Biopesticide registration (FIFRA, EU 1107/2009)

Pesticide regulation was built around a synthetic molecule: one defined structure, a purity specification, a dose that can be measured in soil and on a leaf, and a toxicology package that reads out as a no-observed-adverse-effect level. A microbial or biochemical control agent breaks almost every one of those assumptions, and the registration pathways for biopesticides exist because of the mismatch, not because of a policy preference.

The regulated unit is a strain, not a concentration#

A Bacillus thuringiensis product is not characterised by “how much Bacillus thuringiensis” it contains. Different strains of the same species carry different insecticidal crystal genes, different secondary metabolite repertoires and different host ranges; two isolates that are indistinguishable by 16S sequence can differ in what they do to a non-target insect. So identity is fixed at strain level, tied to a deposit in a recognised culture collection and now routinely supported by whole-genome sequencing. Potency is expressed in viable units — colony-forming units, spores, occlusion bodies — which is a biological count with its own variance, not an analytical concentration.

That has a consequence the dossier has to carry: the hazard question for a micro-organism is usually not the organism but what it makes. Secondary metabolites and any toxigenic potential in the genome are the substantive part of the assessment, which is why the EU data requirements for micro-organisms were rewritten in Commission Regulations (EU) 2022/1439 and 2022/1440 to centre on metabolite characterisation and genome-based screening for antimicrobial resistance determinants rather than on repeated animal studies.

Tiered testing, and why residues mostly disappear#

Non-target testing runs as a tier system. Tier I applies a deliberately exaggerated dose to a small set of representative organisms — bees, birds, fish, daphnids, earthworms, a predator or parasitoid — and a clean result at maximum hazard closes the question. Only a Tier I effect triggers dose-response work at Tier II and field studies at Tier III. This is efficient precisely because most biological agents fail to show an effect at doses no field application could reach.

Residues follow a different logic again. For an organism already ubiquitous in soil or on plant surfaces, a maximum residue level measured against a background of the same organism is not meaningful, so both systems allow exemption from a numerical tolerance where infectivity and toxin production have been excluded — under the US Federal Food, Drug, and Cosmetic Act for the American route, and through the residue definition provisions of Regulation (EC) No 1107/2009 in Europe.

Two systems that divide the problem differently#

The structural divergence is worth stating plainly. FIFRA registers a product: the formulation, its label and its use pattern are approved together by the EPA, with biopesticides handled by a dedicated division under statutory review clocks. Regulation (EC) No 1107/2009 splits the decision — the active substance is approved once at Union level after EFSA peer review, and the product is then authorised separately by member states, by zone. The same organism therefore reaches the two markets through different evidence, and the gap in the European path is usually not the science but the second step.