The ant swarm after baiting is the bait working, and pest crews have been reading it backwards

A laboratory at UC Riverside built transparent Argentine ant colonies, fed them two of the most common bait insecticides, and watched what the workers did before they died ^[1][6]. The finding that matters operationally: the cue almost everyone uses to judge whether a bait worked, namely how many ants they can see, moves in opposite directions depending on the active ingredient, and in one of the two cases it moves the wrong way entirely ^[14][15].

The work comes from the lab of UCR entomology professor Dong Hwan Choe and was published in the journal Insects ^[1]. It compared thiamethoxam and boric acid against Argentine ants, the species most commonly found in homes and gardens in Southern California ^[2]. Both killed most of the workers, but at different rates: thiamethoxam wiped out all workers in about seven days, while boric acid killed roughly 90 percent in about two weeks ^[3]. So boric acid takes about twice as long and still leaves roughly one worker in ten alive ^[4].

The behaviour in between is the point. In healthy colonies, the ants spread through the available nesting space, most staying inside to tend the queen and brood while some workers forage outside ^[7]. After thiamethoxam, many more ants became much more active outside the nest, roaming the foraging area before dying ^[8]. Choe attributes this to the chemical's action on the nervous system: "It's like they're all high on nicotine," he said, adding that "their nervous systems have been compromised by the insecticide" ^[9]. Thiamethoxam, a neonicotinoid, mimics some of nicotine's effects on insect nerves, overstimulating them and eventually causing paralysis ^[10][5].

Boric acid did close to the opposite. Affected ants stopped foraging altogether and gathered tightly inside the nest, particularly near water ^[11]. Boric acid damages the midgut lining, which handles digestion and nutrient absorption, and in some ants damages the excretory organ that regulates water and waste; Choe said that could explain the clustering near water, but that more research is needed to establish it ^[12][13].

That gives two failure modes for anyone judging a bait by eye. Where boric acid is used, foraging stops and the colony can look eliminated, only for survivors to reappear later ^[14]. Where thiamethoxam is used, poisoned workers wander into sinks, garages and windowsills for a day or two before dying, so the infestation looks like it is getting worse at exactly the moment the dose is spreading ^[15].

There is a formulation angle too. Workers share food through the colony, which dilutes the dose any individual receives, and Choe said lower doses of thiamethoxam could potentially prolong the roaming behaviour ^[16]. That is a direct trade-off between how a bait looks in week one and how far the active ingredient travels.

Neither chemical solves the underlying problem. There are no known cases of insecticide resistance in ants, but they remain hard to control, and the treatments may not kill the queen, which is why a single bait application often buys only temporary relief ^[17][18].

What to watch: whether bait manufacturers adjust dosing now that the behavioural signature is documented, since the roaming effect is dose-dependent by Choe's account ^[16]. Watch also for confirmation outside glass-walled lab nests, where foraging space and water sources are not laid out for the camera ^[6]. And treat the water-clustering mechanism as unsettled until someone tests it directly, which Choe himself flags ^[13].