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

Late lessons from early warnings (EEA, 2001), Ch. 13, pp. 135–148. Authors: David Santillo and Paul Johnston (Greenpeace Research Laboratories) and William J. Langston (Plymouth Marine Laboratory); affiliations are from pp. 197–198. No panels. Table 13.1 (p. 143) was compiled by the EEA.

Core story#

TBT antifouling paints took over the market in the 1970s because they were “extremely effective and relatively economical” (p. 135). On the basis of acute toxicity tests it was “initially assumed” that effects needed µg/l concentrations, and the first UK target (20 ng/l) was judged protective on that basis. The potential for effects at parts per trillion had been “grossly underestimated” (pp. 136, 141).

Oddities appeared in 1970–71. Female marine snails near harbours were growing penises, a condition called imposex (Blaber; Smith). The cause remained unknown for about a decade, until analytical chemistry improved (p. 136). At Arcachon Bay the first victim was the oyster drill, a pest that was nearly wiped out. The authors think its loss alone would have prompted “little if any action” (p. 136). Then the oyster crop failed from 1976. Production fell from 10,000–15,000 t to 3,000 t by 1981 (p. 136).

France prohibited application of TBT paints to vessels under 25 m in 1982, “acting on the best information available linking the oyster collapse to the presence of TBT”, before detailed exposure data existed (p. 136). The UK followed: - 1985: sales controls and a 20 ng/l environmental quality target. Dogwhelks turned out to be sterilised at 3–5 ng/l, and summer marina concentrations regularly exceeded 100 ng/l (p. 137). - 1986: the “particularly influential” Bryan et al. study, finding all south-west England dogwhelk populations affected “to some degree” (p. 137). - May 1987: a retail sale ban (p. 138). - About two years later: a revised standard of 2 ng/l (p. 138).

PARCOM, the US, Canada, Australia, New Zealand and the EU (1991) followed (pp. 138, 143).

These small-boat bans reduced inputs and allowed partial recovery. But concentrations plateaued above effect levels in places, sediments acted as reservoirs, and illegal use was implicated in some areas (pp. 138–139). Large ships had been exempted on the argument that they spend most of their time on the open sea (p. 139). The authors call them “potentially the most significant single contributor” to continuing contamination. One estimate puts shipping inputs at 68 t of TBT a year into the North Sea alone, and offshore imposex correlates with shipping density, though its significance was disputed (pp. 139–140). Butyltins turned up in marine mammals, with suggestive immune correlations (p. 140).

In 1994 the IMO’s environment committee said no further controls were needed. “In spite of” that decision, North Sea ministers at Esbjerg in 1995 committed to seek a worldwide phase-out through the IMO, and the committee agreed to look again “On the basis of emerging research” in 1996. It reversed course by 1997. Deadlines for a global phase-out (2003 for application, 2008 for hulls) were adopted by IMO Assembly resolution in 1999, with the convention expected in 2001 (pp. 140–141).

The authors’ lessons (pp. 141–142)#

1. The hazard was underestimated because testing focused on acute toxicity.
2. Persistence was underestimated because predictions of rapid degradation in surface waters “failed to account for” lipophilicity and sediment binding.
3. Bioaccumulation was underestimated; top-predator accumulation was “simply not envisaged”.
4. Early ignorance of the problem’s geographical extent came from low analytical sensitivity (unavoidable) and missing baselines for non-commercial species (avoidable).
5. National bans cannot fix a transboundary problem. Japan’s own ban did not end its contamination, so “It would seem” only global restrictions can.
6. Imperfect alternatives do not justify inaction. Prefer options that release nothing hazardous over “chemical for chemical” substitution.
7. Baseline studies and long-term monitoring are “frequently underrated” but vital for early detection.

The headline verdict follows from the strength of the evidence. Unusually for ecotoxicology, the causal evidence is “irrefutable”. Because action came only after “extensive documentation of ecological impacts”, none of it “therefore” counts as precautionary. That includes the IMO treaty. The authors say its significance “should not be underestimated”, but it is still “fundamentally retrospective” (p. 142). (“Irrefutable” describes the evidence in 2001; the criterion applied to each action is that it followed documented harm.) The chapter ends by noting that organotins will still enter the sea from consumer products, and whether those “emerging challenges” draw on the TBT lessons “remains to be seen” (p. 142).

Main mechanisms#

• Testing paradigm blind spots. Acute endpoints and wrong fate assumptions (pp. 136, 141).
• Commercial salience triggers the first action. The pest species versus the oyster crop (p. 136). Later UK and global action rested on harm to non-commercial species, once it was well documented (pp. 137, 139–140, 142).
• A specific biomarker speeds causal acceptance (pp. 135, 142).
• Regulate the tractable segment first. Exemptions rested on flawed exposure assumptions (pp. 138–139). (That tractability drove the choice is the note-taker’s inference.)
• Jurisdiction mismatch. Regional bodies could not reach shipping; the global body was slow, then reversed (pp. 138, 140).
• “Problem solved” narratives built on partial recovery (p. 138).
• Persistence creates legacy costs borne by third parties such as dredging authorities (p. 141).
• Regrettable substitution (Irgarol 1051) used as an argument against substitution (p. 141).
• Enforcement gaps in point-of-sale and voluntary controls (pp. 138–139).

Transferable insights (technology-neutral)#

1. Thresholds set from the most obvious, high-dose endpoint can be wrong by 10x or more, even when complied with and even with a large built-in margin (pp. 137–138, 141). Strong.
2. Behaviour predicted under one set of conditions fails in others, showing up as unexpected persistence and accumulation (p. 141). Strong.
3. Harm to valued interests triggers faster first action. Harm to parts of a system nobody values commercially needs more documentation before it drives action (pp. 136–137, 142). Moderate. It rests on the authors’ counterfactual, and later UK and global action on non-commercial species tempers it.
4. A specific, sensitive indicator speeds causal acceptance (pp. 135, 142). Moderate–strong.
5. Baselines for things nobody values commercially are invaluable for early detection (p. 142). Moderate.
6. Early controls hit the tractable segment, not the largest source (pp. 138–139). Moderate–strong.
7. Mobile, transboundary sources need an institution with matching reach, which may be the slowest actor (pp. 138, 140, 142). Strong for the jurisdictional point; the Japan illustration is weaker.
8. Partial success gets recast as “solved” (p. 138). Moderate.
9. When action waits for documented evidence of harm, it is retrospective by construction. Where the causal evidence is built from observed damage, its strength also measures the damage already done (p. 142; the second sentence is the note-taker’s generalisation). Strong descriptively; partly definitional.
10. Persistence turns a use decision into a long legacy borne by others (p. 141). Strong.
11. Imperfect alternatives are used to resist substitution; reframing the question around function helps (pp. 141–142). Moderate (normative).
12. Point-of-sale and voluntary controls leave unmeasured enforcement gaps (pp. 138–139). Moderate.

Caveats#

• Standpoint. Two authors are Greenpeace scientists, one representing Greenpeace at marine conventions. The third, a Plymouth Marine Laboratory TBT researcher, co-authored several of the UK studies cited. None of this is disclosed in the chapter.
• Industry conduct is barely documented. One sentence on paint makers’ challenges (p. 135). Nothing on research funding or on IMO lobbying.
• Costs and benefits are not quantified beyond Arcachon tonnage.
• Opponents are cited but thinly engaged. The overexploitation confounder (p. 140) is not directly rebutted, and Abbott et al.’s “treatment approach” is never described. The authors do answer the “alternatives are also harmful” argument, calling it “somewhat negative” (p. 141).
• The “not precaution” verdict is demanding. The French 1982 action and the one-year UK response were fast by Late Lessons standards, though both followed documented harm. The truly “late” elements are the decade before the cause was identified (blamed mainly on analytical limits), the shipping lag of about 20 years, and the original appraisal failures.
• Far-field and top-predator evidence is largely correlational.
• Hindsight is explicitly acknowledged (pp. 141–142).

Later evidence (unverified): treaty entry into force (2008), imposex recovery, a revised mechanism, and Irgarol restrictions.