LL2-12 digest: Part B divider, and Ch12 Booster biocide antifoulants: is history repeating itself?#

Late lessons from early warnings II (EEA 2013), pp. 261–278. Authors: Andrew R. G. Price (marine biologist and consultant; York and Warwick) and James W. Readman (Plymouth Marine Laboratory). The chapter has no panels. Pages 261–264 are only a divider, the Part B contents and a list of panels. Part B’s framing, “degradation of natural systems… feedback loops… substantial unknowns”, is in the report introduction (p. 10).

Core story#

Ships need antifouling. Fouling carries heavy fuel and speed penalties (p. 266), figures the chapter takes partly from a trade article titled “Environmental benefits of TBT antifoulants” (p. 277).

• TBT. It was highly effective and highly toxic. It caused imposex in about 150 gastropod species, cost Arcachon about USD 147m in oyster losses, and harmed molluscs below 10 ng/L (p. 266). The chapter gives six reasons its hazards were underestimated: a wrongly “protective” 20 ng/L target, underestimated persistence and bioaccumulation (two bullets), missed scale, missed transboundary spread, and tolerance of imposex because it had “no immediately obvious economic cost” until the oysters failed (p. 267).
• The substitutes. The 1989 EC ban on small-boat TBT was “triggering the development and use of booster biocides” (p. 276). These were drawn from existing agrochemicals, biocides and even a shampoo antifungal (p. 267), and “believed to be less damaging” (p. 265).
• The warnings.
• 1993: Irgarol 1051 found at up to 1,700 ng/L on the Côte d’Azur, by chance, during a survey of agricultural herbicides that used the same analytical method (p. 267).
• 1996: periphyton effects at concentrations found in coastal waters (pp. 267–269).
• Later work found a natural-phytoplankton EC50 of 70 ng/L and coral-symbiont photosynthesis inhibition down to 60 ng/L (p. 270). The summary says some boosters can damage primary producers “At current environmental concentrations” (p. 265).
• Monitoring. The EC-funded ACE project (1999–2002, 800 samples) mapped contamination across Europe (pp. 268–269, 274).
• The shift in harm. The substitutes swap TBT’s specific endocrine harm for broad-spectrum inhibition of photosystem II. Because the target protein is highly conserved, the base of food webs is at risk (pp. 273–274).
• Action. Denmark and the UK banned Irgarol and diuron on small craft in 2000–2001, and UK Irgarol concentrations then fell (pp. 271, 273). Bermuda funded local coral toxicology and banned both in 2005 (p. 271). EU-wide review under the Biocidal Products Directive was still unresolved in 2008, with no timescales for further data and decisions “expected shortly” (pp. 271–272).

Authors’ lessons#

From the evidence: history is repeating as a pattern: assume a replacement is better, monitor it, concern builds, ban the worst agents, search again (p. 273). Ecological effects “remain poorly understood but of considerable potential concern” (p. 274; “considerable concern” in the summary, p. 265). Targeted bans at manufacturer level reduce concentrations (p. 273). Analytical capacity no longer limits detection (p. 275).

Recommendations: keep monitoring banned and permitted agents, starting with Irgarol, with Chagos as a clean reference (p. 274). Tie monitoring to predetermined action thresholds, or it becomes an “academic pursuit” (p. 274). Use PBT criteria to target legislation (pp. 265, 276). Address “geographical disparities” in national rigour, and appraise future biocidal products better (p. 265). Fast-track non-toxic alternatives; controlled release is only a “stop-gap” (pp. 274–276).

Transferable insights (strength)#

These also cover the main mechanisms, so there is no separate mechanisms section.

1. Banning a hazard creates a market for its substitutes, which can inherit an untested assumption of relative safety (pp. 265, 267, 273, 276). Strong for the substitution sequence; moderate for the untested-assumption part.
2. When the useful function is the hazard, swapping within the same principle relocates harm. Improvement needs a different principle (pp. 266, 273–276). Moderate.
3. A substitute’s hazards are often foreseeable from its known mode of action. The boosters were herbicides “by definition” (pp. 267, 273). Moderate (my inference).
4. Comparative safety claims made before deployment are hypotheses that only monitoring after deployment can test (p. 273). Moderate.
5. Warnings often come from untargeted or academic work (pp. 267, 274). Moderate.
6. Monitoring without action thresholds drifts into academic exercise (p. 274). Asserted.
7. Harm without market value is tolerated until something with market value is hit (p. 267). Moderate (explicit, but one uncited judgement for TBT only).
8. Early “protective” thresholds tend to be too lenient. TBT’s 20 ng/L target; Irgarol’s “proposed EQS” of 24 ng/L against effects at 60–70 ng/L (pp. 267, 269–270, 273). The EU later set 2.5 ng/L [external, 2013]. Moderate.
9. Things that are hard to measure can look safer (p. 270, 273). Suggestive.
10. Partial or national controls leak with mobile sources and legacy stocks, and early partial success gets “hailed… as a ‘solution’” (pp. 267–268, 273). Moderate.
11. Simple rules at the supply choke point can work despite weak retail enforcement (p. 273). Moderate (one study).
12. Small jurisdictions that fund their own evidence act faster than harmonised regimes (pp. 271–272). Moderate.
13. Incumbents innovate within their existing competence; radical alternatives come from outsiders (p. 275). Asserted.
14. Prohibition reduces but rarely ends release (p. 276). Suggestive.
15. Where the most sensitive receptors lie outside the regulating jurisdiction, risk assessment and law can miss the worst harm. Corals are “not directly relevant within the EU”, but dependent territories have reefs (pp. 270–271). Suggestive.

Caveats#

• Insider account. Readman made the first detection and 13 of the 41 references are the authors’ own work, though several pivotal studies are by others. There is no industry or regulator voice, and no companies are named.
• Harm is mostly potential. The evidence is laboratory and microcosm effects plus harbour concentrations at or above effect levels; the harbour-to-reef link is a “possibility” (p. 270). The summary says the chemicals “threaten… coral reefs” (p. 265), but the chapter reports no reef measurements near boating areas. (The authors’ own Chagos study found “negligible risks” (p. 277), but Chagos is a deliberately pristine reference site, so that finding does not bear on reefs near harbours.)
• Effectiveness is overclaimed. “Policy has proven effective… limited adverse ecological impacts” (p. 276) rests only on UK concentration data, from a before–after study co-authored by Readman. There are no Bermuda data and no ecological data for either place.
• One figure is ambiguous. Cresswell reported residual Irgarol at 10–55% of earlier levels (a 45–90% cut). The chapter’s “by 10–55 % of levels” (p. 273) is easily misread as a much smaller reduction.
• The endocrine reassurance goes beyond its evidence. It rests on a finding of no “strong” oestrogenic response (p. 270), which cannot rule out TBT-like masculinising effects. [external] Later work reported endocrine (oestrogenic) effects of DCOIT in fish (Chen et al., 2014), but at concentrations probably well above environmental levels.
• Omissions. Copper, the main biocide, is left in the background. There are no risk–risk trade-offs (fuel use, invasive species) and no costs. The evidence is dated by 2013: legislative tables from 2002 and 2008, latest studies from 2006–2010.
• Fair reading (my judgement). History rhymed, but the loop closed faster than for TBT. National action came about 7–8 years after first detection. [external] EU non-approval of cybutryne came in 2016 (about 23 years), and the IMO ban took effect in 2023 (about 30 years). On the chapter’s evidence the case is closer to a precautionary success in progress than to a classic false negative. Note, though, that the report’s editors group it with cases of “early warnings and (usually) late actions” (p. 644).
• Small source slips. Table 12.4 dates the BPD’s entry into force to 14 May 2000 (probably the transposition deadline; external, unverified). The chapter is more confident about the TBT ban’s success on p. 273 (“well demonstrated”) than on p. 267 (“mixed, or slower than expected”).