Meet the Mosquito With a Big Appetite—for Other Mosquitoes

By Ed Ricciuti

It sounds like classic cloak and dagger. A band of assassins is fielded to infiltrate and kill enemies that are spreading dangerous diseases among the general population. That’s the idea behind using mosquito against mosquito, a tactic described in a new article published this month in the Journal of Insect Science.

Scientists in Harris County, Texas—Houston is county seat—have come up with what could lead to a better way of producing, in their words, “lab-reared, native mosquito assassins,” Toxorhynchites rutilus septentrionalis, that prey on other mosquitoes, notably Aedes species, which spread human disease viruses such as dengue, Zika, chikungunya, and yellow fever. Previously, the tactic has been used infrequently, because it is relatively inefficient. For one, the predatory mosquitoes are not available commercially so must be raised in house. And it is difficult to produce enough of them in a laboratory to make a dent in disease vector numbers after release in the environment. But this new research may change all that.

Instead of raising killer mosquitoes all together in batches, which has been the practice and inefficient, the Texas researchers, led by Anita Schiller, director of mosquito bio-control development at Harris County’s Biological Control Initiative, increased production by rearing them individually. Although it may sound labor intensive, the researchers say their method is more efficient because, when reared all together (the old way) Tx. rutilus larvae cannibalize one another, sharply reducing the output. The researchers also sped up production by feeding the Tx. rutilus larvae foods that are easier to obtain and dispense and promote faster development than the traditional staple of prey mosquito larvae. Schiller and colleagues’ report is published as part of a new Journal of Insect Science special collection on mass rearing of high-quality insects.

Larval Tx. rutilus voraciously hunt and gorge on the larvae of disease-vector mosquitoes sharing the same water. One of them can consume up to 5,000 prey larvae before it matures, which can take several weeks to six months. While the larvae are fierce cannibals, Tx. rutilus adults—large mosquitoes with a wingspan of almost half an inch and legs that would overlap a U.S. quarter—feed peaceably on nectar from flowers. They need sugars from nectar to produce eggs.

Tx. rutilus is sometimes known as the “predatory tree-hole mosquito” because it lays its white, football-shaped eggs in tree cavities and other containers for water. People provide it with nest sites aplenty in the form of flower pots, rain barrels, cans, pet watering bowls, pails, tires, and similar receptacles. Aedes species use the same type of breeding sites, which makes tree-hole mosquitoes ideal for targeting them.

It is not always so easy to match a predator used for mosquito control to prey because mosquito biology can vary considerably between species. “Mosquitoes are as variable as birds,” says Schiller. “Some prefer avian blood meals, others the blood of small mammals, others take it from reptiles, and some do not care if the meal comes from animal or human.” A handful of mosquitoes, like Tx. rutilus, feed on something other than blood.

Reproductive behavior also varies. Some species lay eggs in cavities that eventually fill with rain. Some lay eggs in permanent water while others attach them to floating plants.

“With so many places where biting mosquitoes can breed, it is easy to see how a single control measure is at best complicated and in reality is unlikely to be developed,” says Schiller.

The Tx. rutilus subspecies used in the research was once widespread in Harris County but has declined because the greenery it inhabits disappeared under the concrete of expanding Houston. Some of that vegetated habitat has returned as landscaped yards and parks of maturing neighborhoods.

To breed the predators, Schiller and colleagues housed adults in large mesh cages, each holding 650 individuals of both sexes, with water-filled cups as repositories for eggs laid by the females. The scientists scooped up bunches of eggs, each about the size of a sand grain, with a small net and divided them up in smaller groups among small, water-filled containers. Each group of eggs had to be watched for up to two days as they hatched, to prevent larvae from cannibalizing one another. Developing larvae were removed to individual rearing cells, each holding almost an ounce of water. Picking up larvae with a tiny, 0.1-ounce pipette was delicate work, but it took an experienced hand, on average, 15-seconds per larvae, or 1,000 of them processed in four hours.

“Filling the rearing trays with water and harvesting, rinsing, and pipetting each larva into its own cell—it’s a bit of a careful and precise art,” says Schiller, “but one which my team has mastered beautifully.” Raising and preparing the tiny, squishy worms fed to the Tx. rutilus larvae was similarly challenging, given the fact that they had to be removed from rearing containers and rinsed before they were fed to the larvae.

The type of food provided to larvae was matched larval behavior and size as the young predators developed through instars, stages in development separated by molts. “Early instar Tx. rutilus larvae are astute hunters and will actively swim toward perceived prey, and the feeding response is best triggered by live foods,” says Schiller. Victims are seized by the larvae’s hook-like mandibles and consumed. Young larvae need live food that triggers their feeding response. By the fourth, and last, instar, they will also feed on immobile food, so do not require it to be live.

During the first instar, the food provided was the microworm Panagrellus, a minuscule nematode. Later, the prey was dero worms, commonly known as “microflex” and often fed to fish in home aquaria. Both were raised in the laboratory, while the food for late third and fourth instars, bloodworms—again, a popular aquarium fish food—were bulk purchased frozen.

Pupae that form from the Tx. rutilus larvae were netted and placed in new containers inside cages.

Emergent adults were kept in cages for a few days, allowing them to mate and eggs to begin developing, and then released in a suitable environment. In practice, the Tx. rutilus mosquito can be released as a control tool at any life stage, even as eggs placed in containers where problem mosquitoes breed.

During the summer of 2017, the researchers produced about 1,000 adults weekly, releasing 300 gravid females each week. Then disaster struck. Hurricane Harvey wrecked the rearing facility, although the staff managed to evacuate the mosquito colony and equipment to a room in a community center.

“We set out to marry classical, traditional biological control agents within integrated vector control programs, coming up with ‘one-two punches’ to reduce problem populations,” says Schiller.

Once the laboratory was again operational, a repeat study began in hopes that continued research will lead to full-scale production of predatory mosquitoes ahead of pesticide treatment in IPM programs.

Ed Ricciuti is a journalist, author, and naturalist who has been writing for more than a half century. His latest book is called Bears in the Backyard: Big Animals, Sprawling Suburbs, and the New Urban Jungle (Countryman Press, June 2014). His assignments have taken him around the world. He specializes in nature, science, conservation issues, and law enforcement. A former curator at the New York Zoological Society, and now at the Wildlife Conservation Society, he may be the only man ever bitten by a coatimundi on Manhattan’s 57th Street.