Pest resurgence

Why can spraying a crop leave it with a worse pest problem than it had before?

A field is sprayed. The insecticide works exactly as advertised: within a day the pest is almost gone. Three weeks later the same pest is back at a density higher than before the spray was applied, and the grower sprays again, sooner and heavier.

The tempting explanation is resistance — the survivors were the tough ones, and their offspring inherited it. That does happen, and it is a real and serious problem. But it cannot be this story, because resistance takes generations to build and produces a slow erosion of effectiveness, whereas resurgence can appear after the very first application, with a chemical the pest is still perfectly susceptible to. So what else could a spray do to a field that leaves the pest better off?

Ask what else was living in that field. A crop is not just a pest and a plant. It holds predators — ladybirds, spiders, predatory mites, ground beetles — and parasitoid wasps that lay their eggs in the pest. Most of the time these are invisible, precisely because they are working: the pest sits below the level at which you notice it, and you never see the fraction being eaten.

Now, a broad-spectrum insecticide is broad-spectrum. It does not know which arthropod you resent. So the spray knocks down both the pest and the things that eat the pest.

That alone is not yet a problem — if both recovered at the same rate, you would return to the same balance. So the real question is: after a shared crash, which side gets back up first? Here the two sides are not symmetric, for reasons that have nothing to do with the chemistry.

The pest typically has a shorter generation time and lays far more eggs. Its food — the crop — is still standing, undamaged and abundant. And it was numerous to begin with, so even a small surviving fraction is a lot of individuals.

The predator sits a trophic level up, so it was always the rarer of the two; a heavy kill of a small population can leave almost nothing. It generally breeds more slowly. And it faces a problem the pest does not: its own food supply has just been destroyed, so the few survivors starve or leave the field to look for prey elsewhere. There is an exposure asymmetry too — a predator quartering treated foliage and eating contaminated prey may pick up more residue than the pest that sat still.

Follow that through. The pest rebuilds into a field that has been emptied of the things that used to hold it down. It is not merely recovering; it is recovering with the brake removed. So it does not stop where it was before. It overshoots. That is resurgence, and note that it needs no resistance at all — only a difference in recovery rates.

There is a companion effect worth naming. A field usually contains several herbivores that never become pests because natural enemies keep them beneath the threshold at which they matter. Remove the enemies and one of those can be released into a full outbreak — a secondary pest outbreak, a problem the spray did not worsen but created. Spider mites are the standard example, flaring after treatments aimed at something else entirely.

And the loop closes. The outbreak provokes another spray, which suppresses the enemies further, which deepens the next resurgence. This ratchet is called the pesticide treadmill, and its best-documented cases are real agricultural collapses: cotton in the Cañete Valley of Peru in the 1950s, and rice in Asia, where the brown planthopper rose to major-pest status alongside intensive insecticide use that suppressed its natural enemies. Honesty requires adding that resistance was usually tangled up in these episodes too; resurgence and resistance are separate mechanisms that tend to arrive together.

The essential point is that this is a feedback story, not a toxicology story. The chemical is doing exactly what it claims. The damage lies in what it does to the relationship between two populations with different recovery speeds.

Two refinements. First, "the predators recover more slowly" is the general pattern, not a law — some natural enemies are quick, some pests are sluggish, and whether resurgence follows depends on the particular species pair, the timing, and whether untreated refuges nearby can re-seed the field with enemies. Second, a few insecticides have been shown to raise pest reproduction directly at sub-lethal doses, which compounds the effect; but that is an extra mechanism, not the main one, and resurgence does not need it.

This also explains why integrated pest management is built as it is. Spraying at an economic threshold rather than on a calendar, choosing selective products, leaving refuges: these are not squeamishness about chemicals but attempts to keep the second population — the one you were never trying to kill — alive enough to do the work it was already doing for free.

• How economic thresholds are calculated, and why spraying on a calendar almost guarantees the treadmill.
• Why the same asymmetry between predator and prey recovery shows up in fisheries and in disease control.

Every term the collection defines is gathered in the glossary.