons 2, 3, 4, 6, 8, 12, 13 and 14 (the inquiries; the Keats, Alverson and Harris reports; the 1999 retrospective; the 1939 replacement of experts). - Derived but thinly evidenced: - Lesson 5: the Scottish ban’s “success” rests on “suggested”, p. 19. - Lesson 10: the Native American case is hedged and single-source. The Canadian inshore fishers’ observation was carried into the record by a commissioned scientific re-analysis, and the inshore–offshore link is itself now “open to question” (p. 22). - Advocacy or recommendation: - “Second generation” precaution and ecosystem management (hedged: “appears”, “considerable potential if implemented”; pp. 25–26). - Stakeholder involvement as “essential” (p. 25). - Engaging with economic ministries (p. 26). - Higher returns from lower fishing pressure (a single citation; p. 26). - The conciliatory line on the EU (p. 24). - Lesson 7 is partly an argument for the author’s own discipline, although the Harris and North Sea evidence supports it.

Mechanisms and dynamics#

1. Output mistaken for stock. - Rising effort and technology kept landings up while the stock fell. Landings were then read as abundance (p. 18; the author’s inference, “presumably”). - Effort-adjusted data showed decline decades earlier: Bertram (p. 17) and the Californian signs (p. 19). - The unmodernised inshore fleet saw what the modernised offshore fleet’s catches hid (p. 21). Caveat: DFO in 2000 says the inshore–offshore link is “now open to question” (p. 22). - Technology also removed de facto refuges (p. 18). New technology was “underestimated” in the post-war herring fishery (p. 18), and lesson 8 generalises this (“underestimation of real fishing mortality and technologies”; p. 26).
2. Estimates that are right only in hindsight. - Stock assessments converge only after about five more years of data. Recent estimates therefore echo the assumptions and give false assurance during a decline (p. 21). - For northern cod the bias ran consistently in one direction (p. 21). In the North Sea it recurred in three of five well-documented stocks (not plaice or sole), and the problem had been known since 1977 (p. 24). - It produced circular reassurance: “not yet crashed” was taken to mean “can take more” (pp. 24–25).
3. Natural variability as both a real confounder and a shield. - Genuine fluctuation delayed recognition of fishing’s effect (p. 17). - The same fact was used by an inquiry and by expansionist fishers to resist restriction (pp. 18, 20). - The chapter offers no way to tell honest doubt from strategic doubt.
4. Asymmetric standards of proof. - “Decisively settled” (p. 18), “satisfy everyone” (p. 20) and “overwhelming” (p. 25) are near-unattainable in a highly variable system, so in practice they default to continuation (my reading; the chapter juxtaposes these standards but does not state this mechanism in general form). - The costs of restriction were itemised: lost catches, “drastic” repercussions. The costs of collapse were counted only afterwards (p. 22). - The counter-moves were the 1998 US reversal of the burden of proof (p. 23 n. 3) and EPAP’s “not acceptable excuses” (p. 25). - McIntosh reframed restriction as a low-regret experiment, a way of producing knowledge (p. 18).
5. Positional testimony. - Fishers divided according to how change affected them (pp. 18, 20). - In Newfoundland, dissent switched from inshore to offshore fishers when the official view shifted (p. 21). - Fishers’ testimony was right in some cases (Canadian inshore fishers) and wrong in others (Californian expansionists, Newfoundland offshore fishers in 1989). (The chapter does not present Bertram as a fisher or as an instance of lay knowledge.)
6. Commons and collective-action failure. - Unilateral restraint unravelled (Yarmouth, p. 18), and “mobile outsiders” depleted others’ grounds (p. 18). - Rights and taboos sustained Native catches (p. 17). - My reading: binding institutions, not goodwill, made the difference; the chapter juxtaposes the cases but does not draw this conclusion itself. The concluding chapter’s remark that many fishers who wish to act precautionarily are “prevented from doing so because of a systems failure” (pp. 177–178) is consistent with it.
7. Mobile capital and diverging time horizons. - One-season payback made “mining” and relocating rational for processors. The model was exported to Peru, with the same result (p. 20). - Some Scottish capital-holders had “no particular interest” in sustainability, while place-bound communities felt “forced down a track” (p. 19).
8. A technology treadmill with dissipated gains. - A century of investment produced the same landings with far fewer fishers, more fuel, and 0.1% returns (p. 19). - Fishing moved sequentially down the food web (p. 19). - In the first half of the 20th century, the main restraint on herring came from market turmoil, not regulation (p. 18).
9. Political control of expertise and split jurisdiction. - A change of governor flipped the official assessment within four years (p. 20). - The federal–state disagreement left the dispute “unresolved” for years; the permissive outcome came when the state’s own experts were replaced (p. 20). - The EU’s adviser/manager split produced euphemism (“buffer” rather than “target”) and often unchanged practice (p. 24).
10. The certainty trap.
    • Politicians demanded certainty. Scientists, expecting the top of any range to be used, gave precise low numbers.
    • When these failed, both the decision and the scientists’ credibility failed with them (p. 22).
    • Hiding uncertainty in order to be precautionary backfired. Honest communication of uncertainty and control of range-exploitation are separate problems.
11. Institutional face-saving.
    • An agency whose authority rested on its science (p. 21) resisted revision: critics called “pseudoscience”, a managed release, a summary softer than the body, and “credibility… not questioned”.
    • Correction needed media pressure and a fully independent review (p. 21).
12. Model lock-in and disciplinary monoculture.
    • More mathematics and computing brought “optimism that past mistakes could be avoided” (p. 20). Failed predictions met added complexity (the Ptolemaic analogy, p. 22).
    • Data were read to confirm hoped-for growth (p. 21).
    • Neighbouring ecology saw the problem, but the “gulf” between the fields was “remarkable” (pp. 23, 25).
    • Advice was produced for many stocks in a few weeks (p. 22).
13. Uncertainty versus ignorance, and limits to prediction.
    • Ignorance was deep: of ecosystem dynamics (p. 21); of recruitment, unsolved since 1914 (p. 25); of stock structure, known for herring in the 1890s but only now being applied to cod (pp. 18, 22).
    • Chaotic dynamics and “random number generator” behaviour limit prediction (p. 25).
    • Because the data “legitimately support” everything from growth to decline (p. 21), interests and hopes selected the reading.
14. Cost aversion and the rebuilding trap.
    • Harris split the difference (190,000 t rather than 125,000 t) to avoid “drastic” social and economic repercussions. Even that cut cost tens of millions of dollars and about 1,000 jobs; the collapse that followed cost billions (p. 22). The chapter does not quantify what the deeper cut would have cost.
    • The chapter does not show that the lower figure would have sufficed (“far worse than even the most pessimistic projections”).
    • Depletion erodes the capacity to pay for recovery, and funders are wary because of past subsidy (pp. 24, 26).
15. Framing and language.
    • Reassurance: “inexhaustible” (p. 18); “no reason to be concerned” and “five fold increase” (p. 20); “credibility… not questioned” (p. 21).
    • Ideology: “wise use” versus “hoarding”; restrictions that “hamper… legitimate business enterprise”; catch increases to “assist the war-effort” (p. 20).
    • Discrediting critics: “biased pseudoscience… political agenda” (p. 21), and conversely scientists accused of “medieval scholasticism” (p. 20).
    • Terminology as territory: “approach” versus “principle”, “buffer” versus “target”, a “precautionary gloss” (pp. 23–24).
    • Precautionary self-description (“deliberately conservative”; p. 20) co-existing with overexploitation.
16. Irreversibility and slow, uneven recovery.
    • The sardine took about four decades to recover (p. 20). Cod showed no rebound eight years after the moratorium (p. 22).
    • EPAP warns that exceeding ecosystem limits may be irreversible (p. 25).
    • Herring did recover after moratoria (p. 18), but the chapter does not explain the difference.
17. Learning, forgetting and repetition.
    • “History repeating itself” (p. 17).
    • Heincke’s insight about sub-stocks resurfaces for cod only a century later (p. 22), and the author lists it under “drawing upon historical knowledge” (p. 26). The 1977 bias warning apparently went unheeded (p. 24). The Californian model was re-run in Peru (p. 20).
    • Late learning did happen: open assessments, fisher knowledge, no-take trials, and predetermined reference points (pp. 22–23).

Transferable insights (technology-neutral)#

1. Rising capability can hide decline. When the capacity to extract or produce grows, stable output can conceal the decline of the resource it draws on. Effort-adjusted measures, and people whose capability has not grown, reveal the decline earlier. - Evidence: pp. 17–19, 21, 26. - Strength: strong for herring (Bertram’s quantitative per-boat data against official reassurance, followed by the loss of the inner shoals). Moderate–strong for sardine (the “classic signs” are reported via one secondary source without figures, and the causes of the collapse are contested; see bias check). Weaker for cod: the chapter itself reports that the inshore–offshore link is “now open to question” (p. 22). The diagnosis of Huxley’s error is the author’s inference (“presumably”).
2. Hindsight-only estimates mislead when it matters. Estimation methods that converge only with hindsight give false reassurance precisely when conditions deteriorate fastest. Such biases can be one-directional and can recur long after first being identified. - Evidence: pp. 21, 24. - Strength: strong for the northern cod case (documented by official and independent reviews). Moderate for generality: in the 1999 North Sea analysis the bias appeared in three of five stocks, not plaice or sole, and the chapter implies rather than shows that it went uncorrected after 1977.
3. Demands for certainty breed false precision. Decision-makers’ demand for certainty, combined with experts’ expectation that any range will be exploited at its permissive end, produces falsely precise, deliberately low estimates. These fail twice: the decision is wrong, and expert credibility is lost. - Evidence: p. 22, consistent with p. 21. - Strength: moderate. It rests on one interview study, reported second-hand and framed by the author as “the views quoted there”.
4. Arbiter institutions resist correction. An institution whose authority rests on being the neutral scientific arbiter will tend to defend its estimates, discredit critics and soften unwelcome reviews. Correction often needs external pressure and fully independent review. - Evidence: p. 21. - Strength: moderate–strong. The sequence is well documented, although some characterisations (“downplayed”, “turned on its head”) are the author’s, and the Alverson paragraph carries no specific citation. One complication: Alverson himself publicly minimised the gap (“rather amazing that we are as close to each other as we are”), so the softening was not only the agency’s doing.
5. Impossible proof standards default to inaction. Proof standards for restriction such as evidence that has “decisively settled” the issue, is “overwhelming”, or will “satisfy everyone”, applied in a highly variable system, make inaction the default. The costs of action are counted and the costs of inaction are not. Reversing the burden of proof is the explicit remedy. - Evidence: pp. 18, 20, 25; p. 23 n. 3. - Strength: strong as a description of the standards applied. Moderate as an explanation of outcomes, because the sardine attribution is contested.
6. Genuine variability is also a shield. Genuine natural variability doubles as a shield for those who benefit from continuing. It must be assessed on its merits, not dismissed and not accepted wholesale. - Evidence: pp. 17, 18, 20. - Strength: moderate. The rhetorical use is documented, but the real environmental contribution is not weighed.
7. Mobile capital can deplete and move on. Actors with short payback periods and mobile capital can rationally deplete a shared resource and move on, exporting the pattern. Place-bound communities and future users bear the costs. - Evidence: pp. 19–20. - Strength: moderate. The sardine case is well sourced; the Scottish case is qualitative.
8. Depletion can absorb capability gains. Investment in extraction capability can be wholly absorbed by depletion and competition, producing the same output with more inputs and lower returns. The benefits that do arrive (e.g. safety) differ from those claimed. - Evidence: p. 19. - Strength: moderate. The figures compare endpoints, with no counterfactual.
9. Political control can flip official findings. Where technical advisers lack independence, a political change can rapidly reverse official risk findings. Where jurisdiction is split, disagreement between levels can leave disputes unresolved, and advisers may retreat into euphemism to avoid trespassing on managers’ territory. - Evidence: pp. 20, 24, 26. - Strength: moderate. The Californian episode is brief and single-sourced, and its link to the collapse is asserted. The EU evidence concerns euphemism and “often unchanged” practice, not a reversal of findings.
10. Precaution through biased models fails. Precaution delivered as numerical limits derived from systematically biased models can fail even when its intent is genuinely conservative. Measures that do not depend on model accuracy (set-asides, closed areas) and multiple knowledge sources add resilience.
    • Evidence: pp. 20, 22, 25–26; refuges pp. 18–19.
    • Strength: strong for the failure mode, though it rests on one well-documented case, and DFO in 2000 still could not say why the stock collapsed (p. 22). Suggestive for the remedies as of 2001.
11. Precautionary language can cover unchanged practice. Institutions can adopt precautionary language while keeping prior practice. Contests over terminology often concern ownership and jurisdiction rather than substance.
    • Evidence: pp. 23–24.
    • Strength: moderate. This is the author’s interpretation, backed by cited criticism from peer agencies.
12. A dominant discipline patches its models. When one discipline and its models dominate, failures are met with added model complexity rather than re-examined assumptions. Signals visible to neighbouring fields and practitioners are discounted.
    • Evidence: pp. 21–23, 25.
    • Strength: moderate/suggestive. It rests mostly on critics’ accounts; the other side is not presented.
13. Frontline knowledge sees early but is positional. Frontline knowledge, especially from those without the dominant technology, can detect problems early. But it is positional, and dissent can switch sides as interests shift, so it needs testing rather than deference.
    • Evidence: pp. 17, 20–21.
    • Strength: moderate. The chapter’s own cases run both ways, though it draws only the positive lesson; and for cod, the inshore–offshore link on which the inshore signal depended is itself “now open to question” (p. 22).
14. Unilateral restraint unravels. Unilateral restraint in an open-access system unravels. Durable restraint needs institutions that bind all participants.
    • Evidence: pp. 17–18.
    • Strength: suggestive. There are two brief examples.
15. Depletion traps its own recovery. Deep depletion creates a rebuilding trap: dependants cannot bear the short-term cost of recovery, so decisions split the difference and risk worse outcomes.
    • Evidence: pp. 22, 24, 26.
    • Strength: moderate. The trade-off is well documented, but the counterfactual benefit of the harsher option is not established.
16. What precaution protects is a value choice. The target might be survival of the resource, maximum long-term yield, or dependent systems. There is no single correct precautionary approach, and the objective chosen sets how stringent it is.
    • Evidence: p. 25; p. 178.
    • Strength: strong as a conceptual point (not an empirical claim).
17. Critical knowledge can go unused. Critical knowledge (system structure, known estimation biases) can be established and then left unapplied for decades. Institutions need deliberate memory.
    • Evidence: pp. 18, 22, 24, 26.
    • Strength: moderate/suggestive. There are two instances, but the chapter implies rather than documents that the knowledge was lost or ignored, and it does not analyse why.
18. Systems can fail without villains. Serious failures can arise from competent, well-meaning actors locked into incentive and institutional structures that stop them acting on available information.
    • Evidence: p. 23.
    • Strength: asserted/moderate. This is the author’s judgement about the cod case. It sits somewhat uneasily with his own account of dismissing critics and managing messages in that case (p. 21), although a systems view can include institutional incentives to defend earlier estimates. (The deliberate political override in California, p. 20, is a different case and is not covered by this verdict.)
19. Suspension as a low-cost experiment. Suspending an activity as a reversible experiment can produce evidence cheaply, making precaution a source of knowledge.
    • Evidence: pp. 18–19.
    • Strength: suggestive. There is one proposal and one weakly evidenced outcome.

Limitations, contestation and bias check#

The author’s position - The author is also the report’s Executive editor. His own WWF-UK work supplies the cost-benefit account and the Canadian cost figure (pp. 19, 22; pp. 197–198). - There are no panels or responses from DFO, ICES, the EU, the Californian or federal authorities, or industry. Their views appear only as the author selected them. - The treatment of the EU and ICES is noticeably gentler in tone (“frank”, “constructive”, “concentrating on the future”; p. 24) than that of DFO. This may reflect the report’s EEA setting, and is an observation rather than an accusation. In substance he is still critical: a “precautionary gloss” on an often unchanged regime, the STECF “crisis management” critique that “still appears valid”, and a rebuttal of the Commission’s reasons for avoiding MSY (p. 24).

Secondary sourcing - California rests entirely on McEvoy (1986). DFO’s conduct, the Keats Report and the interviews rest on Finlayson (1994). Huxley is quoted second-hand twice, and the 1376 petition comes via March (1953). - As far as the text shows, nothing is misused. But the key episodes are single-source and not cross-checked.

The sardine attribution (the most important contested point) - The environment appears only as industry’s excuse (p. 20). Table 2.1 says flatly that “continuous inaction leads to the collapse” (p. 27). Yet the chapter opens by stressing that sardine “can show extreme changes over tens of years” naturally (p. 17), and never reconciles the two. - Natural swings (external pointer, to verify in the hindsight strand): palaeo-sediment scale records (e.g. Baumgartner et al. 1992) and work on Pacific climate regimes indicate large natural swings in sardine before industrial fishing. Much later literature treats the collapse as environment and fishing combined. - Dating (external pointer): standard landings histories put the fishery’s collapse in the late 1940s to early 1950s, not 1942. Check the chapter’s “in that year the sardine stock collapsed” (p. 20) against McEvoy. - What survives. If either point holds, the case is weaker as evidence that precaution alone would have averted collapse. It remains strong evidence about how warnings were handled.

The cod counterfactual - The claim. The concluding chapter says the pros and cons of early action were identified beforehand, but only limited action was taken because of uncertainty and cost, and far greater costs followed (p. 176). - Evidence in this chapter that complicates it: - the outcome was “far worse than even the most pessimistic projections”; - DFO in 2000 did not know “why it collapsed” or “why it has failed to recover”; - the inshore–offshore link is “now open to question” (all p. 22). - Factors not discussed, or mentioned only as unknowns: environmental conditions; predation (the chapter lists “what proportion are being taken by predators and other fishing activities” among DFO’s unknowns, p. 22, but does not name seals or weigh predation as a cause); foreign fishing on the Nose and Tail of the Grand Banks outside 200 nm (Day 1995 is cited only for NAFO’s ineffectiveness); and misreporting and discarding. - Later consensus (external pointer): overfishing was the principal cause (e.g. Hutchings and Myers, mid-1990s). This should be checked in the hindsight strand.

Lay knowledge - The lessons credit lay knowledge (p. 26), but the chapter’s own lay testimony points both ways: Californian expansionists (p. 20) and Newfoundland offshore fishers (p. 21) were wrong. - Keats was a scientific re-analysis of DFO’s own data. The reference list names its authors (Keats, Steele and Green) and says it was a report to the Newfoundland Inshore Fisheries Association (p. 28), but the chapter does not say what the authors’ professional standing was (external, to verify: academic scientists). So the case shows lay observation (falling inshore catches) prompting fisher-commissioned scientific dissent. The lay signal and the scientific re-analysis worked together; the case is not evidence for lay knowledge on its own. - The Native American example is hedged and single-source. It includes communities that “may have died out” through overexploitation, which the lesson omits.

Hindsight and selection - Warners. Vindicated warners are praised (“prescient”, “prophetic”). Mistaken warners, and restrictions that proved unnecessary, are largely absent. The one exception is the 1883–85 research finding most complaints against trawling unfounded (p. 18), which the chapter reports fairly. - Huxley. He is called “infamous”, though the author adds “no fool”. External pointer: his 1883 claim was specific to the great sea fisheries under contemporary fishing methods, and he accepted that some fisheries could be depleted. The shorthand “marine fisheries are inexhaustible” compresses this. - Cases. The cases chosen are failures, plus a brief herring recovery and a tentative Georges Bank result. There is no case in which assessment-based management worked, which makes the “stock-model dominance” diagnosis harder to judge. - The 1883–85 inquiry. Its empirical findings are reported but not evaluated. The implied critique is of its burden of proof, not its science. That is a fair distinction, but the chapter leaves it implicit.

Weak evidence upgraded - The Scottish inshore ban’s “success” rests on large initial landings that “suggested” it (p. 19). This becomes the “success of areas protected from trawling” in the lessons (p. 26). - The 1898–1998 Scottish figures compare endpoints only. Quota regimes (external: the Common Fisheries Policy’s conservation regime from 1983), markets, fleet policy and consumer gains are not considered. Even so, they describe a real treadmill.

Factual precision - UNCLOS. “Then novel UN Convention on the Law of the Sea” for Canada’s late-1970s extension (p. 20) is anachronistic. UNCLOS was concluded in 1982, as the chapter’s own reference list shows. Canada’s 1977 zone drew on the emerging consensus in the UNCLOS III negotiations (external, well established). - An uncited quote. The “prominent ecologists” quotation (p. 25) has no source. - An unlisted source. The Alverson Report (1988), quoted several times on p. 21, is not in the reference list. - Table 2.1. It departs from the text in five places: three sharpen the narrative of unheeded warnings, one overstates progress, and one is more sceptical of Canada’s precautionary intent than the text. - The Alverson Report’s date of 1988 may be off by a year (external, to verify). - “Several decades later.” The chapter places Huxley’s 1881 and 1883 statements “several decades” after Bertram (1865), but the gap is 16–18 years; only the 1893 committee is nearly three decades later (p. 18). - Reference-list details (external, to verify): Thompson 1919 is listed as “Fisheries Bulletin (Canada)”, though the paper concerns the California Fish and Game Commission. The DFO 2000 entry gives report number A2–10 but a URL pointing to a 2001 “a3-01” file. Neither affects the argument.

Advocacy - The advocacy in sections 2.6–2.7 is mostly hedged, which is to the author’s credit. - The economic claim (“greatly increase the economic returns”, p. 26) rests on one citation, with no data in the chapter. - The claim that precaution is “generally not something that should need to be laboured” in the industry (p. 26) is qualified by the author himself (“some interest groups” will pursue short-term strategies). Even so, it sits somewhat awkwardly with the chapter’s evidence of repeated industry resistance (Yarmouth, California, Newfoundland offshore fishers). - The chapter never asks whether some of the restrictions it favours might be costly without benefit, or what problems a reversed burden of proof creates (which baseline, which objective, who decides). It does recognise that objectives are value-laden (p. 25).

Fairness in the other direction. The chapter is more balanced than much advocacy on this topic. It: - credits technology’s benefits (safety, capital; p. 19); - reports that the 1883–85 research found most complaints unfounded, and that Bertram valued trawling “used wisely” (p. 18); - refuses to blame individuals (p. 23); - openly questions precautionary approaches based on calculated limits, including the one then enshrined internationally (pp. 22, 25); - concedes the inshore–offshore uncertainty (p. 22); - treats the approach/principle distinction as partly substantive (p. 23).

Notable quotes#

1. “I have always been slow to believe in the inexhaustibility of the shoals” (James Bertram 1865; p. 17)
2. “Presumably landings were equated with stock size” (MacGarvin; p. 18)
3. No prohibition unless experimental evidence “decisively settled” the issue (1883–85 inquiry, as paraphrased by MacGarvin, who adds that the reason was the loss of valuable catches; p. 18)
4. “to us conservation means wise use. We do not believe in hoarding our fisheries resources” (US Bureau of Fisheries, via McEvoy; p. 20)
5. “the method instils a false sense of assurance” (MacGarvin on Alverson; p. 21)
6. “biased pseudoscience written to support a political agenda” (DFO on Keats, via Finlayson; p. 21)
7. Reading data to confirm “the model of growth upon which our hearts had been set” (Harris 1990; p. 21)
8. “It was a systems failure” (MacGarvin; p. 23)
9. “the language has been given a precautionary gloss” (MacGarvin on EU/ICES; p. 24)
10. “these are not acceptable excuses to delay implementing an ecosystem-based management strategy” (EPAP 1998; p. 25)

Open questions#

1. The sardine collapse. What were the relative roles of environment and fishing, and when did it actually happen: 1942, or the late 1940s to early 1950s? Check McEvoy directly, plus the palaeo and landings literature.
2. The 1990 cod decision. Would a 125,000 t limit in 1990 have averted collapse? What do later analyses say about fishing versus environment, seals, foreign fishing and misreporting?
3. Northern cod after 2001 (external pointer, to verify: the moratorium was reportedly lifted in 2024 amid dispute over a changed assessment model). Has the old contest over what the models say returned?
4. The EU after 2001. Did it adopt MSY-based limits (external pointer: the 2013 CFP reform set an MSY objective)? Did North Sea cod recover (external pointer: partial recovery, then renewed decline in the late 2010s)?
5. Retrospective bias. Did it become a routine diagnostic (external pointer: “Mohn’s rho”, 1999), and was it corrected in practice?
6. The “second generation” approach. Did no-take zones, the Georges Bank closures, ecosystem-based management and the inclusion of fishers’ knowledge deliver?
7. “Fishing down the food web”. Has it held up? (External pointer: later analyses contested it using survey data.)
8. Pacific sardine after 2001 (external pointer, to verify: collapsed again, with directed fishing closed from about 2015 despite an environment-linked harvest rule). This would test whether better models are enough.
9. Herring versus cod. Why did North Sea herring recover after moratoria while northern cod did not? The chapter does not ask, yet it matters for the irreversibility lesson.
10. The Scottish inshore trawling ban. When exactly was it imposed, and what was its measured effect?
11. The author’s affiliations. Confirm MacGarvin’s other NGO and industry roles, and whether the WWF-UK cost figures were independently reviewed.
12. The uncited ecologists’ quote (p. 25). What is its source? Check Peters (1991), A critique for ecology, cited on p. 23, first.
13. “Replaced” versus “dismissed”. Chapter 2 says experts were “replaced” (p. 20); the concluding chapter says they were “dismissed” (p. 179). Which matches McEvoy?
14. The Keats authors. The reference list names Keats, Steele and Green (p. 28). What was their professional standing, and does it reframe the lay-knowledge lesson?
15. The Alverson Report. Locate the original (not in the reference list) to check the body/executive-summary contrast and its date (1987 or 1988).

Audit log#

Independent audit against the full text extract (PDF pp. 17–30), with PDF checks of pp. 1, 2, 6, 27, 168–182 and 197–198.

1. Quote error fixed (2.2.2; Notable quote 3): “because of the loss of valuable catches” is the author’s paraphrase, not a quotation from the 1883–85 inquiry. Only “decisively settled” is quoted in the source.
2. Misreading fixed (Mechanism 14): the CAD 26m / 66.6m / ~1,000 jobs were losses from the cut to 190,000 t that was actually made, not costs avoided by rejecting 125,000 t. Clarified in 2.4.2 too.
3. Mechanism 8 corrected: market turmoil was the main restraint on herring only in the first half of the 20th century; the 1970s moratorium and mid-1990s emergency measures were regulatory.
4. ICES “needlessly restricted interpretation” now hedged as in the source (“ICES may concur”), and identified as coming from the ICES-published CWP working-group report that also contains Garcia 2000. The omitted “obliquely … requires discussion with fisheries management agencies” remark and the “split responsibility” point are restored.
5. North Sea retrospective bias corrected: it held for cod, haddock and whiting only, with no correlation for plaice and a possible negative one for sole. “Recurred across stocks” is tempered in Mechanism 2, Insight 2, Timeline D and the digest.
6. “North Sea cod can tolerate more fishing” corrected to “Northeast Atlantic cod”, as in the source. Added the point that northern cod had also survived heavier fishing before its crash.
7. Alverson Report: restored the body’s finding that the stock had grown since 1977 (“probably only very slowly” after 1982); “in official eyes”; the sensitivity analysis; “albeit at the lower end”; Alverson’s own public remark minimising the gap; the internal DFO reappraisal. Noted that the paragraph carries no specific citation.
8. Harris Report: added “probably” and “prior to 1989” to the fishing-mortality finding; restored Harris’s cautious endorsement of DFO’s 1989 reassessment; noted the author’s “(and often still are)” insertion.
9. Table 2.1: added a fifth discrepancy. The late-1970s row (“claiming it is doing so precautiously”) is more sceptical than the text and cuts against the chapter’s argument. The authors-section and digest descriptions of the table are adjusted.
10. Herring (2.2.1): restored the restriction of the industrial herring fishery on reopening, the human-consumption pressure behind the mid-1990s measures, and “learned enough … to stave off total collapse”.
11. Huxley: restored “taking an average of years”. Flagged that the chapter’s “several decades later” is really 16–18 years for Huxley.
12. Scottish ban critique softened: the earlier claim that first catches are high “whatever the policy” misframed the evidence. The weakness is that there are no figures or comparison, and one-off accumulation does not show lasting benefit.
13. 2.2.3: restored the hedges “in the good years”, “may also have” and “probably the majority”; herring added to the early industrial species; the author “surmises” an impact on other species (it was stated as simply unknown).
14. Sardine: restored “While there were uncertainties”; “catch was raised” corrected to “agreed to increase”; the federal Bureau’s concession that the evidence of depletion was strong is added to the timeline and digest; p. 180 reuse added.
15. Cod set-up: added that catches fell “well below” ICNAF-authorised levels after 1968, that offshore Canadian trawler catches rose, and the author’s “many would even now consider as precautionary”.
16. Inshore fishers: added “the only dissenting group” and “generally not upgraded”.
17. 2.4.2: the moratorium was “initially for two years”; the components of the cost figure; “and other fishing activities”; “this most studied stock”. “The stock turns out to be several local populations” softened to “growing awareness … more or less discrete local populations”.
18. 2.4.3: flagged the author’s “the views quoted there” framing; “apparently took seriously”; “point estimates” corrected to “estimates”; the Ptolemaic analogy is cited via Finlayson (ibid., p. 69), so second-hand; the “systems failure” sentence is completed.
19. 2.5: restored “increasing prominence in other fields”; “unjustifiably … substantially reduce or even halt”; the Code’s negotiators did not draw the principle/approach distinction; the Code’s broad scope versus its emphasis on stock models; the MSY-overestimate rationale is identified as the author’s gloss; “possible contradiction” (singular, hedged); UN/FAO emphasis on maximising long-term yields; footnote 3’s “of interest beyond fisheries”; B_pa “set inconsistently”; “often remained unchanged”; the rebuttal framed as “a likely response”; “potential to become a highly contentious issue”.
20. 2.6: Gomes “suggests” / “may work as” (it was “found”); the “know enough” and “burden of proof is on fisheries” claims attributed to the US approach and given their full wording; the attribution of “stakeholder involvement is essential” flagged as ambiguous; Daan 1998 added; Peters 1991 pointer for the uncited quote.
21. 2.7: restored the author’s qualifier (“some interest groups”, “generally”) on precaution not needing to be “laboured”, plus “greater certainty” and “in many cases”. The Advocacy critique is softened to match.
22. Lessons 3, 11 and 13: the omitted case attributions are restored (including “reluctance to address fundamental economic issues”).
23. Author standpoint: “none of it is peer-reviewed” corrected to “none in a peer-reviewed journal” (2001b is an edited-book chapter); “from outside the field” softened (he is a fisheries consultant); “approvingly” on Greenpeace removed; tone and substance distinguished for EU/ICES; Ch. 16 is credited to the “Editorial team”.
24. Reuse list: added pp. 171, 172, 175 and 182, and the p. 176 “benefits of recovery even without collapse” point. The page range is corrected to 171–172, 175–182.
25. Mechanism 1: the claim that technology was “underestimated” in the cod case is not in the text. It is now attributed to herring (p. 18) and lesson 8 (p. 26), with the inshore–offshore caveat.
26. Mechanism 5: removed Bertram from “lay knowledge” (the chapter does not present him so); added Newfoundland offshore fishers as a wrong-testimony case.
27. Mechanism 9 / Insight 9: “the federal–state split let the permissive view prevail” corrected. The dispute stayed “unresolved” until the state replaced its own experts. The generalisation is softened.
28. Mechanisms 12, 13, 17: “overconfidence” → the source’s “optimism”; “forgotten and rediscovered” / “lost and regained” (not stated in the source) → “known for herring, only now applied to cod”; “went unheeded” hedged.
29. Insights 1, 4, 10, 13, 18: strength notes revised. Insight 1 is weaker for cod given the inshore–offshore caveat. Insight 4 adds Alverson’s own public minimising. Insight 10 is a single case with cause still unknown. Insight 18’s tension is re-anchored to the cod case (p. 21), not California (p. 20).
30. Limitations: “mistaken warners … absent” corrected to “largely absent” (the 1883–85 findings are the exception); the lay-knowledge critique softened; the p. 176 counterfactual restated accurately; EU gentleness balanced with substantive criticisms.
31. Timelines: the “decisive” quote corrected to “decisively settled”; date provenance flagged (1883–85, 1977, undated federal view); Harris “probably”; the herring “lag” reframed as repetition across different fisheries; North Sea “without correction” marked as implied.
32. Factual precision: added the “several decades” imprecision and two minor reference-list inconsistencies (to verify).
33. Digest: synced with items 2, 4–8, 9, 14, 21 and 29 (1883–85 findings, the Bureau’s concession, the war-effort increase, Alverson, Harris “probably”, “often unchanged”, 3 of 5 stocks, lessons 8 and 13, the Insight 1/2/8 ratings, five Table 2.1 discrepancies, the lay-knowledge caveat).
34. Scope check: no references to contemporary technologies or companies outside the source were found. Starkist and Van Camp appear in the source (p. 20).
35. Consistency sweep: “unchanged regime/practice” → “often unchanged” (Conclusions B, Mechanism 9); the 1977–1999 bias conclusion limited to 3 of 5 stocks; Insight 3 “point estimates” → “deliberately low estimates”; Insight 5 proof-standard quotes made verbatim (“decisively settled”, “overwhelming”, “satisfy everyone”); Insight 17 reframed from “lost and rediscovered” (not in source) to “left unapplied” and downgraded to moderate/suggestive; “In every case” → “In each of the three main cases”.
36. 2.2.2 wording: the McIntosh findings are restated accurately (“of the commercial species only herring spawn lay on the seabed”; “undamaged for sale”), and quote marks removed from “questionable”, which the chapter does not quote.
37. Digest caveat: “edits the whole report” corrected to “Executive editor, part of the editorial team credited with the concluding chapter”.

Second audit pass#

A second independent audit against the full text extract (PDF pp. 17–30), with PDF text checks of pp. 1, 2, 6, 168, 171–172, 175–182 and 197–198. All concluding-chapter page references and the author’s credentials were confirmed.

38. Alverson executive summary: what was “turned on its head” was the sensitivity demonstration (the width of the plausible fishing-mortality range), not the body’s finding as a whole. Corrected in 2.4.1 and in the digest, which had said it inverted the “considerably in excess” finding.
39. Alverson Report is absent from the reference list (pp. 27–30). Now noted in 2.4.1, Evidence base, Factual precision and the digest, with a new Open question 15.
40. “DFO controlled the release” corrected to the source’s “had responsibility for the management of the release”.
41. Keats Report: the earlier claim that the chapter does not say who its authors were was wrong. The reference list names Keats, Steele and Green, as a report to the Newfoundland Inshore Fisheries Association, cited via Finlayson (p. 28). Fixed in 2.4.1, Lay knowledge and Open question 14. Also restored “consistent and severe” underestimation.
42. Native North Americans: “kept catches sustainable” corrected to the source’s “maintained substantial catches for centuries”. Added the section’s opening framing (“complex”; “nothing new”).
43. Heincke: restored “contrary to general belief” and the shift of fishing offshore.
44. McIntosh: restored “insignificant” for trawling’s immature catch.
45. 2.2.3: restored “fluctuated widely”; the 0.1% figure is specified as average net profit on capital for the Scottish demersal fleet.
46. Cod set-up: the gloss “quotas no longer binding: a sign of depletion” is now marked as my reading. The NAFO “ineffective” judgement is attributed to the author (citing Day 1995), not to “many Canadians”. Restored the “dramatic” 1960s growth in fishing intensity.
47. DFO positives: “fishers’ knowledge is being incorporated” softened to the source’s “attempts are being made”.
48. 2.4.3: restored “still advocated” for the model-based precautionary approaches the cod case challenges.
49. UN Agreement: the source does not spell out the “possible contradiction”. The clauses are now quoted and the interpretation is marked as mine.
50. 2.5: added that US federal guidelines recognise the UN Agreement and FAO Code (p. 23), not only footnote 3.
51. Table 2.1: “four in a pro-precaution direction” replaced with a more precise account. Three rows sharpen the unheeded-warnings narrative, one overstates progress, and one is more sceptical of Canada’s intent. Noted that the 1988 row omits the softened summary. Synced in the authors section, Factual precision and the digest.
52. Lessons classification: lessons 1, 9 and 11 were unclassified; they are now listed as conceptual or procedural.
53. Mechanism 1: the landings-as-abundance diagnosis is marked as the author’s inference (“presumably”). Mechanism 4: “unattainable” softened to “near-unattainable”, with the general mechanism marked as my reading. Mechanism 6: “binding institutions made the difference” marked as my reading, and the p. 177 echo restated accurately (fishers “prevented … because of a systems failure”, pp. 177–178).
54. Insight 1 (notes and digest): sardine downgraded from strong to moderate–strong (single secondary source, no figures, contested collapse).
55. Cod counterfactual: “seal predation not discussed” corrected. The chapter does list predation as an unknown (p. 22), but it does not name seals or weigh predation as a cause.
56. Arithmetic and labels: sardine lag “15–17” → “roughly 15–18” years; North Sea bias “more than 22 years” → “about 22 years”; “What was known, when” is labelled as my synthesis; CFP 1983 is marked as external.
57. Author standpoint: “a main shaper of the report’s framing” hedged as an inference from his title. “Unusually self-critical by the standards of this report” (a cross-chapter comparison not supported here) changed to “notably self-critical about how precaution has been implemented”.
58. Digest: Huxley’s error is attributed to the author’s “presumably”. Harris’s 190,000 t recommendation now includes the 20% goal and his warning “this may contribute to further decline”.
59. Scope check repeated: no mentions of contemporary technologies or companies outside the source.