Pests, Disease, and What Actually Gets Sprayed

TL;DR. Roughly 20 to 40 percent of global crop production is lost to pests and disease despite everything thrown at them. Pesticides are the tool that made modern yields possible and they have a genuine cost: some older classes were seriously toxic, resistance evolves relentlessly, and drift and runoff damage things that were not the target. The residues on the food you buy are, in nearly all cases, very small fractions of levels shown to be safe in animal studies, and this is one of the most heavily monitored things in the food supply. The real risks concentrate on the people who apply the chemicals, not the people who eat the produce.

Key takeaways

• Roughly 20 to 40 percent of potential global crop yield is lost to pests, weeds, and disease even with current controls. Without any control, losses in some crops would exceed 80 percent.
• Pesticide toxicity varies enormously. Comparing a modern selective insecticide to 1950s organophosphates or to DDT is like comparing a modern car to one without seat belts.
• Resistance is inevitable and is managed by rotating modes of action, not by spraying harder.
• Organic farming uses pesticides, drawn from an approved list of mostly natural origin compounds. Some of them, like copper, are more environmentally persistent than the synthetics they replace.
• Washing produce removes some surface residue; peeling removes more; neither is the main reason residues are low. Regulatory limits and pre-harvest intervals are.

What is actually attacking the crop

In short: Four categories, each needing a different control strategy: insects, fungi, weeds, and everything else.

The largest single loss category worldwide is weeds, which is unglamorous and true. Weeds do not make headlines but they quietly take a bigger share than insects do.

A few specific enemies are worth knowing because they shape whole industries:

Apple scab (Venturia inaequalis) is why apple growing in humid climates is chemically intensive. Spores overwinter in fallen leaves, release during spring rain, and infect leaves and fruit, producing corky brown lesions that make fruit unsaleable. A wet spring in a susceptible orchard means a spray roughly every seven to ten days. Scab-resistant varieties exist (Liberty, Topaz, and others carrying the Vf resistance gene) and have struggled commercially because consumers buy the varieties they know.

Late blight (Phytophthora infestans), the organism behind the Irish potato famine, is still the most destructive potato disease worldwide. It can destroy a field in under two weeks in the right weather, which is why potato growers spray on forecast-driven schedules tied to humidity and temperature.

Codling moth is the worm in the apple, literally: the larva tunnels to the core. It is the reason apples are sprayed at specific times, and it is also one of the great successes of alternative control, via mating disruption: dispensers flood the orchard with synthetic female pheromone so males cannot locate real females. This works, is target-specific, harms nothing else, and is widely used.

Varroa mite attacks honeybees, not crops, and matters because of pollination (Chapter 4). It is a major contributor to colony losses worldwide.

What a pesticide is, and the classes worth knowing

In short: "Pesticide" is an umbrella covering insecticides, fungicides, herbicides, and more, spanning enormous ranges of toxicity and selectivity.

The word covers any substance used to kill or control a pest. In practice, more than half of global pesticide use by volume is herbicides.

Insecticides, by generation

The direction of travel is genuine: newer insecticides are generally more selective and less acutely toxic to mammals. That progress is real and it is not the end of the story, because selectivity for mammals is not the same as harmlessness to ecosystems. Neonicotinoids are the clearest case: they are strikingly safe for people compared with what they replaced, and they are strikingly bad for pollinators, because they are systemic and turn up in pollen and nectar.

Fungicides

Two broad families. Protectants (copper, sulphur, mancozeb, captan) coat the leaf surface and stop spores germinating; they must be reapplied after rain and resistance to them develops slowly because they hit many targets at once. Systemics (strobilurins, triazoles, SDHIs) are absorbed and move within the plant, work after infection has begun, and are far more prone to resistance because they hit a single molecular target.

Copper deserves a note, because it is the fungicide approved for organic production and it is not benign. Copper does not degrade. It accumulates in soil indefinitely, is toxic to earthworms and soil microbes at accumulated levels, and long-used organic vineyards and orchards in Europe have measurable copper build-up. The EU has been progressively restricting its use, including in organic systems. This is a good illustration that "approved for organic" is a regulatory category, not a toxicological one.

Herbicides

Glyphosate is the most used herbicide on Earth and the most argued about, so it is worth stating what is actually known. It blocks an enzyme (EPSP synthase) in a metabolic pathway that plants and many microbes have and animals do not, which is the source of its low acute mammalian toxicity. On cancer: the WHO's International Agency for Research on Cancer classified it in 2015 as "probably carcinogenic to humans" (Group 2A), a hazard classification meaning the agency judged there is evidence it can cause cancer under some conditions. Regulatory agencies including the EPA, EFSA, and others conducting risk assessments, which ask whether it does so at realistic exposures, have concluded it is unlikely to pose a cancer risk at expected dietary exposures. Both statements can be true because they answer different questions, a distinction covered properly in Chapter 97. The occupational exposure question, for people who mix and spray it for a living, is genuinely less settled than the dietary one.

The larger and less contested problem with glyphosate is agronomic: two decades of heavy reliance on it, especially with herbicide-tolerant crops, has produced glyphosate-resistant weeds across tens of millions of hectares, which has pushed farmers back toward older and more toxic herbicides.

Resistance: the treadmill nobody escapes

In short: Any control applied uniformly and repeatedly selects for the organisms that survive it, and there is no way around this except deliberate variation.

Spray a field and you kill 99 percent of the target. The 1 percent that survives may carry a mutation that made them survive. They breed. Next generation, the survivor fraction is larger. Repeat for a decade and the product no longer works.

This is not a hypothetical: over 500 insect species have documented resistance to at least one insecticide, over 250 weed species to at least one herbicide, and fungicide resistance can appear within two or three seasons for single-target systemics.

Resistance management is therefore a formal discipline with published rules:

• Rotate modes of action, not just brand names. Products carry a FRAC/IRAC/HRAC group number identifying their biochemical target, and the rule is to alternate group numbers between applications.
• Use mixtures of two modes of action, so a resistant individual still meets a lethal compound.
• Leave refuges. In Bt maize and cotton, growers are required to plant a percentage of non-Bt crop nearby so susceptible insects survive there and dilute resistance genes by mating with survivors. This is a genuinely clever piece of applied evolutionary biology and it has measurably slowed resistance where it has been enforced.
• Do not rely on chemistry alone. Rotation, resistant varieties, sanitation, and biological control all reduce selection pressure.

Integrated Pest Management: what good practice actually looks like

In short: Monitor, decide with thresholds, use the least disruptive tool that works, and treat spraying as one option rather than the default.

Integrated Pest Management (IPM) is the mainstream professional standard, and it is much less about ideology than about not wasting money.

1. Prevention. Resistant varieties, rotation, sanitation (removing infected material), healthy soil, and habitat for beneficial insects.
2. Monitoring. Pheromone traps, sticky traps, weekly scouting, degree-day models that predict when a pest will hatch, and disease forecasting models based on leaf wetness and temperature. This is where the modern effort goes.
3. Thresholds. Do not treat because a pest exists; treat when its density crosses the level at which the damage would exceed the cost of treatment. Some aphids in a field are food for the ladybirds you want to keep.
4. Intervention, least disruptive first. Biological control (releasing predatory mites, parasitic wasps, or nematodes), mating disruption, physical barriers and netting, and then selective chemistry, with broad-spectrum products last.
5. Evaluation. Did it work, and what does that mean for next year.

Biological control is underrated by the public and heavily used commercially, especially in glasshouses, where a Dutch pepper grower may release predatory mites (Phytoseiulus against spider mite), parasitic wasps (Encarsia against whitefly), and predatory bugs (Orius against thrips) as standard practice, with almost no insecticide. It works there because the enclosed environment lets you establish predator populations reliably.

What ends up on your food

In short: Residues are regulated by a chain of limits with large built-in safety factors, monitored by testing programmes, and in the overwhelming majority of samples they are far below the legal limit.

Here is the actual chain, because it is more structured than most people realise.

1. Toxicology establishes a no-effect level. Animal studies determine the highest daily dose producing no observed adverse effect (the NOAEL), typically in two species over a lifetime, plus reproductive and developmental studies.

2. Divide by safety factors. The NOAEL is divided by 100 as standard: a factor of 10 for possible differences between animals and humans, and another 10 for variation between humans, with additional factors applied for particular concerns such as effects on developing animals. The result is the Acceptable Daily Intake (ADI), the amount a person could consume every day for life without expected harm.

3. Field trials set the residue limit. Supervised trials applying the product exactly as the label directs measure what residue remains at harvest. The Maximum Residue Level (MRL) is set at or slightly above the highest residue seen in those correct-use trials.

This is the part that is most misunderstood. An MRL is a trading and good-practice standard, not a safety threshold. Exceeding an MRL means the product was probably used incorrectly. It does not mean the food is dangerous, and in most exceedances the residue remains far below the ADI. The two numbers are often separated by a factor of hundreds or thousands.

4. Pre-harvest intervals enforce the gap. Every product label specifies a minimum number of days between last application and harvest, calculated so residues decay below the MRL. This is legally binding and it is the single most effective control on what reaches you.

5. Monitoring programmes test the food. The EU tests tens of thousands of samples a year across member states; the US FDA and USDA run parallel programmes. Published results are consistent year over year: roughly 95 to 99 percent of samples are within legal limits, about half of samples have no detectable residues at all, and dietary risk assessments based on the measured residues put actual consumer intake at a small fraction of the ADI.

Don't be confused: "detected" and "unsafe" are not related. Analytical chemistry can now detect parts per billion, which is roughly one second in thirty years. A headline saying residues were "found on 60 percent of samples" is describing the sensitivity of the instrument, not the safety of the food. The number that matters is the residue as a percentage of the ADI, and it is almost always well under one percent.

The "Dirty Dozen" list, honestly

The annual list of most-residue-bearing produce published by the Environmental Working Group is widely circulated and is genuinely misleading in one specific way: it ranks by frequency and number of detections, not by the toxicological relevance of the amounts found. Analyses that model actual exposure, including work published in the Journal of Toxicology, have found that consumer exposures to the listed pesticides on those foods were typically several orders of magnitude below levels of concern, and that substituting organic for conventional would not meaningfully change that exposure.

The reason to care about the list is nonetheless real, and it is this: if it persuades anyone to eat fewer strawberries or less spinach, it does clear harm, because the established benefit of eating fruit and vegetables vastly outweighs any residue risk that has been demonstrated. If you can afford organic and prefer it, that is a legitimate choice for environmental and farmworker reasons. Choosing to eat less produce because of residues is the one response the evidence does not support.

What washing and peeling actually do

• Rinsing under running water for 30 seconds removes a meaningful share of surface residues, plus dirt and a good part of any surface microbes. Rubbing with your hands helps more than the water alone.
• Baking soda solution (about 1 percent) has been shown in laboratory work to remove more surface residue than plain water for some compounds, because it speeds their breakdown. It is not necessary, but it is not nonsense.
• Vinegar, salt water, and commercial produce washes are not clearly better than water for residues, though vinegar solutions do reduce surface bacteria somewhat.
• Peeling removes more residue than any wash, and also removes fibre and a disproportionate share of some nutrients and polyphenols. For most produce that is a bad trade.
• Systemic pesticides cannot be washed off because they are inside the plant tissue. This is a genuine limitation of washing and the reason pre-harvest intervals matter more than washing does.
• Food safety is the better reason to wash. E. coli, Salmonella, and Listeria on produce cause far more documented illness every year than pesticide residues have ever been shown to. Melons are the notorious case: the rind is not eaten, so people do not wash it, and the knife carries surface bacteria straight into the flesh.

The costs that are real

Nothing above should be read as a claim that pesticides are harmless. The genuine problems are these, and they are not primarily about your dinner.

Applicator and farmworker exposure is where documented human harm concentrates. Acute poisonings number in the hundreds of thousands globally each year, overwhelmingly in low-income countries where protective equipment and enforcement are weak, and paraquat in particular has caused a very large number of deaths. Chronic exposure associations, including some cancers and Parkinson's disease, are studied primarily in agricultural cohorts for this reason.

Pollinators. Neonicotinoid effects on bees are well documented in both laboratory and field studies, which is why the EU banned the main three for outdoor use in 2018. Pollinator decline has multiple causes (habitat loss, varroa, disease, climate) and pesticides are one of them.

Aquatic systems. Pyrethroids are extremely toxic to fish and aquatic invertebrates at very low concentrations. Runoff and spray drift into watercourses is a real and measurable harm.

Ecosystem simplification. Broad-spectrum insecticides kill the predators of the pest as well as the pest, which can cause a secondary pest outbreak where a previously insignificant species explodes because its natural enemies are gone. This is a classic and well-documented failure mode of calendar spraying.

The bottom line

• Pests, weeds, and disease take a fifth to two fifths of world crop production even with modern control. The counterfactual to pesticide use is not the same food with fewer chemicals, it is substantially less food.
• Pesticide classes differ enormously. The trajectory from arsenicals to DDT to organophosphates to modern selective chemistry is real progress in human toxicity, with ecological costs that are still being reckoned with.
• Resistance is inevitable and is managed by rotating modes of action and not relying on chemistry alone.
• Residues on retail produce are monitored heavily and sit far below the levels used to set safety limits. "Detected" says more about instruments than about risk.
• The documented human harm from pesticides falls overwhelmingly on the people who apply them, not on the people who eat the food.

Sources and notes

Global crop loss estimates follow Oerke's analyses in the Journal of Agricultural Science and subsequent FAO assessments. Pesticide class properties, resistance mechanisms, and mode-of-action group numbering follow IRAC, FRAC, and HRAC published classification schemes. The residue regulatory chain (NOAEL, ADI, MRL, pre-harvest interval) follows EFSA and Codex Alimentarius procedures. Monitoring results are from the EFSA annual EU report on pesticide residues in food and the USDA Pesticide Data Program. The Dirty Dozen critique follows Winter and Katz, Journal of Toxicology, 2011. Glyphosate's divergent classifications are IARC Monograph 112 (2015) versus EFSA and EPA risk assessments. Neonicotinoid effects on pollinators follow EFSA's 2018 conclusions and the field studies underlying the EU outdoor ban. Copper accumulation in organic systems follows EFSA's copper review and EU restriction decisions. Occupational poisoning figures are from WHO estimates and the paraquat literature. Washing and residue removal follows Yang et al., Journal of Agricultural and Food Chemistry, 2017 (baking soda).

Open questions. The long-term health effects of chronic low-level pesticide exposure in agricultural workers, particularly associations with Parkinson's disease and some cancers, remain under active study and are not settled. Whether IPM adoption has actually reduced total pesticide load, as opposed to shifting which compounds are used, is poorly measured at national scale.

👉 Next: organic, regenerative, and what the labels actually mean.