Port biological quarantine

A port inspection never establishes that a consignment is free of a quarantine pest. It establishes that if the pest were present above some chosen level, the inspection would probably have found it. Everything that follows — sample sizes, treatment schedules, the design of molecular screening — is an argument about that word probably.

Sampling against low prevalence

The statistics are unforgiving. Detection follows a hypergeometric relationship between lot size, sample size, the prevalence you are designing against and the confidence you want. The international standard for consignment sampling, ISPM 31, works from exactly these terms, and its familiar worked case is instructive: to be 95 % confident of detecting an infestation present in 0.5 % of units, roughly 600 units must be examined, and for large lots that number barely moves with lot size. Halve the design prevalence and the sample roughly doubles.

Two consequences follow. First, a truly rare pest — one insect in ten thousand cartons — is not detectable by any sampling regime a port can operate; the sample size required exceeds the consignment. Second, because the inspection budget is fixed, the real decision is not how hard to look but where to look, which is why border systems allocate effort by pathway risk rather than uniformly.

Sampling also assumes something usually false: that the pest is distributed at random. Infestation is typically clustered — one nest, one contaminated pallet, one part of a hold — and clustering inflates the effective error, so a sample drawn from a single accessible location is worth much less than its nominal size.

When detection fails, treatment substitutes

Because detection has a floor, the system leans on treatments whose lethality is defined rather than sampled. ISPM 15 for wood packaging is the clearest case: heat treatment to a core temperature of 56 °C held for 30 continuous minutes throughout the profile, or dielectric heating to 60 °C for one minute throughout, or fumigation. The mark on the pallet is not evidence of inspection; it is evidence that a specified lethal condition was applied to the whole volume, and that is a stronger claim than any sample can support. Ballast water is regulated the same way, against a numerical discharge limit — the IMO D-2 standard of fewer than ten viable organisms of 50 µm or larger per cubic metre.

What molecular methods actually move

Rapid nucleic-acid detection changes the shape of the problem in one specific way: it collapses the time from sample to answer, so more of a consignment can be examined inside the dwell time a port can tolerate. More samples at the same cost means a lower design prevalence, and that is the real gain.

What it does not do is give certainty. Amplification detects sequence, not viability, so dead material and residual DNA both read positive; PCR inhibitors in wood, soil and organic debris cause false negatives; and any assay is only as good as the reference sequences behind it, which for many quarantine taxa are sparse or misidentified in public databases. Molecular screening therefore usually triggers a decision to hold and confirm rather than replacing morphological identification outright.