LL1-13 — Ch13 Tributyltin (TBT) antifoulants: a tale of ships, snails and imposex#

Late lessons from early warnings: the precautionary principle 1896–2000 (EEA Environmental Issue Report No 22, 2001). Chapter 13, report pp. 135–148 (PDF pp. 135–148; PDF and printed page numbers coincide in this report). Chapter text pp. 135–142; Table 13.1 p. 143; references pp. 143–148.

Reading basis: the whole text extract, read in order to the final page marker (p. 148). Pages 135, 139, 141, 142 and 143 (Table 13.1) were checked visually against the PDF; the extract matches the printed text. The chapter has no boxes, figures or panels. Its only table is Table 13.1 (p. 143), a timeline marked “Source: EEA”, which indicates it was compiled by the report’s editors, not by the chapter authors. Author affiliations are not given in the chapter. They come from the report’s contributors list (pp. 197–198). Where the editors’ synthesis chapter (Ch. 16, “Twelve late lessons”, pp. 168–191; its TBT mentions fall on pp. 170–181) draws on this case, that is recorded separately below and attributed to the editors. (Audit, first pass: page rendering was unavailable, so Table 13.1, the contributors entries, the Ch. 16 cross-references and several key strings were re-checked against the PDF’s text layer. No images, figures or boxes occur on pp. 135–148. Second pass: the whole extract was re-read against the notes, and the contributors entries, all Ch. 16 cross-references and p. 120 were re-extracted from the PDF text layer and re-checked.)

Authors and standpoint#

Authors (p. 135): David Santillo, Paul Johnston and William J. Langston.

Affiliations and roles as the report states them (contributors list, pp. 197–198, not in the chapter itself): - Paul Johnston (Greenpeace Research Laboratories, Department of Biological Sciences, University of Exeter). An aquatic toxicologist who founded the Greenpeace Research Laboratories in 1987 and is still its principal scientist, with more than 15 years of research on marine pollution (p. 197). - David Santillo (Greenpeace Research Laboratories, University of Exeter). A marine and freshwater biologist and senior scientist with the Laboratories. The report says he “represents Greenpeace International at various international conventions addressing marine environmental protection” (p. 198). - William J. (Bill) Langston (Plymouth Marine Laboratory). A research scientist who has worked on TBT behaviour and impacts in the marine environment “since the early 1980s” (p. 197).

What that means for standpoint: - Two of the three authors are scientists employed by an environmental NGO. One of them represented that NGO in the international marine-protection forums, which plausibly include the regional and IMO processes the chapter describes. The chapter does not say whether Greenpeace took part in the TBT negotiations, and it does not disclose the affiliation in the text. The authors are therefore partly insiders to the policy argument they are narrating. - Langston is an insider to the science. He is author or co-author of several studies the chapter cites (Langston et al., 1987; Langston and Burt, 1991; Langston et al., 1994; Langston, 1996; pp. 137, 139, 141). He works in Plymouth, and much of the key early dogwhelk evidence came from Plymouth Sound and south-west England (Blaber, 1970; Bryan et al., 1986; pp. 136–137). The chapter does not give a study site for the dose-response work of Gibbs et al. (1988). The chapter does not give the affiliations of Bryan, Gibbs and colleagues. The chapter also cites Santillo et al. (1998), a paper on the precautionary principle as protection against “failures of scientific method and risk assessment” (reference, p. 147; cited p. 142). That paper is co-authored by two of this chapter’s authors (Santillo and Johnston). - Evident stance: firmly in favour of the global phase-out and sceptical of the claim that the 1980s controls “solved” the problem (p. 138). It treats defenders of continued TBT use as a named, cited opposing camp, chiefly Abel (2000), Abbott et al. (2000) and Evans (2000) (pp. 135, 138, 141). The chapter also relies on Evans (2000) for neutral facts: market share, Arcachon oyster production and the dates of national bans (pp. 136, 138). It cites Evans et al. (2000a) as evidence of recovery (p. 139), and presents Nicholson and Evans (1997) as having “provided further evidence supporting” the offshore imposex findings while disputing their significance (p. 140). So the “camp” is not dismissed wholesale. The tone is measured and the evidence is referenced throughout, but the chapter makes normative arguments in places, especially on alternatives (pp. 141–142). Unusually, it denies precautionary credit even to the actions it supports. Its conclusion is that none of the TBT actions so far, including the IMO treaty it welcomes, were precautionary (p. 142). - Panels/commentaries: none. (Panels are a feature of the 2013 volume.)

Section-by-section notes#

13.1 Introduction (p. 135)#

• The case for TBT, conceded at the outset. Without control, fouling raises drag and fuel use, “with substantial consequences in terms of economics and emissions”. TBT (and the less used triphenyltins) are “extremely effective and relatively economical as antifouling biocides”, and this drove rapid uptake in the 1970s. These two points have been the defence of TBT ever since the first harms appeared in the late 1970s, and proponents still cite them (Evans, 2000; Abel, 2000) (p. 135).
• The authors’ counter-frame: one-sided cost accounting. What is missing from proponents’ evaluations is “proper consideration not only of the quantifiable financial losses suffered by the aquaculture industry and imposed upon harbour authorities” but also of “broader environmental ‘costs’” (p. 135).
• Why TBT is, “in some ways, a rather unusual case study” (p. 135):
• use was recent (late 1960s onward);
• concerns arose while use was at its peak and “led rapidly to some national and regional restrictions”;
• imposex, caused by “interference with steroid hormone metabolism”, is “a highly sensitive, chemical-specific phenomenon” that “contributed greatly to the early acceptance of a direct causal relationship and, in turn, of the need for controls”.
• Industry opposition, the chapter’s only statement on it: “The paint manufacturers, in concert with other interested parties, mounted challenges to restrictions imposed in the 1980s, but the strength of the evidence for severe effects on biota, including regional extermination of some species, could hardly be denied” (p. 135). No details, sources or outcomes are given.
• Arc of the story:
• Bans on vessels under 25 m (France 1982, UK 1987, wider by the late 1980s and early 1990s) improved marinas and harbours, and some regional recovery followed.
• Improved monitoring then revealed wider contamination and population-level effects. Vos et al. (2000) put imposex at up to 150 prosobranch species worldwide.
• Three things drove calls to extend bans to all vessels: the link between imposex and shipping density, poor recovery in places, and butyltins in marine mammals.
• Written “on the brink” of an IMO/MEPC prohibition, the chapter asks “what can we learn from the process of its evolution?” (p. 135).

13.2 The emergence of the TBT problem (pp. 135–136)#

• Technology lineage:
• Organotins were developed as moth-proofing agents (1920s), then used as bactericides and fungicides.
• Dibutyltin and TBT have been produced since the late 1940s.
• TBT entered antifouling paint in the 1960s “as a booster biocide in copper-based formulations” (p. 135).
• Because it worked better than copper, use “accelerated greatly in the 1970s” and took “a major proportion of the antifouling market” (p. 136).
• Product innovation followed: “free association” paints (quick release, frequent repainting) gave way to “self-polishing copolymer” paints (steadier release, less repainting) (p. 136).
• The core misjudgement: “While toxicity to fouling organisms was intentional, its propensity for wider impacts on the marine environment had been grossly underestimated.” “An early focus on acute effects, especially mortality … failed to identify sub-lethal consequences of prolonged exposure in some taxa” (p. 136). In some gastropods, imposex “can be initiated” by TBT in the low ng/l (parts per trillion) range, concentrations also known to cause shell deformity and larval mortality (Bryan et al., 1986; Alzieu, 1998; Alzieu et al., 1986) (p. 136).
• Effect seen long before cause:
• Blaber (1970) recorded a penis in female dogwhelks in Plymouth Sound, “at much greater prevalence close to the harbour than further away”.
• Smith (1971) described imposex in the American mud-snail (Nassarius obsoletus) near US east-coast harbours. The chapter says the phenomenon “was first described by Smith (1971)”, with Blaber’s observations made “Around the same time” (p. 136).
• “Despite the severity of the phenomenon, however, the causal agent remained unknown. Only as analytical capabilities improved in the late 1970s and early 1980s was the connection with shipping made and the extent of damage already done recognised” (p. 136).
• Two regional histories drove the first controls: Arcachon Bay and southern UK dogwhelks (p. 136).

13.3 Arcachon Bay (pp. 136–137)#

• Setting: oyster (Crassostrea gigas) production was 10,000–15,000 t a year until the mid-1970s. Leisure craft rose from 7,500 (mid-1970s) to 15,000 (early 1980s). TBT inputs peaked at about 8 kg a day (Ruiz et al., 1996) (p. 136).
• The first casualty was a pest. Imposex appeared in 1970 in the oyster drill (Ocenebra erinacea) and led “rapidly to its near extirpation from the bay”. TBT was identified as the cause “only in the early 1980s”. The authors’ key counterfactual: “Had the adverse effects been limited to the loss of this species, considered a pest within the shellfish industry for its damage to oyster stocks, little if any action may have followed” (p. 136).
• Collapse of the commercial crop:
• After a normal spawning in 1976 few larvae survived, and settlement largely failed into the 1980s, with “massive financial losses by the shellfish industry”.
• Production fell to 3,000 t by 1981.
• Adults were made unsaleable by shell deformation (“‘ball-shaped’ specimens”) (p. 136).
• Action before detailed exposure data:
• The observations “pre-dated analytical techniques sensitive enough to describe in detail environmental distributions of TBT”. The first reliable water survey was Alzieu et al. (1986), and sediment data came only in the 1990s.
• Heavy financial losses were nonetheless “sufficient to stimulate relatively swift action”. “Acting on the best information available linking the oyster collapse to the presence of TBT”, France became the first country to prohibit “the application of TBT paints to small (< 25 metre) vessels”, in 1982 (p. 136).
• Effects of the ban:
• The controls “undoubtedly markedly reduced TBT inputs from marinas throughout France”.
• Implementation at Arcachon “was probably aided by the local provenance of many boat-owners and their interest in preserving a local industry”.
• But effectiveness against wider contamination “remains questionable” (Michel and Averty, 1999a) (pp. 136–137).
• Note (note-taker’s reading): the French action was taken under uncertainty about exposure. It came “Acting on the best information available” and before reliable environmental measurements existed. But it was a response to severe harm that had already happened: a pest species nearly extirpated and the oyster industry collapsed, with “massive financial losses”. It is therefore consistent with the chapter’s later verdict that no TBT action was precautionary in the sense of acting before harm (p. 142). The tension is only with respect to certainty about cause and exposure. See the limitations section.

13.4 UK harbours and coastal waters (pp. 137–138)#

• Evidence spreads: imposex in N. lapillus was highest near ports but was spreading along southern UK coasts (Bryan et al., 1986, 1987). Waldock and Thain (1983) attributed failed UK Pacific oyster culture “to TBT exposure rather than to fine sediment particles as previously thought” (p. 137).
• 1985 UK controls: restrictions on sale for small vessels, handling guidelines, and an environmental quality target (EQT) of 20 ng/l TBT (8 ng/l as tin), “a level considered at the time to be sufficiently protective of marine biota” (p. 137). Concentrations “in many coastal waters consistently exceeded the EQT”, and “concentrations well below the EQT could cause severe effects” (p. 137). UK reproductive failures and shell deformities “showed many parallels” with Arcachon (p. 137).
• Table 13.1 describes 1985 as “First controls introduced in United Kingdom limiting concentrations of TBT in paints”, and January 1987 as “further restrictions on TBT content”. The text’s list of 1985 measures does not mention a paint-content limit. But its January 1987 wording (“reduced the maximum permissible content”, p. 138) implies that one already existed. So the table fills a gap in the text rather than contradicting it.
• Dose-response: sterilisation of female dogwhelks at “as low as 3–5 ng/l” (Gibbs et al., 1988); almost all females affected at 10 ng/l; testis development and suppressed egg production at higher levels (p. 137). The “protective” target was therefore 4–7 times the lowest reported sterilising concentration (my calculation).
• Field exposures:
• Marina summer levels “regularly exceeded 100 ng/l and occasionally 1 000 ng/l” (Waldock et al., 1987).
• Similar levels were found in Poole Harbour marinas, which served about 5,000 craft (Langston et al., 1987).
• Levels above 100 ng/l were sometimes found even outside marinas.
• Through the 1980s butyltin concentrations showed strong seasonal cycles tied to boating and maintenance.
• The overlap with effect levels “illustrates the inadequacy of the EQT as a protective instrument” (p. 137).
• The pivotal study. Bryan et al. (1986) found imposex “widespread, that all populations are affected to some degree… Populations close to centres of boating and shipping activity show the highest degrees of imposex”. The chapter calls the study “particularly influential”. It also showed coincidence with rising TBT use and correlation with tissue organotin. The study “was profiled” that September in a Marine Pollution Bulletin news item (“TBT linked to dogwhelk decline”, 1986, p. 390). The chapter says this “in many ways” marked “wider acceptance of the central role of TBT in observed declines” (p. 137).
• Wider concerns by the late 1980s:
• “The geographical extent of the TBT problem in Europe had been recognised by the late 1980s” (Bailey and Davies, 1988a, 1989; Gibbs et al., 1991);
• salmon-cage use contaminated a Scottish sea loch, with more imposex in aquaculture lochs (Balls, 1987);
• organotins accumulate in the sea-surface microlayer, a threat to buoyant fish larvae and intertidal organisms (Cleary and Stebbing, 1987; Cleary et al., 1993);
• sediment persistence (Langston et al., 1987; Langston and Burt, 1991) (p. 137).
• UK escalation:
• January 1987: permitted organotin content of paint cut, “helped by technological developments” (Side, 1987).
• May 1987: this was “rapidly superseded” by a total retail sale ban for <25 m vessels and fish cages, where “earlier voluntary measures having failed adequately to address this latter application” (pp. 137–138).
• The Water Research Centre’s EQS work found “that for TBT, in particular, no-effect levels could not be achieved in the environment without use restrictions”.
• “Two years later, and in recognition of the low-dose effects”, the 20 ng/l EQT was replaced by an EQS of 2 ng/l (Cleary, 1991), a tenfold tightening (p. 138). (“Two years later” appears to count from 1987, i.e. about 1989.)
• Regional action (PARCOM):
• PARCOM (June 1987) found TBT causing “serious pollution in the inshore areas of the Convention waters (NE Atlantic)”. Recommendation 87/1 sought a harmonised retail ban for pleasure boats and fish cages, “and, furthermore, for consideration of restrictions for seagoing vessels and underwater structures”.
• “Though strong in intent, it quickly became apparent that PARCOM could not achieve restrictions within the commercial shipping sector”. Because the issue “had already been referred to the International Maritime Organization (IMO)”, PARCOM turned to docking inputs (88/1).
• Effectiveness is “difficult to evaluate”, but shipyards and docks “remain important point sources” (p. 138).

13.5 A global pollutant (p. 138)#

• Knowledge networks: the Oceans symposia of the US Marine Technology Society “did much to disseminate the growing body of research”. NOAA Mussel Watch found butyltins ubiquitous in US bivalves, and New Zealand work confirmed accumulation in sediments (p. 138).
• Diffusion of small-vessel bans: US 1988; Canada, New Zealand and Australia 1989; EU harmonisation 1991 (p. 138).

13.6 Effectiveness of controls on small vessels (pp. 138–139)#

• What worked: retail restrictions “undoubtedly” moved leisure craft away from TBT “(although the extent of illegal application remains unknown)”, and inputs fell substantially (p. 138).
• Monitoring limits:
• “High detection limits of early analytical methods” limited Arcachon monitoring of butyltins. Levels fell from 900 ng/l (as tin) in 1983 to <100 ng/l by 1985, but it took until the late 1980s to confirm they were below 10 ng/l “in most parts of the bay”.
• Hydrodynamic disturbance “largely prevented” reconstruction of sediment trends (Ruiz et al., 1996) (p. 138).
• The units change here. If the 900 ng/l figure is TBT expressed as tin, it corresponds to roughly 2,200 ng/l TBT (my calculation; the chapter’s own conversion is 20 ng/l TBT = 8 ng/l as tin, p. 137). The chapter does not say whether the figure is TBT or total butyltin.
• The contested “solved” narrative: the authors concede that the mid-1980s recovery of the oyster beds and resumption of commercial operations “provided the first indirect evidence of the effectiveness of the ban”. But the recovery is “frequently cited as an illustration of how the limited controls of the 1980s ‘solved’ the TBT problem (Nicholson and Evans, 1997; Evans, 2000), a view which has met substantial disagreement” (p. 138).
• Evidence for partial recovery:
• flame shell and associated flora recovered at Mulroy Bay (Ireland), after being “decimated” by salmon-cage TBT in 1981–85 (Minchin, 1995);
• less imposex at Sullom Voe/Yell Sound in 1995 than 1991, “although overall prevalence remained high” (Harding et al., 1997);
• a more general UK dogwhelk recovery suggested by Evans et al. (2000a) (pp. 138–139).
• Evidence against optimism (p. 139):
• slower-than-expected recovery at “Ballybegs” (a typo for Killybegs per the reference list; Minchin et al., 1997);
• Shetland recovery set against a high imposex background, with dogwhelks still absent from the worst sites;
• Gulf of St Lawrence butyltins “remarkably high eight years after retail restrictions” (St-Jean et al., 1999);
• French dissolved TBT stabilised “at levels often well in excess of those known to cause adverse effects in some species” (Michel and Averty, 1999a);
• parts of Arcachon still able to cause imposex 10 years after 1982 (Ruiz et al., 1996);
• similar UK concerns (Cleary, 1991; Langston et al., 1994).
• Causes of incomplete recovery (these “undoubtedly differ from location to location”, p. 139):
• “isolated but significant illegal use” on small vessels, in some cases (Huet et al., 1996);
• release from contaminated sediments (Waldock et al., 1990; Langston et al., 1994);
• high larval sensitivity and long life histories (Gibbs, 1993; Rees et al., 1999);
• above all, seagoing vessels: “Potentially the most significant single contributor to continued widespread presence of TBT in the marine environment, however, and currently the most fiercely debated” (p. 139).

13.7 The significance of seagoing vessels (pp. 139–140)#

• The exemption rationale: large commercial and military ships were “viewed as a lesser concern because, it was argued, those vessels spent most of their operational lives on the open seas”. The authors reply that they use inshore waters and ports regularly (p. 139).
• Source evidence:
• Hull-cleaning wash water from a naval frigate carried TBT at about “1 million times the lowest biologically effective concentrations”. One cleaning releases about 100 g of free TBT, and “almost 1 kg” including paint chips, which may act as long-term reservoirs (Waldock et al., 1988). Guidelines exist but are “difficult to evaluate”.
• At Sullom Voe, later inputs came “predominantly from movements of oil tankers themselves” (Bailey and Davies, 1988b).
• Heavy traffic is also a source in the Gulf of St Lawrence (p. 139).
• The live dispute: far-field effects.
• Coastal inputs from shipping “are now generally accepted”, but significance for “remote coastal and offshore areas remains the subject of intense debate” (ten Hallers-Tjabbes, 1997; Evans, 2000) (p. 139).
• Evidence:
  • offshore North Sea whelk imposex correlated with shipping intensity (ten Hallers-Tjabbes et al., 1994);
  • similar correlations in the Strait of Malacca and remote Galicia;
  • a speculative link to whelk loss in the Wadden Sea (Cadee et al., 1995);
  • an estimated 68 t a year of TBT into the North Sea from shipping, with “continuous leaching from painted hulls when in port or under way the major contributor” (Davies et al., 1998) (pp. 139–140).
• Counter-view: Nicholson and Evans (1997) “provided further evidence supporting the pioneering work of ten Hallers-Tjabbes et al. (1994) but questioned the significance of what they termed ‘mild’ imposex against a background of overexploitation by shell fisheries” (p. 140). The authors do not directly rebut this confounder. They note only that detection of offshore population-level effects underlay mid-1990s calls for extension (p. 140).
• Global and food-web reach:
• The problem’s “truly global nature” was acknowledged (Ellis and Pattisina, 1990; Kannan et al., 1995a–c). The chapter cautions that “data remain limited for areas outside Europe and North America”. Studies in Japan and the Philippines had “gone some way towards redressing the balance, confirming (as suspected)” similar relationships (p. 140).
• Marine mammals: up to 10 ppm butyltins in porpoise liver, suggesting low metabolic capacity (Iwata et al., 1995). Further data in Pacific, Asian and Mediterranean species. First European cetacean and seal reports include pelagic species feeding offshore (Ariese et al., 1998; Law et al., 1998, 1999) (p. 140).
• Health effects, hedged:
• “Although uncertainty remains, evidence suggests that accumulation of butyltins in top predators might adversely effect the immune system”. Correlations come from dolphin mortality events (Kannan et al., 1997) and diseased sea otters (Kannan et al., 1998) (p. 140).
• Human dietary intake was long recognised (especially from fish farmed in TBT-treated cages) but evaluated only recently. The assessments conflict. Cardwell et al. (1999) put US intakes “significantly below levels judged to be of concern”. Belfroid et al. (2000) found possible exceedance of tolerable intakes “based on more sensitive immunotoxicological end-points” for some products, with data for most countries “simply unavailable” (p. 140).
• The chapter reports both (“In contrast”) without adjudicating. Its wording suggests, though it does not say outright, that the difference turns partly on the choice of endpoint (my inference).

13.8 Progress towards a global phase-out (pp. 140–141)#

• Framing: “Despite uncertainties regarding ‘far-field’ effects, contamination of the marine environment with organotins is clearly a persistent and pervasive problem” (p. 140). The case for action is detached from resolving the far-field dispute.
• Inertia, then reversal:
• 1994: MEPC decided “that no further controls were necessary”.
• 1995: Esbjerg North Sea ministers agreed “to undertake concerted action within IMO aiming at a worldwide phase-out of the use of TBT on all ships”. Like OSPAR, they “recognised the limits of regional measures in addressing a transboundary problem”.
• 1996: MEPC agreed to look again “On the basis of emerging research”.
• 1997: MEPC 40 recognised the possibility of a ban, “a sea change in the thinking of that committee”.
• 1998: draft mandatory regulations (p. 140).
• Deadlines: the draft Convention on the Control of Harmful Anti-fouling Systems sets 2003 for ending application and 2008 for presence on hulls. These were adopted under IMO Resolution A.895(21) (November 1999), with convention adoption “expected in 2001” (pp. 140–141).
• Table 13.1 addition: OSPAR (1998) prioritised organotins, with cessation of all releases to the marine environment targeted for 2020 (p. 143).

13.9 The question of alternatives (p. 141)#

• Context: “With the advent of such a progressive measure” (the authors’ description of the IMO phase-out), attention turns to “effective and economically viable alternatives” (Evans et al., 2000b, titled “The TBT ban — out of the frying pan into the fire?”, p. 145) (p. 141).
• Unacceptable predecessors: “some of the preparations in use prior to the widespread introduction of TBT (including those containing mercury or arsenic compounds) are not acceptable alternatives” (p. 141). Going back is not an option. (The chapter does not say TBT displaced these because they were worse. The only reason it gives for TBT’s market gains is its “superior effectiveness over copper”, p. 136.)
• Regrettable-substitution risk (the note-taker’s term): copper systems “generally require booster biocides”, which “may present additional problems”. Irgarol 1051, “used extensively in some areas as a replacement for TBT on leisure craft”, occurs widely “in certain estuaries” and has “direct effects on plant growth in the field”. These findings “in some ways mirror early findings with respect to TBT” (p. 141).
• Rebuttal: “Some commentators use such examples as justification against substitution of TBT in commercial shipping applications (Abel, 2000; Abbott et al., 2000). This is a somewhat negative argument. The more appropriate question is, can fouling control be achieved without recourse to such highly toxic, persistent and bioaccumulative substances?” (p. 141).
• This reframes the question from choosing a chemical to delivering a function.
• Abbott et al. (2000) is a cost-benefit analysis subtitled “The case for a treatment approach” (p. 143). The chapter does not explain its argument, so readers never see the opposing proposal. Abel is a co-author of Abbott et al. (reference list, p. 143), so the two citations overlap.
• Non-chemical options:
• research on natural antifoulants is “developing rapidly”, though “commercial application may remain some time away”;
• “Perhaps the most promising contemporary solution is biocide-free, ‘non-stick’ coatings which simply present a physical barrier to settlement”. They are established for leisure craft, proven for fast craft, and “under evaluation” for large vessels;
• “Although significant technical issues remain to be resolved”, non-release mechanisms “would seem desirable” (p. 141).

13.10 Late lessons from the TBT story (pp. 141–142)#

• Testing paradigm: “It was initially assumed, for example, based on acute toxicity tests that concentrations in the ug/l (micrograms per litre) range were necessary to initiate biological effects”, hence the 20 ng/l EQT. Even compliance would have left “severe biological effects” (p. 141). The EQT therefore already sat well over an order of magnitude below the assumed µg/l effect range, yet it was still 4–7 times too high (my calculation from pp. 137, 141). A generous margin on the wrong endpoint did not protect.
• Mechanism counterfactual: “Had the endocrine-mediated mechanisms underlying imposex (Matthiessen and Gibbs, 1998) been identified earlier, the potential for low-dose effects would have been clear” (p. 141).
• Hindsight admitted: “much of our knowledge of organotin toxicity has been obtained through hindsight and is likely to continue to develop”. The authors hope “the lessons gained will allow earlier identification and, wherever possible, avoidance of future problems in advance” (p. 141).
• Persistence:
• “Early predictions that TBT would degrade rapidly in surface waters (see reviews by Simmonds, 1986 and Lee et al., 1989) failed to account for the high lipophilicity and sediment-binding properties of organotins in seawater”. Simmonds and Lee et al. are cited as reviews of these predictions, not as their authors. Simmonds (1986) is titled “The case against tributyltin” (p. 147).
• Half-lives are days in eutrophic surface water but “up to several years” in nutrient-poor waters and sediments, “especially anaerobic sediments”.
• Persistence and toxicity to sediment communities raise “the prospect of delayed recovery of damaged ecosystems” and represent “a substantial legacy to be borne by authorities responsible for dredging operations”. This is “only now receiving formal recognition” under the London Convention 1972 and OSPAR 1992 (p. 141).
• Bioaccumulation: biofilm amplification was recognised only recently (Labare et al., 1997); “Accumulation in top predators was simply not envisaged” (p. 141).
• Complexity: “Hindsight is a wonderful thing, of course”. Increasing awareness of marine “pathways and interactions”, “coupled with an appreciation of their complexity and indeterminate nature (Santillo et al., 1998)”, “should equip us to reduce or even avoid the potential for future TBT-style scenarios involving other persistent organic pollutants” (p. 142).
• Two sources of early ignorance: “low sensitivity of analytical methods and absence, in most regions, of adequate baseline data on the distribution and ecology of non-commercial species”.
• The first was “unavoidable”, partly offset by imposex as a specific biomarker.
• The second “could have been avoided”. South-west England’s baselines were “invaluable”.
• “Though frequently underrated, baseline studies play a vital role in the early detection of adverse trends and may consequently serve the application of precaution”. Long-term monitoring of recovery is “Equally important” (p. 142).
• Transboundary: contamination continued in Japanese coastal waters despite a national ban on all TBT antifouling uses, which “clearly illustrates the transboundary nature of the problem”. Hence, hedged: “It would seem that universal, global restrictions are the only way to address the totality of the TBT problem and to achieve environmental quality standards in all coastal systems” (p. 142).
• Alternatives:
• “though the selection of alternatives to hazardous chemicals is never a simple task, this does not justify inaction”. Some alternatives have problems, “though thankfully, on a global scale, none has yet proved as damaging as TBT”.
• Non-release options “should, perhaps, be viewed most favourably”.
• “Broader consideration of problems may give rise to more beneficial solutions than simple ‘chemical for chemical’ substitution” (p. 142).

13.11 Conclusions: precaution or retrospective action? (p. 142)#

• Evidence: the use of organotin antifoulants “has and continues to result in widespread and sometimes severe environmental effects”. “Unusually in the field of ecotoxicology, the evidence linking cause and effect (with respect to TBT-induced imposex) is irrefutable”. Vos et al. (2000) call it “the best example of endocrine disruption in invertebrates that is causally linked to an environmental pollutant” (p. 142).
• Verdict:
• “It would be difficult to argue, therefore, that any of the actions to address TBT to date have been precautionary, resulting as they have from extensive documentation of ecological impacts. Actions have undoubtedly contributed towards remediating the most severe problems, but this is not precaution” (p. 142).
• The 1987 UK restrictions followed realisation that concentrations were “already well above those known to cause chronic effects” and had already caused widespread gastropod declines.
• On the IMO convention the authors hedge (“It could be argued, similarly”). It “has come only after the consequences of continued use have been well documented”. They concede that “the significance of this global, legally binding treaty should not be underestimated”, but conclude “it too is fundamentally retrospective in action” (p. 142).
• The logic runs from the strength of the evidence to the verdict. Because the causal evidence is “irrefutable”, and action came only after “extensive documentation of ecological impacts”, the actions were “therefore” not precautionary (p. 142). Note that “irrefutable” describes the state of evidence in 2001. The chapter’s actual criterion is that each action followed documented harm, not that the evidence was already conclusive when each action was taken (France acted in 1982 before detailed exposure data existed, p. 136). Note-taker’s gloss: where the causal evidence is itself built from observed harm, as with imposex, the strength of that evidence is also a record of how much damage had already occurred.
• Outlook:
• The 2008 phase-out “will mark the closure of an important chapter”. Sediment and biota persistence remain, “but widespread inputs from shipping should, at least, be a thing of the past”.
• “Organotin inputs will continue, of course, through their use as additives in a wide range of consumer products. Whether efforts to address these emerging challenges will draw on the lessons of the past remains to be seen” (p. 142).

Table 13.1 “TBT: early warnings and actions” (p. 143; Source: EEA)#

Checked visually against the PDF. The table runs from the early 1970s to 2001; its entries are folded into the case timeline below. Points to note: - It is marked “Source: EEA”, which indicates it was compiled by the editors, not the authors. - It omits MEPC’s 1994 “no further controls” decision (p. 140) and the UK EQS revision to 2 ng/l (p. 138). - For 1985 it records a paint-content limit that the text’s list of 1985 measures omits, although the text’s January 1987 entry implies such a limit (see 13.4). - For 1982 it says France prohibited “the use” of TBT paints on small vessels; the text says “application” (p. 136). - It adds OSPAR’s 2020 cessation target (1998 entry).

How the report’s editors use this case (Ch. 16 “Twelve late lessons”, editorial team; outside this section, recorded for cross-reference only)#

• p. 170: top-predator accumulation of organotins “was simply not envisaged”, given alongside DES as an example of surprise under the editors’ first lesson, on responding to ignorance (16.2.1). The preceding sentence, about chemicals that were “relatively inert and benign under ‘normal’ conditions” but behaved differently “under conditions that were not considered in the risk appraisal”, refers to CFCs; it should not be read as the editors’ description of TBT, which was intentionally toxic.
• p. 171: the context is the editors’ claim that, once irreversibility became apparent, “regulators were often slow to react”. “It was relatively quickly established that TBT was more persistent than first assumed”. Paired with asbestos: “In neither case did hazard reduction actions take account of the long-tem [sic] effect early enough”. The editors propose persistence and bioaccumulation as screening criteria.
• Also p. 171: TBT is listed among cases where readily identified acute effects preceded less obvious chronic ones. The editors caution that this is not a general rule.
• p. 172: quotes the baseline-studies conclusion.
• p. 173: “In the case of TBT, rates of degradation were based on assumptions about the nature of the marine environment that were evidently incorrect for many areas”.
• p. 177: quotes the “chemical for chemical” line under “Evaluate alternatives and promote robust, diverse and adaptable solutions”.
• pp. 180–181: “global problems clearly require a global response”, illustrated by TBT and the IMO.
• p. 120 (Ch. 11, MTBE, a different author team): TBT is listed with CFCs and PCBs as persistent chemicals that “caused unwelcome ‘surprises’”.
• p. 168: the editors’ four questions to authors were early warning, actions or inactions, costs and benefits, and lessons. For the second, “The key point here is the length of the gap between the specific problem being identified and effective action being taken”. Costs and benefits “proved to be the most difficult question for the case study authors to answer”. This chapter fits that pattern.
• No other TBT mentions occur elsewhere in Ch. 16 (to p. 191) or in Ch. 17 (checked against the PDF text layer).

Case timeline#

Lags (my calculation from the chapter’s dates): - First field observation of imposex (1970–71) to identification of TBT as the cause (the connection with shipping was made in “the late 1970s and early 1980s”; at Arcachon “only in the early 1980s”): roughly a decade, 8–12 years. The chapter attributes this to analytical limits: “Only as analytical capabilities improved … was the connection with shipping made” (p. 136). It attributes early ignorance of the geographical extent of the problem to analytical limits plus absent baselines (p. 142). It does not suggest suppression. - Onset of Arcachon oyster failure (1976) to French ban (1982): about 6 years. From causal identification (early 1980s) it was about 1–2 years, which the chapter calls “relatively swift” (p. 136). - Bryan et al. (1986) to UK retail ban (May 1987): about 1 year. From Blaber (1970) it was 17 years. - Recognition that commercial shipping was a significant source (referral to IMO before mid-1987; Waldock et al. 1988; Bailey and Davies 1988b) to IMO’s 1999 resolution: about 12 years. To the planned end of TBT on hulls (2008): about 20 years. The IMO explicitly decided against further controls in 1994 (p. 140). - First observed biological effect (1970) to planned global elimination on hulls (2008): about 38 years. - The chapter’s own “late lesson” is not the length of any one lag. Its point is that even the fast actions came only after harm was “extensively” documented (p. 142).

Harms and costs described: - Collapse of Arcachon oyster production from 10–15 kt to 3 kt/yr, and “massive financial losses” (p. 136). Adult oysters made unsaleable by shell deformation (p. 136). - Financial losses “imposed upon harbour authorities” (p. 135; not quantified). - Failed UK Pacific oyster ventures (p. 137). - Near extirpation of the oyster drill at Arcachon (p. 136). - Regional extermination of some species (p. 135) and absence of dogwhelks from the worst-affected sites (p. 139). - Flame-shell beds decimated at Mulroy Bay (p. 138). - Imposex documented in the wild in “as many as 150” prosobranch snail species worldwide (Vos et al., 2000; p. 135). - Sediment legacy for dredging authorities (p. 141). - Possible immune effects in top predators (p. 140). - Possible exceedance of tolerable human intakes for some seafood (p. 140). - No aggregate monetary estimate is given for either costs or benefits.

The authors’ own lessons and conclusions#

Lessons drawn from their evidence (analytical): 1. The hazard was underestimated because testing focused on acute mortality. Acute tests led to the assumption that effects needed µg/l concentrations, when effects began at low ng/l. The protective target set on that basis (EQT 20 ng/l) was inadequate even where it was met (pp. 136, 141). Had the endocrine mechanism been identified earlier, “the potential for low-dose effects would have been clear” (p. 141). 2. Persistence was underestimated. Degradation estimates from eutrophic surface waters ignored lipophilicity and sediment binding, and half-lives in sediments and nutrient-poor waters reach years. The result is delayed recovery and a dredging legacy (p. 141). 3. Bioaccumulation was underestimated. Biofilm amplification was recognised only recently; accumulation in top predators was “simply not envisaged” (p. 141). 4. Early ignorance of the problem’s geographical extent had two causes: low analytical sensitivity, which was unavoidable and partly offset by imposex as a specific biomarker, and absent baselines for non-commercial species, which was avoidable (p. 142). 5. Transboundary persistence. Japan’s continuing contamination despite a national ban shows national measures are insufficient (p. 142). 6. None of the TBT actions were precautionary. All were responses to extensively documented harm, including the forthcoming IMO convention, which is “fundamentally retrospective” (p. 142).

Recommendations and advocacy (normative): 1. Invest in baseline studies and long-term monitoring, including of non-commercial species. Baseline studies “may consequently serve the application of precaution” (p. 142). 2. Pursue “universal, global restrictions”, which “It would seem” are “the only way to address the totality of the TBT problem” (p. 142). 3. Do not let imperfect alternatives justify inaction. Ask whether the function can be delivered without toxic, persistent, bioaccumulative substances (pp. 141–142). 4. Prefer alternatives that do not release hazardous substances, such as biocide-free non-stick coatings, and consider problems broadly rather than doing “chemical for chemical” substitution (pp. 141–142). 5. Use awareness of marine complexity and “indeterminate nature” to avoid “future TBT-style scenarios involving other persistent organic pollutants” (p. 142). 6. Apply these lessons to organotins in consumer products, an “emerging” challenge (p. 142). This is implied, not stated as a recommendation: “Whether efforts to address these emerging challenges will draw on the lessons of the past remains to be seen.”

Mechanisms and dynamics#

1. How the warnings arose: the effect came first, the cause later. - The earliest warnings were field observations of a strange biological effect (imposex) with a spatial gradient toward harbours (Blaber, 1970; Smith, 1971). They had no identified cause (p. 136). - The causal link needed analytical chemistry that did not yet exist (p. 136). Harm was therefore visible long before it was attributable. - What made attribution possible, and then quickly accepted, was the specificity of the biological response: imposex is “chemical-specific” (p. 135) and works as a biomarker (p. 142). - The chapter’s account implies that the kind of evidence that eventually settled the case, an unusually specific marker plus dose-response data, is the exception in ecotoxicology (“Unusually in the field of ecotoxicology”, p. 142).

2. Commercial salience decided whether the first warning counted. - The first casualty at Arcachon was a pest species, and the authors believe its loss alone would have prompted “little if any action” (p. 136). The first, fastest action came when the harm struck a valued economic activity (oyster culture) and imposed “massive financial losses” (p. 136). - In the UK too, failed oyster culture was part of the early evidence (p. 137). - Later action rested heavily on evidence from non-commercial species. The 1987 UK restrictions followed “widespread declines in gastropod populations” (p. 142), documented mainly in dogwhelks (p. 137). The push for a global ban drew on offshore whelk imposex and residues in marine mammals (pp. 135, 139–140). So the chapter does not support a claim that harm to non-commercial species was simply tolerated. - Where baseline data existed (south-west England), they were “invaluable in the early detection of effects on non-commercial species”. Their absence “in most regions” delayed recognition of the problem’s geographical extent (p. 142). - My synthesis: economic loss triggered the first and fastest response. Ecological harm to non-commercial species drove later action, but only once it had accumulated into documented population declines.

3. Interest alignment at local scale eased regulation. - At Arcachon, boat owners who were local and had an interest in preserving the local oyster industry “probably” aided implementation of the small-boat ban (p. 137). The polluters and the harmed community overlapped. - For commercial shipping, polluters (global fleets) and harmed parties (coastal ecosystems and fisheries everywhere) do not overlap. That is where action stalled (pp. 138–140). This contrast is my inference; the chapter does not draw it.

4. Exposure assumptions and regulatory boundaries. - The first regulations drew a line at 25 m vessel length. Large commercial and military ships were exempted on the argument that they spend most of their time on the open sea (p. 139). - That assumption ignored port time, hull cleaning (up to ~1 kg TBT per operation, p. 139) and continuous leaching in port and under way (68 t/yr into the North Sea, pp. 139–140). - The boundary also matched what was politically and jurisdictionally achievable. Regional bodies could regulate retail sale for leisure craft, but “could not achieve restrictions within the commercial shipping sector” (p. 138). - The regulated segment was the easier one, not necessarily the larger source. (That tractability drove the choice is my inference, supported by p. 138.)

5. Jurisdiction mismatch and the need for a global institution. - A transboundary, mobile source needs a regulator whose reach matches the problem: IMO/MEPC (pp. 138, 140, 142). - Regional bodies (PARCOM/OSPAR, North Sea ministers) could regulate retail sale for small craft. But on seagoing ships they “could not achieve restrictions” (p. 138) and could only press the global body. Their route there was a political declaration (Esbjerg 1995) (p. 140). - The global body first decided that no further controls were needed (1994), then changed position within three years, “a sea change in the thinking of that committee” (p. 140). - The chapter explicitly credits the 1996 re-examination to “emerging research” (p. 140). The role of regional pressure is implied only by the sequence it narrates (Esbjerg 1995 “in spite of” the 1994 decision). It does not describe the internal politics of MEPC: flag states, shipping associations, or paint-industry lobbying.

6. Standards that move as knowledge grows. - The EQT of 20 ng/l was “considered at the time to be sufficiently protective” (p. 137). It was replaced by an EQS of 2 ng/l within about four years (p. 138). - The EQS work itself showed that “no-effect levels could not be achieved in the environment without use restrictions” (p. 138). The management approach shifted from setting environmental concentration targets to restricting use.

7. Burden and standard of proof. - The chapter implies a high de facto standard of proof. Action followed “extensive documentation of ecological impacts” (p. 142). - For offshore effects, the debate continued despite correlational evidence across several regions (pp. 139–140). Sceptics, notably Nicholson and Evans (1997), who themselves found offshore imposex, offered overexploitation as an alternative explanation and disputed the ecological significance of “mild” imposex (p. 140). - The authors’ move is to decouple the policy case from resolving the far-field dispute: “Despite uncertainties regarding ‘far-field’ effects, contamination … is clearly a persistent and pervasive problem” (p. 140). The chapter does not explicitly discuss who carried the burden of proof or how regulators weighed the evidence.

8. The “problem solved” narrative. - Partial recovery (the Arcachon oysters) became a rhetorical resource for those opposing extension of the bans: the 1980s controls “‘solved’ the TBT problem” (p. 138). - The authors counter with evidence of plateaued concentrations, slow recovery and sediment reservoirs (pp. 138–139). - Mechanism: an intervention that fixes the most visible, commercially salient symptom can be read as fixing the whole problem, reducing pressure to address remaining sources.

9. Measurement capacity shapes what is known. - High detection limits limited monitoring. It took until the late 1980s to confirm that Arcachon concentrations were below 10 ng/l (p. 138), and sediment trends could not be reconstructed (p. 138). - Early ignorance of geographic extent was partly “an unavoidable process of methods development” (p. 142). - Knowledge dissemination mattered for the speed of acceptance (pp. 137, 138): networks (the Oceans symposia, NOAA Mussel Watch) and a journal news profile that “in many ways” marked wider acceptance.

10. Complexity and unanticipated pathways. - The chapter explicitly says several routes were underestimated or unforeseen: - biofilm amplification, “only recently been recognised” (p. 141) - sediment binding and anaerobic persistence, which degradation predictions “failed to account for” (p. 141) - accumulation in top predators, “simply not envisaged” (p. 141) - Other routes emerged as “additional concern[s]” once TBT was in wide use. An example is sea-surface microlayer accumulation, a threat to fish larvae and intertidal organisms (p. 137). The chapter does not say these were unforeseeable. - Human dietary exposure is not in this category. Intake from seafood, especially farmed fish, “has long been recognised”, though it was “Only more recently” evaluated (p. 140). - The authors present the marine environment as having a “complexity and indeterminate nature” (p. 142). This is framed less as a failure of individual actors than as a limit of the appraisal paradigm.

11. Persistence, irreversibility and legacy costs. - Because the chemical persists, a use decision in the 1970s carries forward into the costs of dredged-material management decades later. Those costs are “to be borne by authorities responsible for dredging operations” (p. 141), and ecological recovery is delayed (pp. 139, 141). - Recovery depends on species life histories and larval sensitivity (p. 139), which adds biological lags on top of chemical ones.

12. Distribution of costs and benefits. - Benefits went to vessel operators (fuel, drag, maintenance intervals) and, implicitly, paint manufacturers (p. 135). The chapter does not quantify them. - Costs fell on aquaculture (p. 136), harbour and dredging authorities (pp. 135, 141), non-commercial species and ecosystems (pp. 135–139), and potentially marine mammals and seafood consumers (p. 140). - Costs are also displaced geographically: Japan’s contamination continued despite its own ban (p. 142). - The authors’ opening criticism is that proponents’ evaluations omit these costs (p. 135).

13. Substitution dynamics and regrettable substitution. - TBT entered as a booster in copper paints and then, because of its “superior effectiveness over copper”, “captured a major proportion of the antifouling market” (pp. 135–136). Earlier mercury- and arsenic-containing preparations are ruled out as alternatives (p. 141). The chapter does not describe how or why they fell out of use. - Its replacements (copper plus boosters such as Irgarol 1051) showed early signs of repeating the pattern (p. 141). - Opponents used the risk of regrettable substitution as an argument against substitution (p. 141). The authors respond by reframing around function and non-release technologies (pp. 141–142). - Innovation effects are visible in three places: - the move to self-polishing copolymers (p. 136) - “technological developments” enabling lower-content paints in 1987 (p. 138) - biocide-free “non-stick” coatings, commercially available “for many years”, which “retain a significant market share for leisure craft” (p. 141) - The chapter does not examine whether regulation stimulated or slowed antifouling innovation.

14. Enforcement and voluntary measures. - The UK small-boat controls, and the PARCOM and EU harmonisation, worked mainly through retail sale restrictions (pp. 137–138). France’s 1982 law prohibited application of TBT paints to small vessels (p. 136). The “extent of illegal application remains unknown” (p. 138), and “isolated but significant illegal use” has been implicated in continued contamination in some places (p. 139). - Voluntary measures for fish cages “failed adequately to address” that use (p. 138). - Guidelines for hull-cleaning inputs exist but their effectiveness is “difficult to evaluate” (pp. 138–139).

15. How proponents and regulators thought (mental models). - Proponents framed TBT through its efficacy and economy (p. 135), a benefits-first frame. - Regulators in 1985 believed a 20 ng/l target was protective. That belief was grounded in acute-toxicity data (pp. 137, 141). - Degradation predictions for surface waters were generalised without accounting for lipophilicity and sediment binding (p. 141). The editors (p. 173) put it as “assumptions about the nature of the marine environment that were evidently incorrect for many areas”. - The exemption for large vessels rested on an “open seas” model of exposure (p. 139). - The IMO’s 1994 view was that no further controls were necessary (p. 140). - In each case confidence rested on a model built from a limited set of conditions. The chapter does not attribute bad faith to regulators. It presents their errors as paradigm-bound. It says nothing about the paint manufacturers’ motives beyond noting that they acted “in concert with other interested parties” (p. 135).

16. Framing and language. - The authors’ language escalates from the neutral (“adverse effects”, p. 135) to strong terms: “grossly underestimated” (p. 136), “regional extermination” (p. 135), “near extirpation” (p. 136), “decimated” (p. 138), “irrefutable” (p. 142). - The key conceptual framing is the contrast between “precaution” and “retrospective action” (p. 142). - They characterise the opposition’s “alternatives are also harmful” argument as “somewhat negative” (p. 141). - The phrase “far-field” effects (p. 140) marks the site of contested uncertainty.

Transferable insights (technology-neutral)#

1. Safety thresholds derived from the most obvious, high-dose endpoint can miss low-dose, mechanism-specific effects by an order of magnitude or more. A “protective” standard built on them can be badly wrong even when complied with, and even when it already includes a large margin below the observed effect level. Evidence: acute-toxicity assumptions (µg/l) lay behind an EQT of 20 ng/l, already well over tenfold below that range. Sterilisation occurred at 3–5 ng/l. The EQT was replaced by a 2 ng/l EQS (pp. 136, 137, 138, 141). Strength: strong. The numbers are specific and sourced to primary studies within the chapter, and the revision of the standard is documented.

2. Predictions about how an agent behaves after release, calibrated under one set of conditions, can fail badly in others. The failure shows up as unexpected persistence and accumulation. Evidence: half-lives of days in eutrophic surface water versus years in sediments and nutrient-poor waters. Predictions of rapid degradation “failed to account for” sediment binding and lipophilicity. Biofilm amplification recognised “only recently”. Accumulation in top predators “not envisaged” (p. 141). (Sea-surface microlayer accumulation, p. 137, is a related concern that emerged in the 1980s, but the chapter does not call it unforeseen.) Strength: strong for the case. The editors generalise it on p. 173.

3. Warnings of harm to economically valued resources can trigger faster action than warnings confined to parts of a system nobody values commercially. Harm to those parts tends to need more extensive documentation before it drives action. Evidence: the authors judge that near extirpation of the pest oyster drill would, alone, have prompted “little if any action”, whereas the oyster crop collapse prompted “relatively swift action” (p. 136). UK oyster failures also fed the evidence (p. 137). Counterpoint: the 1987 UK restrictions and the global phase-out rested largely on harm to non-commercial species (dogwhelks, offshore whelks, marine mammals) once it was well documented (pp. 137, 139–140, 142). Strength: moderate for the first-response claim, which rests on the authors’ counterfactual. The stronger claim, that such harm is simply “tolerated”, is not supported by this case.

4. A highly specific, sensitive indicator of harm can shorten the time to causal acceptance and action, partly making up for weak measurement of the agent itself. Evidence: imposex as a “chemical-specific phenomenon” that “contributed greatly to the early acceptance of a direct causal relationship” (p. 135). It partly overcame low analytical sensitivity (p. 142). The journal news profile of Bryan et al. (1986) “in many ways” marked wider acceptance (p. 137). Strength: moderate to strong. Asserted clearly and consistent with the quick action after 1986. The chapter offers no comparison with cases lacking such a marker.

5. Early detection of harm depends on baselines that nobody has a commercial reason to collect. Where they exist they are invaluable; where they are absent, the extent of harm can go unrecognised for longer. Evidence: absence of baseline data on non-commercial species “in most regions” was one of two causes of early ignorance of the problem’s geographical extent. South-west England baselines were “invaluable” (p. 142). Strength: moderate. A single positive example, stated as a general rule. The editors adopt it (p. 172).

6. Early controls tend to target the most tractable segment of a problem, not the largest source. Exemptions often rest on exposure assumptions that later prove wrong. Evidence: <25 m vessel bans; large vessels exempted on the “open seas” argument (p. 139); PARCOM unable to reach commercial shipping (p. 138); later evidence that shipping was “potentially the most significant single contributor” (p. 139). Strength: moderate to strong. The exemption rationale and later source evidence are documented. That the choice was driven by tractability is my inference, supported by p. 138.

7. Where the source crosses jurisdictions, effective control depends on an institution whose reach matches the problem. That institution may be the slowest actor, and smaller jurisdictions acting alone cannot close the gap. Evidence: regional bodies “could not achieve restrictions within the commercial shipping sector” and deferred to IMO (pp. 138, 140). MEPC’s 1994 “no further controls” decision (p. 140). Japan’s national ban did not end contamination (p. 142). Editors generalise the point (pp. 180–181). Strength: strong for the jurisdictional point, which is well documented. Weaker for the Japan illustration, which the chapter asserts from contamination data without showing that foreign vessels were the source (see limitations). “May be the slowest actor” is my inference from the 1994 decision.

8. Partial success can be recast as full success, and the “problem solved” narrative can lower pressure to address remaining sources. Evidence: Arcachon recovery “frequently cited” as showing the 1980s controls “‘solved’ the TBT problem” (p. 138). Contrary evidence of plateaued concentrations and incomplete recovery (pp. 138–139). IMO’s 1994 decision (p. 140). Strength: moderate. The claim and counter-evidence are documented. The chapter does not show causally that the narrative produced the 1994 decision.

9. When action waits for documented evidence of harm, it is retrospective by construction. Where the causal evidence is itself built from observed damage, its strength is also a measure of how much damage has already been done. Evidence: the authors’ reasoning runs “evidence … is irrefutable … It would be difficult to argue, therefore, that any of the actions … have been precautionary, resulting as they have from extensive documentation of ecological impacts”. The 1987 UK restrictions came after concentrations were already far above effect levels and gastropod populations had declined. The IMO convention came after consequences were “well documented” (p. 142). Strength: strong as a description of this case. As a general claim about precaution it is a definitional stance, since any action after harm is counted as non-precautionary. The second sentence is the note-taker’s generalisation: the chapter’s “irrefutable” describes the evidence in 2001, and the French action of 1982 came before detailed exposure data existed (p. 136). See limitations.

10. Persistence turns a use decision into a long-lived legacy whose costs fall on third parties and later generations, and recovery lags well behind controls. Evidence: years-long half-lives in sediments; dredging legacy; delayed recovery; long life histories and larval sensitivity (pp. 139, 141, 142). Strength: strong. Mechanistically clear and documented with sources.

11. The existence of imperfect alternatives is routinely used to argue against substitution. A more productive question is whether the underlying function can be delivered in a way that avoids the hazardous property altogether. Evidence: Irgarol 1051 findings “mirror early findings with respect to TBT”; opponents’ use of such examples; the authors’ reframing; non-stick coatings (pp. 141–142). Editors adopt it (p. 177). Strength: moderate. The regrettable-substitution risk is evidenced. The recommended reframing is normative, and its technical feasibility for large vessels was unproven in 2001 (p. 141).

12. Controls that act at point of sale, or rely on voluntary compliance or guidelines, leave enforcement gaps whose size is often unknown. Evidence: unknown extent of illegal application; “isolated but significant” illegal use implicated in continuing contamination in some places; failed voluntary fish-cage measures; hull-cleaning guidelines whose effectiveness is “difficult to evaluate” (pp. 138–139). Strength: moderate. Several instances, but thin quantification (“remains unknown”).

13. Measurement capability shapes what counts as evidence of harm or of recovery. High detection limits can hide both. Evidence: cause unknown until analytics improved (p. 136); detection limits slowed confirmation of recovery (p. 138); sediment trends could not be reconstructed (p. 138); “low sensitivity of analytical methods” (p. 142). Strength: moderate. Well illustrated, although the chapter treats the analytical lag as largely unavoidable.

14. Where those generating a risk share a stake in the harmed activity, controls are easier to implement. Where generators and harmed parties are separated, action stalls. Evidence: local boat owners at Arcachon (p. 137) versus the global shipping impasse (pp. 138–140). Strength: suggestive. The Arcachon point is hedged (“probably aided”), and the contrast with shipping is my inference.

15. Institutional positions can reverse quickly once coordinated pressure from sub-global actors combines with new evidence of wider harm. Evidence: MEPC went from “no further controls” in 1994 to a “sea change” in 1997 after the Esbjerg declaration and offshore evidence (p. 140). Strength: moderate. The sequence is documented, but the causal drivers inside the institution are not analysed. The chapter itself credits the 1996 re-examination to “emerging research”; the role of regional pressure is implied only by the sequence (p. 140).

16. Cost accounting that includes only the user’s private benefits and omits external, spread-out costs systematically favours continued use. Evidence: proponents’ defence rests on efficacy and fuel economy; what is “missing” is aquaculture losses, harbour-authority costs and environmental costs (p. 135). Strength: asserted. The chapter quantifies neither side beyond Arcachon tonnage.

Limitations, contestation and bias check#

Standpoint and disclosure. - Two of the three authors are Greenpeace scientists. One represented Greenpeace International at marine-protection conventions (pp. 197–198). The third is a Plymouth Marine Laboratory researcher who had worked on TBT since the early 1980s (p. 197). He co-authored several of the UK studies the chapter cites (pp. 137, 139, 141), though not the key Bryan et al. (1986) or Gibbs et al. (1988) papers. - None of this is disclosed in the chapter text. Readers find it only in the contributors list. - This does not undermine the ecotoxicological record, which is heavily referenced to peer-reviewed literature. It does bear on selection and emphasis: which recovery studies are foregrounded, how opponents are characterised (“somewhat negative argument”, p. 141), and the advocacy for a global ban and non-release alternatives. - The chapter says nothing about the role of NGOs (including Greenpeace) in the IMO process, a notable silence given the authors’ position.

Industry and interests are under-documented. - The paint manufacturers’ 1980s challenges get one sentence (p. 135), with no references, specifics or outcomes. - The chapter does not say: - who funded the research on each side (it gives no funding information for any cited study, supportive or sceptical); - how shipping interests or flag states behaved at IMO. - For a project interested in how interests shape evidence, this chapter offers little direct evidence. The dynamics have to be inferred from the structure of the case, not from documented behaviour.

Costs and benefits (question 3 of the editors’ brief, p. 168). - Benefits (fuel, emissions, efficacy, economy) are conceded (p. 135) but not quantified. - Costs are quantified only for Arcachon production (p. 136). - There is no estimate of the cost of alternatives, of the fuel and emissions penalty of less effective coatings, or of the costs of dredging and legacy management. - The editors acknowledge this as a general weakness of the case studies (p. 168). - The opening charge that proponents omit external costs (p. 135) is fair as a critique, but the chapter does not itself provide the missing accounting.

Opposing views: present but thinly engaged. - Evans (2000), Nicholson and Evans (1997), Abel (2000) and Abbott et al. (2000) are cited and their positions sketched in a sentence or two (pp. 135, 138, 140, 141). - The confounding argument that offshore whelk declines reflect overexploitation (p. 140) is not directly rebutted. - The “treatment approach” of Abbott et al. is not explained at all. - On the other hand, the chapter does present evidence of genuine recovery (Mulroy Bay, Shetland, Evans et al. 2000a; pp. 138–139). It reports the contradictory human-intake assessments evenhandedly (p. 140), and it hedges on marine-mammal immune effects (“Although uncertainty remains … might”, p. 140). That is fair practice.

The “not precaution” verdict is partly definitional. - The authors hold that any action taken after documented harm is “not precaution” (p. 142). This is a demanding standard. - Their own account of France in 1982 describes action “on the best information available” before reliable environmental measurements existed (p. 136). That arguably involved some precaution about the causal link and exposure. But it still followed severe, documented harm (the oyster collapse), so it is consistent with the authors’ verdict. - The UK 1987 ban came about a year after the “particularly influential” Bryan et al. study (p. 137). By contrast, the editors report that for many case studies the gap “was long, certainly many years or decades, and in some cases over a century” (p. 168). The authors themselves note that early concerns “led rapidly to some national and regional restrictions” (p. 135). - The editors’ key measure is “the length of the gap between the specific problem being identified and effective action being taken” (p. 168). By that measure, small-vessel TBT regulation was fast once the cause was identified. - The case’s genuine “late” elements are narrower: - the decade between the first observations and causal identification, which the chapter attributes mainly to unavoidable analytical limits; - the roughly 20-year lag for seagoing vessels; - the original appraisal failures (acute-only testing, degradation assumptions). - My reading (not the chapter’s framing): TBT shows fairly responsive retrospective regulation of small craft, jurisdictionally limited and much slower regulation of seagoing ships, and poor prospective appraisal. It is less a story of warnings ignored than of warnings that could not be attributed, plus a standard set on the wrong endpoint. The exception is the IMO’s explicit 1994 decision against further controls (p. 140).

Hindsight. - The authors explicitly acknowledge hindsight (pp. 141, 142). - The counterfactual “Had the endocrine-mediated mechanisms … been identified earlier” (p. 141) is not a claim that the mechanism was knowable in the 1970s. - The case therefore supports lessons about the design of appraisal (endpoints, baselines, persistence screening) more than lessons about the culpability of decision-makers.

Generalisability and case selection. - The authors themselves call TBT, “in some ways, a rather unusual case study” (p. 135): a single chemical class, a uniquely specific biomarker, commercially visible harm and a definable regulated product. - Lessons about speed of acceptance (insight 4) may not transfer to harms without such markers. The authors imply as much with “Unusually in the field of ecotoxicology” (p. 142).

Evidence thinness in places. - Offshore and far-field effects rest largely on correlations with shipping density (pp. 139–140). - Top-predator immune effects rest on correlations in stranded or diseased animals (p. 140). - Illegal use is unquantified (p. 138). - The Japan transboundary claim (p. 142) is asserted from contamination data. The chapter does not show that the contamination came from foreign vessels, rather than, say, legacy sediments or domestic non-compliance.

Minor textual issues: - “Ballybegs” (p. 139) should be Killybegs (reference, p. 146). - Units shift between ng/l TBT and ng/l as tin (pp. 137, 138). - Table 13.1’s 1985 entry records a paint-content limit that the text’s 1985 list omits, though the text’s January 1987 wording implies one (pp. 137–138, 143). - “Paris Convention (1978)” (p. 138): 1978 is the convention’s entry-into-force year. It was signed in 1974 (general knowledge, not verified here). - Imposex is dated 1970 on p. 136, while the introduction says undesirable consequences “became apparent in the late 1970s” (p. 135). This is reconcilable: effects were seen in 1970, attribution came in the late 1970s. - “effect” for “affect” (p. 140).

Later evidence (from general knowledge, NOT verified against sources in this pass; flagged for the hindsight strand): - The treaty: the AFS Convention was adopted in October 2001 but entered into force only in September 2008. The “2003 worldwide prohibition” therefore did not bind globally in 2003. Some regional instruments (e.g., EU rules from 2003) implemented the dates for their own fleets and ports. To check. - Recovery: widespread declines in imposex and recovery of dogwhelk populations have been reported since the ban, alongside persistent sediment contamination. To check against OSPAR assessments and primary monitoring studies. - Mechanism: later research implicated retinoid X receptor (RXR) signalling in imposex, refining the “steroid hormone metabolism” account the chapter gives (p. 135). Mammalian “obesogen” research on TBT emerged later. To check against primary literature. - Alternatives: Irgarol 1051 (cybutryne) was later restricted in the EU and added to the AFS Convention. If confirmed, this bears out the chapter’s warning (p. 141). To check. - Coatings: biocide-free foul-release coatings gained use on some larger and faster vessels, but copper-based biocidal coatings remained dominant. To check.

Notable quotes#

1. “While toxicity to fouling organisms was intentional, its propensity for wider impacts on the marine environment had been grossly underestimated.” (p. 136)
2. “An early focus on acute effects, especially mortality … failed to identify sub-lethal consequences of prolonged exposure in some taxa.” (p. 136)
3. “Had the adverse effects been limited to the loss of this species, considered a pest within the shellfish industry for its damage to oyster stocks, little if any action may have followed.” (p. 136)
4. ”…illustrates the inadequacy of the EQT as a protective instrument.” (p. 137)
5. ”…no-effect levels could not be achieved in the environment without use restrictions.” (p. 138)
6. “This is a somewhat negative argument. The more appropriate question is, can fouling control be achieved without recourse to such highly toxic, persistent and bioaccumulative substances?” (p. 141)
7. “Accumulation in top predators was simply not envisaged.” (p. 141)
8. “Though frequently underrated, baseline studies play a vital role in the early detection of adverse trends and may consequently serve the application of precaution.” (p. 142)
9. “Broader consideration of problems may give rise to more beneficial solutions than simple ‘chemical for chemical’ substitution.” (p. 142)
10. “Actions have undoubtedly contributed towards remediating the most severe problems, but this is not precaution.” (p. 142)
11. “It would be difficult to argue, therefore, that any of the actions to address TBT to date have been precautionary, resulting as they have from extensive documentation of ecological impacts.” (p. 142)

Open questions#

1. Industry conduct: what were the paint manufacturers’ 1980s “challenges” (p. 135)? Legal, scientific, lobbying? Who were the “other interested parties”? Primary sources would include UK government consultation records, the US Organotin Antifouling Paint Control Act 1988 hearings, and industry submissions to MEPC.
2. IMO politics: what drove MEPC’s 1994 “no further controls” decision and the 1997 reversal (p. 140)? What were the roles of flag states, shipping associations, paint makers and NGOs, including the authors’ own organisation? MEPC session reports for 1994–1998 are the primary source.
3. Funding and sponsorship: who funded the key studies on each side of the recovery, far-field and seafood-intake debates? The chapter gives no funding information for any study.
4. Counterfactual on appraisal: could chronic, sub-lethal testing on sensitive gastropods have been done in the 1960s–70s? Did any contemporary regulatory regime require it? This would test whether the failure was avoidable, or whether it was “an unavoidable process of methods development” (p. 142).
5. Was France 1982 precautionary? How much evidence did the French authorities actually have in 1982? This would test the chapter’s own “not precaution” verdict (pp. 136, 142).
6. Cost-benefit: what were the actual fuel and emissions benefits of TBT, and the costs of the alternatives? Were Abbott et al.’s (2000) figures credible?
7. Post-2001 outcomes: - When did the AFS Convention actually bind? - How fast did offshore imposex decline? - Did Japanese coastal contamination fall after the global ban (a test of the transboundary claim, p. 142)? - Did the 2020 OSPAR cessation target (p. 143) hold?
8. Regrettable substitution: did copper and booster-biocide systems produce TBT-like problems at scale? Did non-release technologies become viable for large commercial vessels (p. 141)?
9. Consumer-product organotins (p. 142): did regulation of organotin additives draw on the TBT lessons, as the authors hoped?
10. Mechanism revision: how does the later mechanistic understanding of imposex change the chapter’s claim that earlier identification of the mechanism would have revealed the low-dose risk (p. 141)?

Audit log#

Independent fact-check against the full text extract (pp. 135–148) and the PDF text layer (Table 13.1; contributors pp. 197–198; Ch. 16 pp. 168–191; p. 120). Page rendering was unavailable, so visual checks were replaced by text-layer checks.

Notes file - Header: corrected Ch. 16 span (pp. 168–181 → pp. 168–191, TBT mentions on pp. 170–181); “means” → “indicates” for the “Source: EEA” inference; added a note on how this audit verified. - Standpoint: removed the unsupported implication that Bryan, Gibbs and colleagues worked at Plymouth (the chapter gives study sites, not affiliations); noted Santillo et al. (1998) is co-authored by two chapter authors. - Standpoint: qualified the “opposing camp” list. The chapter uses Evans (2000) as a neutral factual source, cites Evans et al. (2000a) recovery evidence, and presents Nicholson and Evans (1997) as supporting the offshore findings. Replaced “self-critical on precaution” with an accurate description. - 13.2: added the chapter’s statement that Smith (1971) “first described” imposex. - 13.3: softened the note that the French 1982 action was “close to precautionary”; it followed severe documented harm and is consistent with the chapter’s verdict. - 13.4: corrected the claimed table–text “inconsistency” on 1985. The text’s January 1987 wording (“reduced the maximum permissible content”) implies a prior limit, so the table fills a gap rather than contradicting the text. - 13.4: restored “as low as” on 3–5 ng/l; marked the 4–7x figure as the note-taker’s calculation against the lowest reported sterilising level. - 13.4: Bryan et al. described as “particularly influential” (chapter’s term). The Marine Pollution Bulletin item was a profile or news item, not an editorial. Restored the “in many ways” hedge. - 13.4: “full European extent” replaced with the verbatim “geographical extent of the TBT problem in Europe”. - 13.4: PARCOM 87/1 scope completed (pleasure boats, fish cages; seagoing vessels “and underwater structures”). - 13.6: restored “in most parts of the bay”; added that hydrodynamic disturbance prevented sediment-trend reconstruction; caveated the 900 ng/l-as-tin conversion (the chapter does not say whether it is TBT or total butyltin). - 13.6: restored the “in some species” qualifier on French effect levels. - 13.6: restored “isolated but significant” illegal use “in some cases”, and “undoubtedly differ from location to location”. - 13.7: Nicholson and Evans “confirmed” → verbatim “provided further evidence supporting”. - 13.7: added the chapter’s caveat that “data remain limited for areas outside Europe and North America”. - 13.7: flagged the endpoint explanation of the conflicting seafood-intake assessments as inference. - 13.9: removed the unsupported claim that TBT was “a substitute for worse chemistry” (mercury and arsenic); the chapter describes TBT displacing copper on effectiveness grounds. - 13.9: restored Irgarol qualifiers (“used extensively in some areas”, “certain estuaries”); flagged “regrettable substitution” as the note-taker’s term. - 13.9: restored “Although significant technical issues remain to be resolved”. - 13.10: added that the 20 ng/l EQT already sat over tenfold below the assumed µg/l effect range, yet still failed to protect. - 13.10: added the authors’ stated hope that the lessons will allow earlier identification of future problems. - 13.10: corrected attribution. Simmonds (1986) and Lee et al. (1989) are cited as reviews of the degradation predictions, not as their authors. - 13.10: restored the “It would seem” hedge on global restrictions. - 13.11: restored the “It could be argued, similarly” hedge and the concession that the treaty’s significance “should not be underestimated”; set out the chapter’s “therefore” logic (strong evidence at the time of action = action came late). - Table 13.1 notes: reframed the 1985 point; added the table’s “use” versus the text’s “application” for France 1982. - Editors’ use: added context for p. 170 and the verbatim p. 168 gap metric; recorded that there are no other TBT mentions in Ch. 16–17. - Timeline: 1970 oyster-drill row no longer says “ignored because the species was a pest”; it now gives the authors’ counterfactual. - Timeline: 1983–85 row corrected from “TBT” to butyltins (as tin); added late-1980s <10 ng/l confirmation. - Timeline: 1986 row “decisive”/”editorial” corrected; actor labels not found in the chapter (“Plymouth scientists”, “UK government scientists”, “Dutch scientists”, “Japanese, US, European scientists”, “Academic malacologist”) replaced with cited authors. - Timeline: 1996 row now carries the chapter’s stated reason (“On the basis of emerging research”). - Lags: separated the cause-identification delay (analytical limits, p. 136) from ignorance of geographical extent (analytics plus baselines, p. 142). - Harms: added unsaleable deformed oysters and losses “imposed upon harbour authorities” (p. 135). - Authors’ lesson 1: “thresholds were set in µg/l” corrected to “assumed” (the EQT was 20 ng/l); added the mechanism counterfactual. - Recommendations 2 and 6: restored the “It would seem” hedge; marked the consumer-products point as implied, not recommended. - Mechanism 2: rewritten. Commercial salience explains the first (French) action. UK 1987 and global action rested on documented harm to non-commercial species, so “the warning system was keyed to economic loss, not ecological loss” was removed as contradicted by the source. - Mechanisms 3–4: interest-alignment contrast and tractability explicitly flagged as the note-taker’s inference. - Mechanism 5: corrected “regional bodies could only lobby” (they did regulate small craft). The chapter explicitly credits “emerging research” for MEPC’s shift; regional pressure is only implied by sequence. - Mechanism 7: “Opponents” → sceptics; noted that Nicholson and Evans themselves found offshore imposex. - Mechanism 9: “journal editorial” → news profile, with the “in many ways” hedge. - Mechanism 10: removed human dietary exposure from the “not anticipated” list (the chapter says it “has long been recognised”); separated routes explicitly called underestimated from later-emerging concerns such as the microlayer. - Mechanism 13: removed the claim that TBT displaced mercury and arsenic formulations; “niche market” corrected to “significant market share for leisure craft”. - Mechanism 14: corrected “retail sale restrictions, not use bans”. France’s 1982 law prohibited application. - Mechanism 15: “assumed a uniform marine environment” replaced with the chapter’s and editors’ actual wording. - Insight 1: added the large-margin point and its evidence. - Insight 3: reformulated; added the UK and global counter-evidence; strength qualified (the strong “tolerated” form is unsupported). - Insight 4: “editorial” → news profile. - Insight 5: “decisive” and “harm goes unnoticed” softened to the chapter’s “invaluable” and “early ignorance of the geographical extent”. - Insight 7: the Japan example is no longer called a “clean illustration” (that clashed with the limitations section); strength split between the jurisdictional point and the Japan example; “slowest actor” flagged as inference. - Insight 9: rewritten to follow the chapter’s logic. Action came because evidence of harm was extensive, which is why it was retrospective. It no longer says strong evidence “does not produce timely action”. - Insight 12: added source qualifiers (“isolated but significant”, “difficult to evaluate”). - Limitations: Langston “produced much of the underlying UK science” → co-authored several cited UK studies, but not Bryan et al. (1986) or Gibbs et al. (1988). - Limitations: removed the insinuation naming specific studies (Cardwell et al.; Evans/Nicholson) as possibly industry-sponsored; now states the chapter gives no funding information for any study. - Limitations: in the “not precaution” discussion, noted that the French action followed documented harm; quoted the p. 168 gap metric verbatim; added the authors’ own “led rapidly” (p. 135); labelled the overall reading as the note-taker’s; noted the 1994 IMO exception. - Minor textual issues: reframed the Table 1985 item; added the Paris Convention date (1978 entry into force vs 1974 signature, flagged as unverified general knowledge). - Open question 3: neutralised so that no specific study is singled out. - Notable quotes: added the p. 142 “therefore” verdict sentence.

Digest file - Core story: “nobody expected effects at parts per trillion” → effects assumed to need µg/l on the basis of acute tests. - Core story: Bryan et al. “decisive” → “particularly influential”. - Core story: “illegal use continued” → implicated in some areas. “They proved to be ‘potentially…’” → the authors call them. 68 t marked as an estimate. Offshore significance noted as disputed. - Core story: IMO sequence clarified (1996 re-examination “On the basis of emerging research”; deadlines by Assembly resolution in 1999; convention expected 2001). - Lessons: restored the “It would seem” hedge; added the baseline and monitoring lesson. - Headline verdict: corrected the inverted logic (“despite irrefutable evidence” → “therefore”, because action followed extensive documentation); added the treaty concession. - Mechanisms: commercial salience limited to the first action, with the later non-commercial-species evidence noted. - Insights 1, 3, 5, 7 and 9 revised in line with the notes. - Caveats: Langston description corrected. The “not precaution” caveat now notes that both fast actions followed documented harm, and that the decade before the cause was identified is also a late element.

Checked and found accurate (no change): all author affiliations (pp. 197–198); all Ch. 16 cross-references (pp. 170, 171, 172, 173, 177, 180–181) and p. 120; Table 13.1 content; the dose, concentration, tonnage and date figures; the verbatim quotes in “Notable quotes”; the “Ballybegs”/Killybegs and “effect”/”affect” textual notes; no mention of any out-of-scope contemporary technology or company.

Second audit pass#

Re-read the full extract (pp. 135–148) against the notes and digest after the first pass. Re-extracted the contributors entries (pp. 197–198), all Ch. 16 TBT mentions (pp. 168–181), p. 120 and a whole-report search for TBT, organotin and imposex from the PDF text layer. Affiliations, Table 13.1 and the Ch. 16 quotations were confirmed.

Notes file - Standpoint: removed Gibbs et al. (1988) from the list of south-west England studies; the chapter gives no study site for it. - 13.1 and Limitations: restored the “in some ways” hedge on “a rather unusual case study”. - 13.2: restored “in some gastropods” and “can be initiated” on low-ng/l imposex; shell deformity and larval mortality are attributed to “concentrations also known to cause” them. - 13.3: completed the quote “Acting on the best information available linking the oyster collapse to the presence of TBT”; France prohibited “application”, not use in general. - 13.4: restored “in many coastal waters” (EQT exceedance); added the UK–Arcachon “many parallels” point. - 13.4: separated the seasonal-cycle point (all 1980s butyltin data) from the “outside marinas” point, which the notes had run together. - 13.4: added “in recognition of the low-dose effects” to the EQS revision; noted “Two years later” counts from 1987. - 13.6: added the authors’ concession that oyster-bed recovery “provided the first indirect evidence of the effectiveness of the ban”. - 13.7: “mostly” continuous leaching → leaching was “the major contributor” (a plurality is not stated to be a majority); 68 t marked as an estimate. - 13.9: added the authors’ description of the IMO measure as “progressive” and the “economically viable” qualifier; added that Abel co-authored Abbott et al., so the two opposing citations overlap. - 13.9: “displacement … of copper” → TBT “captured a major proportion of the antifouling market” through “superior effectiveness over copper” (the chapter does not say “displaced”). - 13.9: restored the hedge that commercial natural antifoulants “may remain some time away”. - 13.10: “The consequences are delayed recovery” → persistence “raises the prospect of” delayed recovery. - 13.10: corrected a non-verbatim quote (“and their complexity …”) to the source wording (“pathways and interactions”, “coupled with an appreciation of their complexity and indeterminate nature”). - 13.11: added the opening sentence on “widespread and sometimes severe environmental effects”; noted that “irrefutable” describes the 2001 evidence, not the evidence at each action (France 1982 preceded detailed exposure data); labelled the “strength of evidence = damage done” point as the note-taker’s gloss. - Table 13.1: “compiled by the EEA” softened to what the “Source: EEA” label indicates. - Editors’ use, p. 170: corrected a misattribution. The “relatively inert and benign … not considered in the risk appraisal” sentence refers to CFCs, not TBT. - Editors’ use, p. 171: added the context “regulators were often slow to react” once irreversibility became apparent. - Lags: the imposex-to-cause lag is now 8–12 years, since the connection was made in “the late 1970s and early 1980s”. - Harms: “about 150” species → “as many as 150” (Vos et al., 2000). - Authors’ lesson 4: specified that the early ignorance concerned the problem’s “geographical extent”. - Mechanism 13: same “captured a major proportion” correction as 13.9. - Mechanism 15: removed “paint manufacturers’ challenges as interest-driven”. The chapter says only that they acted “in concert with other interested parties” and attributes no motive. - Insight 2: the microlayer is no longer offered as evidence of an unforeseen pathway; the chapter calls it an “additional concern”, not unforeseen. - Insight 9: “conclusive evidence” → “documented evidence”; the second sentence is limited to evidence built from observed damage and flagged as the note-taker’s generalisation. - Insight 15: added that the chapter credits “emerging research” for MEPC’s re-examination. - Limitations: the unsourced “Many other Late Lessons cases involve lags of decades” is replaced with the editors’ own p. 168 statement.

Digest file - Core story: “led regulators to assume” → it was “initially assumed” (the chapter names no actor); “not anticipated” → “grossly underestimated”. - Core story: France “banned TBT” → prohibited application, with the full “best information … linking the oyster collapse” quote and a note that this came before detailed exposure data. - Core story: marinas “often exceeded” → summer concentrations “regularly exceeded”; Bryan et al. “to some degree” restored. - Core story: “North Sea ministers pushed back” → the Esbjerg commitment “in spite of” the 1994 decision. - Lessons 2 and 4: wording aligned with the source (“failed to account for”; “geographical extent”). - Headline verdict: added the 2001-evidence caveat and the chapter’s closing point on consumer-product organotins. - Mechanism “tractable segment”: flagged as the note-taker’s inference. - Insight 9: reworded to match the notes; insight 12 page reference corrected to pp. 138–139. - Caveats: “left unanswered” corrected. The confounder is not directly rebutted and the treatment approach is not described, but the authors do answer the substitution argument.

Checked and found accurate in this pass (no change): author affiliations and biographies (pp. 197–198); Ch. 16 quotations on pp. 168, 170, 171, 172, 173, 177 and 180–181, and p. 120 (Ch. 11, MTBE); no TBT mentions elsewhere in the report outside Ch. 13, Ch. 16, p. 120, the contents, the biographies and the index; all dose, concentration, tonnage, date and lag figures; the Notable quotes; no out-of-scope contemporary technologies or companies.