Why do some ant workers still wander at night?

As far as I have seen, the vast majority of workers in colonies of diurnal, non-nomadic species return to their nests and remain there at night. The colonies seem to forage largely during the day. However, a few individuals of said colonies would still wander far from the nest at night. I don't believe they are lost because they still seem to be following pheromone tracks.

Why do they do that? What drives them to still come out at night and not remain in the nest like most other workers do? Are they any different in terms of caste? Are they always the same individuals? Is there a selection process to decide who will wander off at night? Are they somehow invalid and can't tell day from night? Is that somehow evolutionarily advantageous (e.g. having a few workers take some risk at night, maintaining the pheromone tracks throughout the night so that the other workers don't have to start all over the next morning)?

Someone (jokingly) told me those ants must still be hungry, but I don't immediately buy that explanation because my understanding is that most species will store food in the nest in one form or another anyway, or they just share food by trophollaxis.

Swarm behaviour

Swarm behaviour, or swarming, is a collective behaviour exhibited by entities, particularly animals, of similar size which aggregate together, perhaps milling about the same spot or perhaps moving en masse or migrating in some direction. It is a highly interdisciplinary topic. [1] As a term, swarming is applied particularly to insects, but can also be applied to any other entity or animal that exhibits swarm behaviour. The term flocking or murmuration can refer specifically to swarm behaviour in birds, herding to refer to swarm behaviour in tetrapods, and shoaling or schooling to refer to swarm behaviour in fish. Phytoplankton also gather in huge swarms called blooms, although these organisms are algae and are not self-propelled the way animals are. By extension, the term "swarm" is applied also to inanimate entities which exhibit parallel behaviours, as in a robot swarm, an earthquake swarm, or a swarm of stars.

From a more abstract point of view, swarm behaviour is the collective motion of a large number of self-propelled entities. [2] From the perspective of the mathematical modeller, it is an emergent behaviour arising from simple rules that are followed by individuals and does not involve any central coordination. Swarm behaviour is also studied by active matter physicists as a phenomenon which is not in thermodynamic equilibrium, and as such requires the development of tools beyond those available from the statistical physics of systems in thermodynamic equilibrium. In this regard, swarming has been compared to the mathematics of superfluids, specifically in the context of starling flocks (murmuration). [3]

Swarm behaviour was first simulated on a computer in 1986 with the simulation program boids. [4] This program simulates simple agents (boids) that are allowed to move according to a set of basic rules. The model was originally designed to mimic the flocking behaviour of birds, but it can be applied also to schooling fish and other swarming entities.

Edible Insects Farming: Efficiency and Impact on Family Livelihood, Food Security, and Environment Compared With Livestock and Crops

Weaver Ants

Weaver ants are used as food by some local communities, and they are particularly preferred by traditional healers due to their medicinal uses. Also, their eggs are a popular delicacy (condiment). The ant workers construct nests by weaving leaves together using larval silk. The colonies can be maintained in host plants (particularly mango trees) if protected from predators and substrate and water are provided, as the ants need it to produce acetic acid ( Fig. 4.2 ). In order to accelerate multiplication, highways are made from jute or cotton woven rope or rattan cane. Ants are harvested once a year by using a long bamboo pole with a bag or basket attached with strings to the tip. A hole is poked into the nest with the tip of the pole and it is shaken. In this way, larvae and pupae fall down into the bag. The content of the bag is poured into a big plastic container in which some rice and tapioca from cassava (Manihot esculenta) flour is added to prevent the ants from climbing and escaping. A branch is inserted into the container so that adult ants can climb up. Then, the branch is whipped against a tree to release them and the larvae and pupae left in the bag are collected for human consumption. Ants can be farmed in home gardens also by feeding them with food scraps and water ( Hanboonsong et al., 2013 ).

About 300–400 g of larvae and pupae and about 2–8 kg of adults can be collected every day. On average, a collector can earn US $8–15 (250–500 ThB) per day. With current market demand for fresh produce in Thailand, sellers can earn more from farmed insects than farm conventional crops of rice or cassava ( Hanboonsong et al., 2013 ).

Constraints: Because of the market demand, ant populations are decreasing in the wild. This trend has a negative impact on the ecology, as weaver ants are predators of crop pests. Collecting queens to start new colonies is a hard task because queens are found in small nests at the highest point of the tree, where they are difficult to reach.

Why do some ant workers still wander at night? - Biology

The Florida carpenter ant complex is comprised of several species, two of which are common around structures: Camponotus floridanus (Buckley) and Camponotus tortuganus (Emery). These bicolored arboreal ants are among the largest ants found in Florida, making them apparent as they forage or fly indoors and out.

Figure 1. Major workers of Camponatus sp. Photograph by Rudolf H. Scheffrahn, University of Florida.

In a survey of common urban pest ant species covering four metropolitan areas of Florida (Daytona-Orlando, Tampa Bay area, Sarasota-Ft. Myers, and the greater Miami area), Klotz et al. 1995 found that infestations of Florida carpenter ants accounted for approximately 20% of all ant complaints by homeowners. Klotz et al. (1995) found only a few instances where other ants, including imported fire ant (Solenopsis invicta Buren), crazy ant (Paratrechina longicornis (Latreille), ghost ant (Tapinoma melanocephalum (Fabr.), and pharaoh ant (Monomorium pharaonis (L.)) were more frequently encountered in buildings than carpenter ants.

During the flight season, carpenter ants can often be found in alarming numbers. Sometimes homeowners are concerned about damage to the structural integrity of their homes, which they sometimes incorrectly learn, is caused by Florida carpenter ants. However, unlike the wood-damaging black carpenter ant, Camponotus pennsylvanicus (DeGreer), found in Florida's panhandle and a few other western U.S. species, Florida carpenter ants seek either existing voids in which to nest or excavate only soft materials such as rotten or pithy wood and Styrofoam. Other concerns are that these ants sting (they do not) and bite (they do).

In recent years, a small and exotic daytime-foraging carpenter ant, C. planatus Roger, has become common in many parts of central and southern Florida.

Figure 2. Camponotus floridanus is found widely distributed throughout Florida and some neighboring states, while C. tortuganus is limited to central and southern portions of Florida. The ratio of C. floridanus to C. tortuganus is about 2:1 in south Florida. There are several other Camponotus species found in Florida, however, these are rare or usually not associated with buildings. These species include Camponotus caryae (Fitch), C. castaneus (Latreille), C. decipiens Emery, C. discolor (Buckley), C. impressus (Roger), C. nearcticus Emery, C. pylartes Wheeler, C. sexguttatus (Fabr.), C. snellingi Bolton, and C. socius Roger.

Description (Back to Top)

The antennae of Florida carpenter ants are 12-segmented, with the terminal segment being slightly elongated and bullet-shaped, and without a club. There is a circular ring of hair at the end of the abdomen. The waist consists of one petiolar segment. The antennal scape is flattened basally and broad throughout. Workers vary in size, ranging from 5.5 to 11 mm in length. Smaller workers are called minors while larger workers are called majors. Winged females (alates) are the largest caste reaching up to 20 mm in length. There is no sting, but workers can bite and spray formic acid for defense. The thorax is evenly convex a key characteristic of carpenter ants. The thorax and head are ash brown to rusty-orange and the gaster is black. Body hairs are abundant, long, and golden. Male reproductives are much smaller than queens with proportionally smaller heads and larger wings. Specific characters for C. floridanus include legs and antennal scapes with numerous long, coarse brown to golden erect hairs, shorter than those on the body. For C. tortuganus, specific characters include a major worker with head longer than broad tibia of all legs and antennal scapes without erect hairs and thinner than C. floridanus and paler with less color contrast.

Figure 3. Female alate (reproductive) of the Florida carpenter ant, Camponatus floridanus (Buckley). Photograph by Rudolf H. Scheffrahn, University of Florida.

Figure 4. Male and female reproductives of the Tortugas carpenter ant, Camponatus tortuganus (Emery). Photograph by Rudolf H. Scheffrahn, University of Florida.

Figure 5. Worker and male reproductives of the Tortugas carpenter ant, Camponatus tortuganus (Emery). Notice that worker's head is longer than broad. Photograph by Rudolf H. Scheffrahn, University of Florida.

Life Cycle (Back to Top)

As with all members of the Order Hymenoptera, carpenter ants develop by complete metamorphosis, going through stages of the egg, larva, pupa, and adult worker or reproductive. Larvae are maggot-like, and pupae reside in silk cocoons and are often mistaken for eggs. Winged reproductives fly in the evening or night during the rainy season (May through November). Typically, a single queen, fertilized by a smaller short-lived male, will start a new colony, caring for her first brood of larvae until they develop into workers, which then begin to forage for food. Workers then care for the queen and ensuing brood. The colony will continue to grow and populations may reach several thousand workers. When the colony is two to five years old, depending on environmental conditions, new winged reproductives, or alates, will usually be sent out. Alates (winged reproductives) are observed from spring to fall, depending on the area and environmental conditions. Queenless satellite nests are often founded within 20 to 100 feet of a mature nest. Proximity of nests can lead to fighting among neighboring colonies.

Figure 6. Florida carpenter ant, Camponotus floridanus (Buckley), dealate queen tending brood. A dealate is a reproductive that has shed its wings. The adults that emerge from this brood will be small ants called minums, and they take over the queen's brood-tending functions so she can concentrate on laying eggs. The brood that emerge after the minums should be normal sized worker ants. Photograph by John Warner, University of Florida.

Figure 7. Adult workers, and brood (larvae and pupae) of the Florida carpenter ant, Camponatus floridanus (Buckley). Photograph by Rudolf H. Scheffrahn, University of Florida.

Figure 8. Florida carpenter ant workers, Camponatus floridanus (Buckley), from neighboring colonies fighting. Photograph by Rudolf H. Scheffrahn, University of Florida.

Pest Status (Back to Top)

Carpenter ants are one of the most common indoor insect pests in Florida. Alarmed homeowners often see these ants foraging (especially at night) and either attempt to control the ants with spray insecticides or call their local pest control operator (PCO). PCOs report going to innumerable homes to speak with frantic homeowners who have failed to control foraging or flying carpenter ants. Many PCOs and seasoned entomologists have resigned to failure after not being able to treat hidden nesting sites, sometimes after years of trying! An experienced pest control operator, or a determined homeowner, can usually follow a trail of ants back to the ants' nesting site and treat it (see Management below).

Complaints are numerous during the spring swarm season, usually between April and June, when winged reproductives are often found in homes in such places as along window ledges and near sliding glass doors. It is common to mistake winged ants for winged termites. Differences between ants and termites are given below:

  • Elbowed antennae
  • Fore wings larger
    than hind wings
  • Waist constricted
  • Beaded antennae
  • Fore and hind
    wings of equal size
  • Waist broad

Foraging and Feeding (Back to Top)

Florida carpenter ants tend to forage at night. The peak foraging hours are just before sunset until two hours after sunset, then again around dawn. Foraging proceeds in very loosely defined trails or by individual ants that seem to wander aimlessly. These ants have a fondness for sweets and can be found in campgrounds near soda machines and other areas where sweets are readily accessible. Similarly, they are fond of sweet floral nectars and honeydews produced by sucking insects, especially aphids, scales, and mealybugs. Trees and shrubs which are infested with these honeydew-producing insects or produce nectars will have ants wandering in all directions over leaf surfaces, up and down the stems and trunk. Carpenter ants will also seek out other insects, both living and dead, for food.

Carpenter ants foraging in homes can be in search of sweets or moisture, or even new nesting sites, especially in kitchens and bathrooms, or other rooms that have water leaks from plumbing or leaks around doors and windows. Otherwise, they might simply be trailing from an interior nest to an exterior food source.

Carpenter ants, like many other ants, will trail along wires or cables that may be attached to homes and serve frequently as access routes for them to enter attics and other above ground areas. Tall trees touching structures cause "bridges" which provide foraging access into buildings.

Nest Sites (Back to Top)

Carpenter ants seem to prefer voids for nesting which have these characteristics:

  • Close to moisture and food sources
  • Safe from predators such as birds and lizards
  • Safe from flooding, heat, and other environmental stresses
  • Easily accessible (for them, but inaccessible for the Pest Control Manager!)

They will hollow out wood softened by moisture and/or fungi to create nests. This wood can be in tree stumps or dead tree limbs, or in any part of a structure having damaged wood. They will not excavate nesting galleries in sound wood. Bits of debris, called frass, are often ejected from nesting sites. Frass consists of bits of excavated materials and pieces of dead insects, including carpenter ants.

Common exterior nesting sites include: old drywood termite galleries and wooden objects that have had previous damage from other organisms including insects or fungi rotting tree stumps and tree holes or crotches between limbs under old leaf petioles in palms, especially in and around the inflorescence of coconut palms under bark, in roots of trees, especially citrus trees in old wooden fences, sheds, old wooden decks, bamboo poles (even thin or short pieces) or tree supports, debris of almost any kind, coconuts left on the ground, under mulch, inside logs or wooden borders in gardens, railroad ties, old shoes, in voids in ceramic or concrete decorations, walls or support pillars, in expansion joints either not filled in or filled with rubbery materials, under stones, in home exterior coverings, especially wood panels, and so on.

Common interior nesting sites include: wall voids (especially walls that have moisture seepage), under attic insulation and usually near the eaves where they are very difficult to reach, under bath tubs, very common under windows and door frames which have moisture intrusion from rain or sprinklers, around skylights, in boxes or paper bags, in closets which are not often used, under appliances, especially dish washers, in flat roofs (one of the most difficult problems due to lack of adequate access), behind wood panels, in wood furniture, cracks in floors, under bathroom fixtures, and many other places! Carpenter ants are sometimes found in electrical boxes, such as fuse, meter, or timer boxes or appliances. Unusual nest sites have included a computer printer, a radio, and a pay phone. Also check hollow supports of patio screens or voids in patio ceilings.

Management (Back to Top)

Direct treatment of nesting sites is recommended because these sites harbor the brood, queen, and a bulk of the workers and winged reproductives, however, finding the nest sites can be difficult. A small amount of insecticidal dust or spray applied directly to the nest area is usually successful. Excessive treatment can become repellent, actually causing the nest to move to another location if the dust or spray is applied near-to but not directly on the nest.

Observation of foragers entering voids is the best means of finding the nest. Watch for trailing ants at their peak nocturnal foraging hours and follow them. Look for areas of higher ant population density indicating closer proximity to the nest. Placing a few drops of sugar water, honey or dead insects along a trail can cause other nestmates to be recruited to the area. Try to follow the foragers back to the nest and then treat the nest.

Figure 11. Worker of the Florida carpenter ant, Camponatus floridanus (Buckley), entering a void. Photograph by Rudolf H. Scheffrahn, University of Florida.

There are many situations in which the nest is not accessible, or cannot be found. In those cases use one of the baits made for carpenter ants, and follow the label directions. Usually, baits are simply placed along the trail and foragers bring the toxic baits back to the nest where food and toxicant are shared via trophallaxis (communal food sharing). Carpenter ants are finicky eaters and tend not to recruit in large numbers to any food source thus decreasing the efficacy of insecticidal baits. Residual sprays in foraging areas can also be helpful. Be sure to spray areas where ants are feeding, such as trees and shrubs. A systemic insecticide can help control aphids and other honeydew producers to reduce food for the carpenter ants.

Eliminate "bridges" caused by trees and shrubs touching house exteriors. If wires or power cables are being used as bridges, it may be possible to have a professional treat the wires or areas where the wires attach to the structure. There are a number of "pest barrier" substances available that are sticky and can be used on tree trunks and other places to stop ants from passing. Caulking exterior openings and weather stripping may also aid in control. Read and follow label instructions and precautions before using any insecticide.

Selected References (Back to Top)

  • Deyrup MA. 1991. Exotic Ants of the Florida Keys (Hymenoptera: Formicidae). Proceedings of the 4th Symposium on the Natural History of the Bahamas, June 7-11, 1991. 21 pp.
  • Deyrup MA, Carlin N, Trager J, Umphrey G. 1988. A Review of the Ants of the Florida Keys. Florida Entomologist 71: 165-6.
  • Klotz JH, Mangold JR, Vail KM, Davis Jr. LR, Patterson RS, 1995. A Survey of the Urban Pest Ants (Hymenoptera: Formicidae) of Peninsular Florida. Florida Entomologist 78: 112-3.
  • United States Department of Agriculture, 1945. House-Infesting Ants of the Eastern United States, Their Recognition, Biology, and Economic Importance. Technical Bulletin No. 1326, Agricultural Research Service. 61-74.
  • Vail K, Davis L, Wojcik D, Koehler PG, Williams D, 1994. Structure-Invading Ants of Florida. Cooperative Extension Service, University of Florida, Institute of Food and Agricultural Sciences. SP164. 13-14.
  • Wheeler WM. 1932. A List of the Ants of Florida with Descriptions of New Forms. New York Entomological Society, Vol. XL, No. 1. 13-16.

Authors: John Warner and Rudolf H. Scheffrahn, University of Florida
Photographs: Rudolf H. Scheffrahn and John Warner, University of Florida
Video: John Warner, University of Florida
Web Design: Don Wasik, Jane Medley
Publication Number: EENY-272
Publication Date: July 2002. Revised: September 2004. Reviewed: December 2017. Latest Revision: April 2021.

An Equal Opportunity Institution
Featured Creatures Editor and Coordinator: Dr. Elena Rhodes, University of Florida

Insects in the City

Odorous house ants are attracted to honey and other sweets.

Ants can be a challenge to identify without the proper equipment and experience. In many cases, the best way to confirm the identity of an ant is to enlist the help of a pest management professional. Nevertheless, it is possible to identify some of the most common species of household ants without a microscope.

The following pictures and descriptions can be used to help you identify some of the most common Texas ant species. Once you know the species of ant in your home, you can determine where it is likely to nest, what kind of damage it causes, and what kind of control measures are most effective.

Characters used to identify ants

Before you can identify an ant, you should familiarize yourself with a few key things to look for. Before, or while, you collect an ant to identify, look around the area for other similar ants. All ants live in colonies and rarely travel alone. They can usually be seen hunting as a colony for food along well-established foraging trails. When you see an ant trail, observe whether the different workers on the trail vary in size, or whether they are all identical in size. If you can follow the ant trail back to the nest, observe the nest itself. Also, take note of when the ants are active–are they active mostly during the day, or are they most active at night.

If you are still unsure of the ant species after observing them in action, collect a few to examine under a hand lens, magnifier, or microscope, if you have one. The easiest way to capture and kill ants for examination is to fill a small vial, pill container, or jar with a small amount of ethanol or rubbing alcohol. Use the end of a soft cloth, or a brush, wet with alcohol to pick up the ants and drop them in your container. The alcohol will kill your ants quickly and allow you to examine them more closely.

Body characters

Ants, like all insects, have three main body regions. Unlike many other insects, however, ant body regions are very distinct, with obvious constrictions between the head and thorax, and thorax and abdomen (or gaster)(see drawing). Most ants also have antennae with a long first segment that creates a bend, or “elbow” in the middle of the antennae. Although worker ants, the most abundant members of the ant colony, are wingless, the reproductive caste in the ant colony may have wings. One or more times a year, these reproductive ants may fly from the nest (swarming) as part of the mating process. Termites, also social insects, also swarm periodically in structures however, the distinct waist and elbowed antennae are reliable characters for distinguishing ants from termites.

Ants exhibit three distinct body regions, the gaster, thorax and head. The presence of pedicels between abdomen and gaster is what distinguishes ants from similar insects like wasps.

Ants have one or two connecting segments between the abdomen and gaster. These segments are called pedicels, or nodes. Because some wasps resemble ants, the presence of these pedicels can distinguish ants from wasps. Wasps may have narrow, even elongated, waists but they lack the distinct nodes characteristic of ants.

Additional useful body characters of ants to include: a characteristic smell when crushed, presence or absence of a sting, and presence or absence of spines on the head and thorax.

Ant nests

Ant nests may be conspicuous or their nests may be hidden. Soil nests may have a distinctive shape, or they may assume the shape of their hiding place under rocks or other objects. When you find an ant nest in the soil, note its shape or pattern, as well as the number and placement of nest entrances. Some ants nest in trees, either making their own cavities or, more commonly, taking advantage of existing cavities (from rot or termite activity). Some ants, like carpenter and acrobat ants, may use your home as a substitute for their normal, preferred nesting site in a tree or shrub.

Worker size

Ants are classified in this guide as tiny (less than 1/16 inch), medium sized (1/16 to 3/8 inch-long), or large (greater than 3/8 inch-long). Besides length, an important feature that can help distinguish different ant species is whether the worker ants in a colony are all equal in size (monomorphic) or variable in size (polymorphic).

Ant behavior

Ants can often be identified by behaviors that are unique to their species. Notice how fast they run, how they form trails to the nest, how they carry food and distinctive postures when disturbed. Take a few minutes to observe what kinds of food they are carrying, if any (some ants feed on liquids and may not be carrying anything visible). A little time observing behavior may provide clues to the ant’s identity.

Common Texas Household Ants

Red imported fire ants, Solenopsis invicta

One of the most common ant species in eastern Texas and throughout the southeastern U.S. is the red imported fire ant. Fire ants can be identified by their reddish-brown coloration, double pedicel, and by workers that range in size from 1/16 to 3/16 inch. In addition, fire ants make conspicuous mounds with no visible entrance holes on the mound itself. One of the most characteristic behaviors of the fire ant, however, is its aggressive response to nest disturbance including a vigorous and painful sting. Fire ants almost always nest outdoors, although they will enter buildings in search of food and water, especially in late summer.

Carpenter ants, Campanotus species

Carpenter ants are generally large (1/4-1/2 inch), and may be solid black, brown, or a combination of black and red-orange. Workers are variable in size, have a single pedicel, and a smooth, curved thorax in profile. Carpenter ants are mostly nocturnal, coming out at night to follow trails along fences, tree limbs, water hoses or other linear objects. They do not sting, but can bite. They nest in hollow trunks and branches of trees, but will also make their homes in hollow doors, boxes, and the walls and ceilings of buildings.

Acrobat ants, Crematogaster species

These medium-sized ants are frequently confused with carpenter ants because of similar coloration and nesting habits. Upon close inspection, however, acrobat ants are quite distinct from carpenter ants. Acrobat ants have two pedicels, workers are all the same size, and they have two spines on the thorax. Unlike carpenter ants, they are mostly active during the day. When disturbed, acrobat ants lift their distinctive, heart-shaped gasters into the air as a defensive posture, much like an acrobat might balance on her hands. Acrobat ants most commonly nest in trees, but will also make their homes in the walls and insulation of structures.

Pharaoh ants, Monomorium pharaonis

Pharaoh ant. Photo by Forest & Kim Starr, U.S. Geological Survey,

These are tiny ants (1/32 inch-long) that frequently nest and live indoors. They are yellowish in color with a dark-tipped gaster. They have two pedicels and all workers are the same size. They make their nests in any dark, narrow space and may be found nesting in cardboard boxes, electrical boxes, in wall voids, etc. Pharaoh ants are picky eaters, but alternatively feed on foods high in sugars and proteins. Specially designed pharaoh ant baits are usually necessary for good control.

Crazy ants, Paratrechina and Nylanderia species

Crazy ants are usually more hairy than other fast-moving house ants. The tawny crazy ant is lighter colored, and very abundant where it occurs.

Crazy ants are famous for their fast and erratic running behavior. They also may be distinguished from other ants, like the Argentine ant, by the many hairs on their body, including four pairs of hairs on the top of their thorax. One common crazy ant species in east and south Texas is the true crazy ant, Paratrechina longicornis. This dark brown ant has long legs and antennae, is extremely fast moving. It commonly infests buildings, especially in the warmer, more humid parts of the state. A new, exotic species of crazy ant, in the genus Nylanderia, has recently become established in upper Gulf Coast and some parts of the hill country of Texas. It promises to become an important pest of homes in areas where it is introduced. Crazy ants nest in a variety of locations indoors and out. Indoors, look for nests in the soil of potted plants.

Odorous house ant, Tapinoma sessile

Odorous house ant is a common ant that makes its nests outdoors under mulch, stones, and inside and under other objects. They are medium-sized, with a single pedicel that is hidden from above and difficult to see. They are smooth bodied with relatively few hairs, and produce a distinctive, licorice-like, smell when crushed. They are attracted to sweets. Odorous house ants most commonly nest outdoors, but will also nest in bricks and wall voids and other interior locations.

Rover ant, Brachymyrmex patagonicus

This is a relatively new pest in Texas. These are tiny ants, similar in size to the Pharaoh ant, but stocky and darker brown in color. All workers are the same size. They are frequently seen along sidewalk and house edges, on the trunks of trees and in a variety of outdoor locations. They do not sting, but will enter homes where they are attracted to sweet foods. Like odorous house ants, they nest in the ground but colonies will thrive in a variety of locations including the walls of structures.

Leaf-cutting ant, Atta texanus

This large ant is easily identified by its distinctive nest, and leaf-carrying behavior. Leaf cutter ants live outdoors in the soil. The mounds constructed by the ants are crater-shaped and usually clustered together in a “town”. The ants carry freshly cut and dried leaves and other plant material to the large, subterranean nest. These materials are carefully placed into underground “gardens” in which they grow a special kind of fungus. The fungus is the food source for the ants. Worker ants vary in size and are reddish-brown in color. They have long legs and three sets of spines on the thorax, spines on the head, and a smooth gaster. They rarely enter homes but can be common in urban backyards, especially in eastern third of the state, and in the central hill country of Texas.

These are some of the more easily recognized and common house and backyard ants of Texas. If you are still unsure about what kind of ant you have, check with your county extension agent or a local pest control professional for assistance.

For more information

For more information about other indoor ants, see Extension Leaflet B-6183. For a more complete guide to the common ant genera of Texas, check out the Ant genera identification guide (for professionals) (B-6138-for sale).


Carpenter ants damage wood by excavating and creating galleries and tunnels for their nest. These areas are clean, do not contain sawdust or other debris, and are smooth with a well-sanded appearance.

The damage to wood structures is variable. The longer a colony is present in a structure, the greater the damage that can be done. Structural wood can be weakened when carpenter ant damage is severe. Generally, damage occurs slowly, often taking years to occur.

Prevention and control

To prevent carpenter ant problems indoors, eliminate high moisture conditions.

  • Replace moisture-damaged wood.
  • Prevent moisture from wood or lumber that is stored in a garage or near the house by elevating it to allow air circulation.
  • Store firewood as far away from buildings as possible.
  • Remove tree and shrub stumps and roots.
  • Trim branches that overhang the home, so branches don't touch the house (including roof and eaves).
  • Prune branches that touch electrical lines. Carpenter ants can travel from branches to lines and use them to get into buildings.

How to find carpenter ant nests

In order to eliminate carpenter ants nesting indoors, you need to locate and destroy their nest. This is often challenging as nests are hidden and not easily discovered. Careful observations of worker ants will help you find the nest. Observe worker ants between sunset and midnight during spring and summer months.

  • Use a flashlight with a red film over the lens. Ants cannot see red light and won’t be disturbed by it.
  • Or cover part of the flashlight with your hand so it is less bright to follow carpenter ants.

You can increase your chances of following workers to their nest by setting out food that they like. Many foods are attractive to carpenter ants.

  • During spring, carpenter ants are particularly attracted to protein sources, such as tuna packed in water. (They don't like tuna packed in oil.)
  • Set out small pieces of tuna for the ants to take back to their nest. It is easier to follow the ants when they are carrying food.
  • Place food close to where carpenter ants are active so they can easily find it.

Carpenter ants are attracted to honey and other sweet foods.

Other signs that indicate a carpenter ant nest is present:

  • Finding coarse sawdust.
  • Consistent indoor sightings of large numbers of worker ants (20 or more).
  • Large numbers of winged ants indoors (late winter through spring).

Examine areas where steady moisture is or has been a problem. These are a few common locations:

  • Firewood stored in an attached garage, next to the foundation, along an outside wall, or in a basement.
  • Areas around the plumbing or vent entrances.
  • Trees with branches overhanging and touching the house or contacting utility wires.

Pest management professionals may use a moisture meter to find areas prone to carpenter ants.

Listen for sounds that may indicate a nest:

  • An active colony may make a dry, rustling sound that becomes louder if the colony is disturbed.
  • Tap the suspected area and press an ear to the surface to hear any sound.

Pest management professionals may use a stethoscope to locate a nest.

If one nest is found, there may be more nests in a structure.

Indoor control of carpenter ants

The best way to control carpenter ants is to locate and destroy the nest, replace damaged or decayed wood, and eliminate any moisture problems.

Eliminating a carpenter ant nest can be difficult because of the hidden nature of the nest. Carpenter ant control is usually best done by an experienced pest management professional. They have the experience, equipment and a wider array of products to more effectively control a carpenter ant problem.

You can help by telling the pest management professional about when, where and how many ants you've seen.

There can be more than one nest in a building, but only treat nests that have been discovered.

Indoor treatment with dust or liquid pesticides

Spraying foraging workers is not effective. It may temporarily reduce the number of ants you see. However, this will not eliminate a nest because:

  • Ants carry very little insecticide back to their nests.
  • Most ants forage outside and do not come in contact with the insecticides.
  • Only a relatively small percentage of a colony's population is out foraging at any given time.

Nests are often hidden in wall voids, ceilings, subfloors, attics or hollow doors. It is sometimes necessary for a professional pest management applicator to drill small (about 1/8 inch) holes and apply insecticidal dust into the nest area. Don't do this yourself.

Determine the nest's location as specifically as possible. Control should not be applied randomly through the home.

If the nest is exposed (for instance due to remodeling or reroofing), you can use a liquid or aerosol ready-to-use insecticide.

Constant Zombie Threat

The ants can stop fungus from growing within the nest, but the fungus still kills some ants in every colony. The researchers watched 4 colonies for 7 months, and counted all of the dead ants that sprouted a fungus within a 100 m2 area around each nest. That is the same area as 4 full-sized basketball courts, over which they looked for infected ants smaller than the tip of your finger. Ant deaths varied throughout the seasons. One hundred forty-six newly infected ants died during the wettest month measured (December), while only 12 died in one of the driest months (March).

This graph shows the number of dead ants researchers found around the nests. Click for more information.

These researchers also found that dead ants appeared just about everywhere around the nests. The location of ant trails did not move away from any fungal threats.

So how can ants stay safe from ant-zombies? Well, remember the ant safety net? They only send out the oldest individuals to collect food on these trails. This means they only expose a portion of their nest to the outside world, where the zombie threat lurks around every corner.

While ants may have figured out how to avoid the spread of zombie fungus, remember this the next time you watch a zombie movie. While you may be safe and sound from zombies, some insects have to live alongside zombies every day.

Additional images via Wikimedia Commons. Zombie snail by Gilles San Martin.

Small wonders: What ants can teach us

When most people think about ants - if they think about them at all - they think of pests in the pantry or on a picnic.

But in Belize, the ant is the king of the jungle . constantly on the prowl, under every rock and in almost every flower.

In the Belizean rainforest there are several hundred different ant species - dozens in the treetops alone, says Mark Moffett - biologist, author, photographer and ant-enthusiast almost from birth.

"I learned early, like when I was in diapers, that ants are controlling the world under our feet," he said. "Down there as an infant I would watch them doing all of these things that were very human-like. Building roads, working together to collect food, all kinds of things. Ants do all kinds of things that even primates, like a chimpanzee, don't have to deal with."

Trending News

Take the leaf-cutter ant. These insects live in colonies composed of millions, and feeding all those millions of mandibles requires a lot of work.

"This is a tough job, and their jaws get quite worn down by it," Moffett said. "Their jaws, however, contain a lot of zinc, so they're essentially living can openers that can grab onto the leaf from one side and tear through with that other tooth from the other side the way you use a little portable can opener."

A lot of ants carry leaves with hitchhikers on them. "This was something that early explorers even pointed out: Why are these little ants climbing on top of the leaves and getting hauled along?" Moffett said. "Well, one reason is it probably costs the colony less energy for them to stand on the leaves than to walk themselves. So this is just good economics."

"Carpooling!" Salie said. "These leafcutters are carrying their booty back to the colony. But they're not gonna eat the leaves."

"No, they don't actually eat these leaves. And you would think they would 'cause they're carrying literally pound after pound of leaf down this tree. But they actually turn them into a mulch on which they raise a fungus. They're fungus-eating ants."

They're entirely farmers. In fact, says Moffett, "they do everything you think human farmers do."

And in case you were wondering what that farm-fresh fungus looks like, we dug some up.

"This is ambrosia for the ants," said Moffett ,"and they don't need anything more."

"This is kind of their Power Bar," Salie said.

"It is, it's very nutritious. You wanna try some?"

"No, but I'd like to watch you try some," Salie said.

The verdict? "It needs a little chocolate."

Of course, ants don't just create farms they make assembly lines, highways and even underground cities.

"You get a variety of different sizes of workers with different shapes. And they're all built specifically to do certain tasks or jobs. So they are born with this identity," said Moffett."

There are soldiers, nurses, sanitation specialists, highway construction workers.

There are actually "suicide bomber" ants. "The ant simply walks up to the enemy and explodes - spraying this toxic yellow glue over itself and everything around it," Moffett said. "That's right out of sci-fi. It doesn't even need any TNT. It just has it built into its body."

With behavior this "human-like," they must be pretty smart, right?

No, according to Deborah Gordon, professor of biology at Stanford: "Ants are not smart. In fact, if you watch an ant for any length of time, you're gonna end up wanting to help it, because ants are really very inept.

"But colonies are smart. So what's amazing about ants is that in the aggregate, all of these inept creatures accomplish amazing feats as colonies," she said.

And according to Gordon, they do it all without a boss.

"In an ant colony, there's nobody in charge. There are no bureaucrats. There are no foremen. There are no managers. There is nobody telling anybody what to do," she said.

Wait a minute? Don't ant colonies have queens?

"The queen does not give rules," explained Moffett. "She does not make proclamations. She just sits there and lays eggs. Being the queen would be the most boring job in the ant society."

Okay, so the queen isn't in charge. But ladies do rule the colony. Virtually any ant you ever see is a female - males just mate once and die.

So all these females survive and thrive together, all without a leader - which can be hard for us humans to understand.

"We put a lot of effort into thinking through how to organize some of the things that we try to do as groups," said Gordon. "Ants don't put in any effort at all. They're pretty messy about it, and it works really well."

So, if no one is in charge, how DO ants make decisions?

"Ants largely communicate by scent over very short distance," Moffett said. "It makes sense to actually put out little puffs of odor. If they're trying to do something in a longer range, like there's a battlefield, at some distance, they will lay down that little odor trail."

Most ants, it turns out, simply "follow the crowd" the more ants follow a trail, the stronger the trail's scent is - and the stronger the trail's scent is, the more likely ant will follow it.

Ants provided some guidance for Doug Lawson, a systems analyst at Southwest Airlines.

"Because we know that ants have accomplished these amazing things right based on very simple rules, we know that if we want to see something complicated happen - like completely filling the interior of an aircraft with people - we know that simple ant-type behavior is adequate to represent what's occurring," Lawson said.

"So Southwest Airlines said, 'Help us figure out the most efficient way to help us get our passengers on a plane,' and you said, 'I know - I'll use ants'?" asked Salie.

"Yeah, right. Because they do complicated things with very simple rules," Lawson said.

Lawson used mathematically-modeled ants to determine the most efficient way of boarding a plane, which turns out to be open seating.

"So Southwest's way of boarding without seat numbers is actually more efficient than when I board another airline and know exactly what my seat is?" asked Salie.

"Right. When we simulated what the different airlines are doing, it turns out that with assigned seats, there's a one-third chance that you're going to ask two people to get up, whereas open seating - since the middle seat is the undesirable one - generally that's the one that's last to be filled, [so] only one person is likely to get up, the person sitting near the aisle," said Lawson. "I may have to ask somebody to get up and get out of the way to let me get to a seat, and that's about it. So it's really simple.

"So the ants are sort of an analogy - simple rules produce complicated transactions and events and major structures. and we can do the same thing in a little simulated world and represent very complicated processes," said Lawson.

"Now, did these ants have carry-on baggage? Were these ants cranky?" asked Salie.

"Yeah, we left out bad behavior," Lawson replied.

For human behavior, it turns out ants have a lot to teach us about activities that don't require a lot of brain power, like what happens at the check-in counter.

"Basically if you have enough employees or machines - or ants in a colony - they can all have very specific tasks," said Lawson. "And that's how ant society is. And that's how they evolved jobs over millions of years. It's come to be that we need a nursing ant we need a soldier ant."

Or a Southwest employee directing passengers to the next open check-in machine.

Lawson believes airline operations are just the beginning of what mathematically modeled ants have to offer .

"Anything that provides a service can be converted into an ant, and so then they'll adapt, too," Lawson said. "So a service device could be an ant also, and it could change its behavior - what service is it offering? Where should it be? It could wander around the lobby trying to figure out where it ought to stand, and what kind of service it should offer.

"So when you think about it that way, anything - an ATM, a parking space, an aircraft - convert them into an ant and let them behave, too, under the influence of the customer who wants them, and they'll change their behavior too."

"Arguably, humans are too smart for the functioning of the whole society - it pays to be individually stupid," Moffett laughed. "This is the wisdom of the crowds idea brought to ants. Basically, all those little ants with their mostly ignorant choices, out of all that emerges a smart society."


Institute for Computational Medicine, NYU Langone Health, New York, NY, 10016, USA

Institute for Computer Science & Department of Biology, Heinrich Heine University, 40225, Düsseldorf, Germany

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar


IY and MJL developed the ideas and wrote the manuscript together. The authors read and approved the final manuscript.

Corresponding authors

Extreme lifespan extension in tapeworm-infected ant workers

Social insects are hosts of diverse parasites, but the influence of these parasites on phenotypic host traits is not yet well understood. Here, we tracked the survival of tapeworm-infected ant workers, their uninfected nest-mates and of ants from unparasitized colonies. Our multi-year study on the ant Temnothorax nylanderi, the intermediate host of the tapeworm Anomotaenia brevis, revealed a prolonged lifespan of infected workers compared with their uninfected peers. Intriguingly, their survival over 3 years did not differ from those of (uninfected) queens, whose lifespan can reach two decades. By contrast, uninfected workers from parasitized colonies suffered from increased mortality compared with uninfected workers from unparasitized colonies. Infected workers exhibited a metabolic rate and lipid content similar to young workers in this species, and they received more social care than uninfected workers and queens in their colonies. This increased attention could be mediated by their deviant chemical profile, which we determined to elicit more interest from uninfected nest-mates in a separate experiment. In conclusion, our study demonstrates an extreme lifespan extension in a social host following tapeworm infection, which appears to enable host workers to retain traits typical for young workers.

1. Introduction

Reproductive and task division of labour have been long acknowledged to be hallmarks of eusociality in insect societies. The queen specializes in reproduction and the workers collectively perform all other tasks, including caring for the queen and her offspring, building and defending the nest, and foraging for food [1]. An intriguing feature of many social insects, and of ants in particular, is the stark divergence in lifespan between female castes, which are typically not genetically determined [2]. Many ant queens exhibit extraordinarily long lifespans of several decades. They remain fertile during their lifetime spending it almost exclusively inside the nest tended by their worker daughters [3]. By contrast, ant workers—being sterile—exhibit much shorter lifespans of only a few weeks, months or rarely years [4–6]. These sterile workers take over all chores in the nest, and as they grow older, they switch from inside work such as brood care to the much riskier outside tasks [7].

The social environment is known to improve the health and survival of animals typically living in groups [8–10]. In social insects, this is especially apparent for queens, whose long lifespans can be in part explained by the high levels of social care they receive from their workers and the reduced extrinsic mortality due to the safe environment in the nest [9]. Despite these obvious benefits of sociality, a social lifestyle with the close proximity of genetically similar group members also provides favourable conditions for parasites to spread and thrive [11,12]. Parasites typically reduce the fecundity and survival of their insect hosts due to their reliance on their hosts' resources [13]. Social insects serve as hosts for diverse parasites and workers leaving the nest to forage for food are frequently exposed to parasites [12–15]. Infections in social insect workers often lead to behavioural changes and can even accelerate the behavioural maturation of young workers [16–18]. This highlights the importance of parasites for the phenotypes of social insect hosts and reveals how social traits can be intertwined with parasite-induced alterations.

Parasite infection generally incurs fitness costs to their hosts, such as a slower development, reduced survival or a lower fecundity [19]. Surprisingly, some parasites extend the lifespan of their hosts, for instance by interfering with the fecundity–longevity trade-off by reducing the reproductive success of hosts up to complete sterilization [20–22]. We have previously demonstrated increased survival over a few weeks of tapeworm-infected workers of the small Central European ant Temnothorax nylanderi [23]. Ants of this species serve as an intermediate host to the trophically transmitted tapeworm, Anomotaenia brevis [23]. Infected workers do not show reduced reproductive potential compared with their uninfected worker sisters on the contrary, they develop their ovaries even more strongly when the queen is removed [24]. Ant workers get infected during the larval stage when fed with tapeworm eggs [25], which develop within the ants to parasitic cysticercoid larvae. A single ant can be parasitized by as many as 70 cysticercoids (T. Sistermans 2020, unpublished data [26]), that reside in the haemocoel of their abdomen. The complex life cycle of A. brevis is completed when woodpeckers prey on parasitized ant colonies that live in cavities of sticks or acorns on the forest floor. Then, inside the bird's gut, the cysticercoids develop into adult tapeworms [27]. Next to increased survival, tapeworm infection leads to a multitude of phenotypic changes in infected workers, which are easily identified by their yellow, less sclerotized cuticle compared with their brown nest-mates [28]. Infected workers are less active, stay on the brood pile and exhibit reduced flight behaviour [23,26]. These behavioural changes are also reflected in an altered gene expression in the brain and abdomen [29,30].

The main focus of our multi-year study was to investigate the long-term consequences of A. brevis infection on T. nylanderi workers. We were interested in how long infected workers can live or whether their reported increase in survival might be due to traits more typically expressed by younger ants. We therefore tracked worker and queen survival in parasitized and unparasitized ant colonies over 3 years until over 95% of the uninfected workers had died. The factors that directly or indirectly regulate lifespan in social insects are still poorly understood. In particular, the role of social behaviours of group members on longevity has hardly been studied so far. In our focal species, infected workers are known to be fed more often than their nest-mates [26]. We thus were interested to gain insights into whether ants in parasitized societies receive more social attention depending on their infection status or caste. We further examined whether the deviant cuticular hydrocarbon profile of infected workers is more attractive to their nest-mates, which could explain why they receive more attention [28,31]. Physiological traits are known to be associated with the pace of life and could therefore be good markers of longevity [32–34]. We thus finally analysed the metabolic rate and lipid content of workers and queens in parasitized and unparasitized colonies.

2. Methods

The ant colonies in this study were collected in forested areas close to Mainz-Wiesbaden, Germany, from three different sites: (i) 50°00′36.4″ N, 8°10′47.3″ E (ii) 50°02′29.4″ N, 8°02′46.6″ E and (iii) 50°05′42.8″ N, 8°09′55.1″ E.

2.1. Survival of uninfected and infected workers and queens

We noted the survival of workers and queens of different age groups from parasitized and unparasitized colonies for 3 years (1110 days, start: 22 September 2014, end: 6 October 2017). We collected 8 and 22 parasitized T. nylanderi colonies and 9 and 19 unparasitized colonies in September 2013 and April 2014, respectively. All colonies were queenright and comprised between 22 and 245 workers (121 ± 58 workers: mean ± 1 s.d.). Colony size (i.e. the number of workers) did not differ between parasitized and unparasitized colonies (Wilcoxon test: W = 429.5, p = 0.89). Temnothorax ants display a synchronized annual reproductive cycle. During four weeks in summer, all new workers and sexuals emerge from the pupae. We took advantage of this synchronized emergence in mid-September 2014 to differentiate between young and old workers. Young workers—so-called callows—are easily distinguishable from older ones by their light, not yet fully sclerotized cuticle. These workers focus on brood care and are henceforth referred to as nurses. Workers leaving the nest to forage for food—so-called foragers—are usually older and were at the time of collection likely to be 1 year or older [31,32]. Temnothorax nylanderi ants are brownish with a characteristic dark abdominal stripe that is visible in callows, but is missing in infected workers with their unpigmented, yellow cuticle [25,26,28].

Within a day of emergence, we wire-marked (ELEKTRISOLA, 0.025 mm) five nurses (n = 285) and concurrently five foragers (n = 285) in each colony. All nurses and foragers were identified to be uninfected based on their body coloration (e.g. brownish, with abdominal stripe) in this and all following experiments. In parasitized colonies, we additionally marked between one to five infected workers (depending on availability), which were identified by their unpigmented cuticle (n = 98). To be able to distinguish young from old workers, wire-markings were unique for nurses and foragers within each colony, but were randomized between colonies. Ant queens can be easily distinguished based on their distinct morphology (e.g. larger body and structured thorax) and were not wire-marked. As T. nylanderi is strictly monogynous [24], queens were likely to already be a few years old at the beginning of the observations, as they had successfully established a colony. At the onset of our observations, all infected workers and selected foragers were at least 1 year old, whereas callows were only 10 days old.

The ant colonies were maintained in observation nests consisting of two glass slides separated by a piece of plexiglas providing a cavity of 4.9 × 1.1 × 0.3 cm. These observation nests were placed in three-chambered boxes with a moistened plaster floor and kept in climate chambers, set to temperatures and photoperiods typical to the season (December–February: 10°C : 5°C day : night (DN) temperature and 10 h : 14 h light : dark (LD) period March–May: DN 20°C : 15°C and LD 12 h : 12 h June–August: DN 25°C : 18°C and LD 12 h : 12 h September–November: DN 18°C : 13°C and LD 12 h : 12 h). We recorded the survival of all marked workers and the queens at 10-day intervals. The day of death was recorded as the last day the ant was observed alive. During the 3 years, 16.8% (48 of 285) of marked young workers, 21.0% (60 of 285) of marked foragers and 15.3% (15 of 98) marked infected workers disappeared, with no corpses found. For these individuals, the day of disappearance was entered as the day of death. We found the corpses of all queens who died. On each observation day (i.e. every 10 days), colonies were fed with pieces of crickets and a droplet of honey, except during hibernation (December–February), when we provided colonies with a droplet of honey at every second observation. Water was offered ad libitum throughout the entire time. Ant survival was examined using Kaplan–Meier analyses allowing for right-censored data, that is the number of days until death per individual. The log-rank Mantel–Cox test was employed to determine statistical differences and groups were compared for their hazards. All statistical analyses were performed in the Graphpad Prism v. 8.3.0 software.

2.2. Worker and queen metabolic rate, body mass, lipid content and social care

We collected an additional set of 14 parasitized and 14 unparasitized colonies in April–May 2018. All colonies were headed by a single queen and contained 69 ± 33 workers (mean ± 1 s.d.). Parasitized and unparasitized colonies did not differ in colony size (i.e. number of workers or brood, Wilcoxon tests, both p > 0.7). Infected workers and queens were easily recognized based on colour or morphology (see above). As all uninfected workers were non-callow adults (i.e. sclerotized cuticle), nurses and foragers were identified by their behaviour and location in the nest: nursing the brood inside the nest or foraging outside [24,35]. We cannot provide the exact age information about these ants. However, as colonies contained more than 20 workers, queens were likely to be a few years old. All workers were more than 10 months old, as the experiment began in June before the emergence of the new annual worker generation, which emerges in late summer. In Temnothorax, foragers are generally older than brood care workers [36,37].

We first measured the O2-consumption of workers and queens from each parasitized and unparasitized colony. Measurements were taken with the MicroRespiration system from UNISENSE (Denmark), following their protocol. Ants were isolated from their colony and individually placed in a micro-respiration chamber (v = 0.448 ml) and sealed with agar and paraffin oil. The glass chamber was transferred to a water bath at a constant temperature of 23°C. A thin capillary in the chamber lid served as an oxygen microsensor to measure the O2-consumption. Real-time O2-consumption was recorded for 10 min and viewed using the free software SensorTraceBasic v. 3.0.200. All ants were weighed directly after the measurement (accuracy of 1 μg PESCALE Wägetechnik). We calculated the respiration rate using the linear section of the O2-consumption slope (from minute 5 to minute 10), adjusted for the live body mass (mg). The variable we used was the slope of O2-consumption (µmol l −1 ) plotted against time (s), divided by the ant mass (mg) and multiplied by the chamber volume (ml), hereafter ‘metabolic rate’.

Following the respiration measurements, the ants were randomly marked with coloured wires and returned to their colony. On the following day, we conducted 20 behavioural scans of each focal individual and noted down whether or not an individual received social care from their nest-mates (i.e. being groomed, fed, carried or antennated). Thereafter, we calculated the rate of social care occurrences from all 20 observations. We also noted down whether we observed active begging behaviour for food. The next day, all marked ants were individually frozen at −20°C. To extract their lipids, they were placed in a chloroform/methanol mixture (2 : 1, v/v) for 24 h [28]. Nonadecanoic acid (C19 : 0) was added as the internal standard (20 µl in DCM/MeOH, 2 : 1 v/v 0.2 mg ml −1 ). The extracts were then fractionated in Chromabond SiOH columns (1 ml Macherey-Nagel). Each column was conditioned with chloroform and hexane, and the lipids were eluted with chloroform. The samples were dried under a nitrogen stream and dissolved in 250 µl of a 2 : 1 dichloromethane/methanol mixture (v/v). Lipid extracts were analysed with a 7890A gas chromatograph (Agilent) coupled to a 5975C mass-selective detector (Agilent). The oven programme started at 60°C for 1 min, then increased by 15 K min −1 to 150°C, followed by an increase of 3 K min −1 to 200°C and finally a ramp of 10 K min −1 up to 320°C, where it remained constant for 10 min. Peak areas were integrated manually using the Agilent software MSD Chem Station E.02.02. The data were then exported to MS Excel and manually aligned. The fatty acids had chain lengths between C12 and C20 and were identified based on diagnostic ions, retention time and the molecular peak. We calculated the total lipid content from the quantity of the internal standard and the quantity of all fatty acids together, and divided it by live body mass (mg) to obtain the relative lipid content.

The datasets of metabolic rate, body mass (both square-root transformed) and lipid content were separately analysed using linear mixed models (LMM). The relative frequency of social interactions was assessed using general linear models (GLMM) following binomial distribution with a logit-link function. We examined whether ants differed in their metabolic rate, body mass, lipid content and the frequency of social care depending on the parasitism status of the colony, their own caste and infection status. Hence, each model (LMMs and GLMM) incorporated colony parasitism (parasitized/ unparasitized) in interaction with the category of the individual ant (infected/nurse/forager/queen) as fixed predictors and colony identity as the random factor. We applied a backward stepwise selection procedure for model selection (α = 0.05). Neither metabolic rate, body mass, lipid content nor the frequency of social care was impacted by the parasitism status of the colony (see results). We thus only focused on parasitized colonies, analysing whether and how the metabolic rate (LMM), body mass (LMM), lipid content (LMM) and the frequency of social care (GLMM) differed among infected workers, nurses, foragers and queens. The individual ant category served as a single explanatory variable and colony identity was included as the random factor. Pairwise comparisons between groups were obtained by re-running the models by relevelling factor levels. Models were run in R v. 3.3.2 using package lme4 (commands lmer, glmer) [38].

2.3. Response to cuticular hydrocarbons of infected and uninfected workers

Chemical cues trigger behavioural changes in social insects. We thus investigated whether uninfected workers in parasitized colonies are attracted by the cuticular hydrocarbons (CHCs) of infected nest-mates, which provides a first possible explanation for increase care. For this, we collected 12 parasitized colonies 83 ± 40 workers (mean ± 1 s.d.) and 12 unparasitized colonies 110 ± 25 workers (mean ± 1 s.d.) in autumn 2019. Parasitized colonies contained at least five infected ant workers. From each colony, we froze five infected and five uninfected, brood-caring workers at −20°C. CHCs were extracted by adding 0.5 ml hexane to each vial with either five nurses or five infected workers. The ants were removed after 10 min. The extracts were transferred into inlays, concentrated to approximately 15 µl by evaporating the hexane under a gentle nitrogen stream and frozen until use. A day before the experiment, we relocated 10 uninfected workers, 10 larvae and the queen of each colony in a Petri dish nest (Falcon, ∅︀ 5 cm) with a plastered floor. Ten microlitres of CHC extract was added to a glass bead (Roth, ∅︀ 1.7–2.1 mm), and the hexane was given 20 min to evaporate. We presented to each parasitized sub-colony two glass beads, one covered with the CHCs of uninfected nest-mates and the other one with the CHCs of infected nest-mate workers. The position of the glass beads within the nest (left or right) was randomized and they were equidistant to the centre. Each sub-colony was video-recorded for 20 min (Canon, HD Camcorder Legria HFR706). The experiments were conducted at 21°C in a climate chamber with 80% humidity. We used unparasitized colonies to investigate whether nest-mate CHCs generally elicit the interest of workers. The procedures were the same except that the sub-colonies were presented with two glass beads covered with either 10 µl of nest-mate CHC extract or 10 µl hexane. The videos were analysed with the VLC media player with the observer being blind as to the experiment and treatment of the glass beads. We noted the number of interactions directed towards each glass bead. The data were analysed with R (v. 3.5.1) using GLMM (package glm2, command glm) with a binomial or quasi-binomial distribution. The model was used to calculate whether the number of contacts towards the two alternative glass beads differed from a 50 : 50 chance. The results of one trial of a parasitized sub-colony were removed, in which the ants had little contact with the glass beads (less than 10 times). For data visualization, we used the proportion of contacts towards glass beads covered with either infected nest-mate CHCs for parasitized colonies or nest-mate CHC for parasitized colonies (command ggplot).

3. Results

3.1. Survival of uninfected and infected workers and queens

Infected workers survived considerably longer than uninfected workers (i.e. nurses and foragers), but their survival did not differ from that of queens in both parasitized and unparasitized colonies (table 1 and figure 1a). The presence of infected workers reduced the survival of their uninfected nest-mates. In particular, foragers from parasitized colonies survived less long than foragers from unparasitized colonies, while this lifespan reduction was less pronounced and borderline non-significant in nurses (table 1 and figure 1a,b). Queens from parasitized colonies showed no change in survival compared to queens from unparasitized colonies (figure 1a). Nurses in parasitized colonies survived longer than foragers, whereas nurses from unparasitized colonies survived as long as foragers (table 1 and figure 1b).

Table 1. Results of log-rank Mantel–Cox tests comparing survival of ant categories.

Figure 1. (a) Survival of queens (green), infected (yellow) and uninfected workers (brown) of T. nylanderi from unparasitized (dash) and parasitized colonies (solid) over 1110 days. (b) Survival compared between infected workers (yellow), nurses (red) and foragers (blue) from unparasitized (dash) and parasitized (solid) ant colonies. Statistical results are provided in table 1.

None of the 285 marked, uninfected workers in parasitized colonies survived until the end of the observation period, whereas 52 of 98 (approx. 53%) infected workers and 15 of 29 (approx. 55%) queens were still alive after 3 years. In unparasitized colonies, none of the 140 nurses, only 3 of 140 foragers (approx. 2%) and 17 of 28 (approx. 61%) queens were alive after 3 years. Across parasitized and unparasitized colonies, uninfected nurses survived on average for 296 ± 216 days, foragers for 254 ± 217 days, infected workers for 842 ± 300 days and queens for 862 ± 309 days (mean ± 1 s.d.). Overall, 12 of 57 (approx. 21%) colonies died within the 3-year observation period, that is both the queen and all workers perished. Parasitized colonies were as likely to die as unparasitized colonies (npara = 8, nunpara = 4, Fisher's exact test: p = 0.33).

3.2. Metabolic rate, body mass and lipid content of workers and queens

Infected workers and nurses exhibited similar metabolic rates (t(23.35) = 0.88, p = 0.39). Compared with all other categories, queens had a lower (all p < 0.002) and foragers a higher metabolic rate (all p < 0.006 figure 2a). Infected workers and nurses also had a similar body mass (t(25.36) = 1.36, p = 0.19) and relative lipid content, although there was a trend for infected workers to contain a lower fraction of lipids (t(25.32) = 1.51, p = 0.07 figure 2b,c). Queens had the highest body mass, while foragers had the lowest, yet both had similar relative lipid contents (t(18.99) = 0.39, p = 0.70). Infected workers did not differ from queens and foragers in their relative fat contents (t(22.99)queens = 1.09, p = 0.29, t(21.97)foragers = 1.46, p = 0.16 figure 2c). Ants from parasitized and unparasitized colonies did not differ in their metabolic rate, body mass and lipid content (table 2).

Figure 2. Differences in (a) metabolic rate, (b) body mass, (c) lipid content and (d) received social care of queens (grey), infected workers (yellow), nurses (red) and foragers (blue) from T. nylanderi colonies. Vertical lines represent standard error. Different letters indicate significant differences between categories. Statistical results are provided in tables 2 and 3.

Table 2. Results of LMM comparing metabolic rate, lipid content and body mass measurements ant categories and colony types.

3.3. Social care provided to workers and queens

In parasitized colonies, social care behaviours directed towards infected workers exceeded those towards nurses, foragers and even the queen (table 3 and figure 2d). Active begging behaviour for food was very rarely observed (5 out of 1780 total interactions), and infected workers were never involved in these interactions. Foragers received less care than nurses and nurses received less care than the queen (table 1 and figure 2d). The rate of social care was unaffected by colony parasitism status as a main effect nor in interaction with individual ant category (table 3).

Table 3. Results of GLMM comparing the frequency of social behaviour between different ant categories and colony types.

3.4. Response towards cuticular hydrocarbons of infected and uninfected workers

In unparasitized colonies, significantly more attention was directed towards nest-mate-derived CHCs, relative to the hexane control (GLM: z = 3.07, p = 0.002, figure 3b). In parasitized colonies, the CHCs from infected nest-mates elicited more responses than the CHCs from uninfected nest-mates (GLM: z = 6.67, p < 0.0001 figure 3a), suggesting that the chemical profile of infected workers is more attractive for nest-mates.

Figure 3. Response of ants to glass beads covered with cuticular hydrocarbon extracts. (a) Contacts of uninfected workers from parasitized colonies to glass beads covered with CHCs of infected nest-mates to an uninfected nest-mate control. (b) Contacts of uninfected workers from unparasitized colonies to glass beads covered with CHCs of nest-mate workers to a hexane control.

4. Discussion

Parasites generally reduce the fitness of their hosts. In those rare cases in which parasites extend their hosts' lifespan, they typically decrease the fecundity of their hosts [20,21]. Here, we report an at least threefold prolonged lifespan of ant workers infected with a helminth, and these infected workers exhibit a similar if not larger reproductive potential than uninfected members of their caste [24]. During our 3-year observation period, the survival of infected workers was similar to that of queens, which can live for up to two decades in this species [39]. The observed differences in survival were extreme. While more than half of all infected workers were still alive after more than 1000 days, all of their uninfected nest-mate workers had already died.

Endoparasites are scavengers of host resources. Infection with tapeworm larvae may therefore cause a reduction in body mass and fat content, and possibly an increase in metabolic rate of host workers. Alternatively, as infected workers survive as long as the queen does, they might resemble the royal caste also in physiology. We found more support for the first hypothesis as infected workers exhibited a similar metabolic rate, body mass and lipid content as nurses do, which are the youngest members of ant colonies [36]. Infected workers also resemble nurses in their cuticular hydrocarbon profile and are able to invest in the development of their ovaries in the absence of the queen, just like many nurses do [24,36]. In addition, infected workers are often inactive and stay in close contact with the brood, which is the typical location of nurses within an ant nest [24,26,37].

In many traits, infected workers thus appear to age more slowly, but what are the proximate causes for their extended lifespan? Presently, we are lacking mechanistic explanations and multiple factors including intrinsic physiological changes and extrinsic conditions seem to play a role [40–43]. Whether and how social behaviours contribute to lifespan differences in social insects is unknown. We show that infected workers not only obtain more care than the average uninfected workers, but also more than the queen, which is usually the best cared for ant in the colony. Could such differences indirectly modulate lifespan and if yes, what are the underlying mechanisms? Studies addressing these questions will be the scope of future studies.

Infected workers are well provided with food [26], and our behavioural observations in this study do not indicate that they actively beg for food more than other workers. Rather, our experiments with cuticular hydrocarbon extracts indicate that the chemical signals of infected workers are highly attractive for their nest-mates. It is not surprising that infected workers signal their needs in this way as communication in ant societies is primarily mediated through chemical cues [44]. The increased attention directed towards infected workers might be caused by specific cuticular compounds, or simply by their profile being different from that of other workers [28]. The latter explanation is less likely as we could previously show that the chemical profile of infected workers is closer to that of uninfected nurses, than the latters’ CHC profile is to that of foragers [31]. It remains unclear, whether and how the infection changes chemical signalling in infected workers.

Being infected by a parasite is usually costly for the host. In this host–parasite system, there are no direct negative effects on the infected workers themselves rather the cost of parasitism becomes visible in their uninfected nest-mates. Indeed, uninfected foragers of parasitized colonies exhibit elevated mortality compared with workers residing in unparasitized colonies, as shown here and in an earlier study [23]. This increased mortality was apparent, although our ant colonies were well maintained in our laboratory, regularly receiving ample food and water. Potentially, the care and high food demands of infected workers increase the workload of their nest-mates, which in turn might cause their increased mortality [26]. Yet, our physiological data did not point to increased physiological stress, as infected workers in parasitized colonies exhibited a similar metabolic rate, body mass and lipid content to those from unparasitized colonies. Moreover, the survival of queens from parasitized colonies was unaffected by colony parasitism. The well-being of the queen is critical for the entire colony as only she can ensure long-term colony survival. Thus, though workers from parasitized colonies provide more care for their infected nest-mates, they should avoid neglecting the queen. Preliminary observations indicate a decrease in queen care with an increasing fraction of infected workers (A. Lenhart 2020, personal observation). Moreover, parasitized colonies raise more queen–worker intercastes and show a male-biased sex ratio, both are potential signals of stress on the colony level [26]. However, these negative consequences of parasitism are rather weak as A. brevis does not negatively affect the reproductive output of T. nylanderi field colonies and parasitized colonies are as often queenright as unparasitized colonies [26].

Parasites with complex life cycles regularly alter the phenotype of their intermediate hosts [45,46]. Besides changes in life history, the behaviour and morphology of hosts are often profoundly modified [47]. Such alterations might represent host compensatory responses or side effects of infection, but could also be adaptations of the parasite to secure survival or enhance transmission [46]. Workers of T. nylanderi ants infected by the tapeworm A. brevis are highly modified organisms. Next to their extraordinary long lifespan, the infection reduces the behavioural repertoire of the workers [24,26]. They remain predominantly inactive inside the nest and show reduced anti-predatory responses [23,26]. This stands in stark contrast to behavioural changes observed in ants used as intermediate hosts by other parasites with complex life cycles. Ant workers infected by the lancet liver fluke leave the nest to expose themselves to the definitive host of this parasite, large herbivore mammals [48]. Temnothorax workers are tiny, about 2–3 mm in length, and live in the leaf litter, where they are difficult to spot for their definitive hosts, woodpeckers [25]. Sending infected workers out of the nest and exposing them to high extrinsic mortality might be a rather unsuccessful strategy. Instead, infected workers remain in the safety of their acorn or stick nests, which is the very place woodpeckers are opening in search of insect larvae. Thus, it is likely that these behavioural alterations observed in tapeworm-infected workers actually may predispose ants to predation by birds. Accordingly, the extended lifespan could be caused by the parasite in order to prolong the period of possible transmission. The lifespan extension might be especially feasible in ant workers, which can rely on the care from their nest-mates and where genes for a long life, that is queen-specific genes, could be activated [30]. Lifespan extension following infection with parasites has been found in other host–parasite systems [20,22,49–52]. In these cases, the parasite either reduces the fecundity of its host thereby triggering a lifespan extension or it appears to increase host survival to be able to finish its own development before the host dies. Our case lays somewhat different, as here the parasite may benefit from a lifespan extension of its intermediate host to facilitate its transmission to the final host.


Ant collection permits were obtained from local forestry departments. We followed the guidelines of the Study of Animal Behaviour and the legal and institutional rules.