Thermoregulation strategies in ants in comparison to other social insects, with a focus on red wood ants (Formica rufa group)

Nest architecture and properties of nest material

Ants nest in wide range of materials; in soil, under stones, in leaf litter and even in living trees (see Table 1). Some of them build above-ground nests, usually from soil or organic material, called ant hills or mounds. These nests show an advanced ability in regulating inner temperature.

The red imported fire ant Solenopsis invicta and members of the genus Pogonomyrmex build soil mounds from excavated soil^7,8. These nests gain heat through solar radiation and the brood is moved along an increased thermal gradient^1,9. Weaver ants from the genus Oecophylla construct their nest from living leaves, with the help of their own larvae, which produce a special “glue” from their salivary glands¹⁰.

Red wood ants from the genus Formica build large nests from organic material that is based on a mixture of soil, twigs, coniferous needles and pebbles^3,11–13. In these nests a stable heat core can be maintained thanks to the good insulation properties of these materials and the metabolic heat produced by the ants or their associated microflora^14,15.

The composition of the organic material is not the same throughout the whole nest volume^3,16 and the mound structure is not rigid either. The ants loosen and renovate the nest structure and the organic material is continuously moved from the inside to the outer layers³. Nest structure and architecture plays a vital role in nest thermoregulation. The tunnels and passages build by ants are important mainly for ventilation and humidity control, which affect the nest temperature. Pieces of tree resin are often incorporated into the Formica nest material because of its antimicrobial effect¹⁷. The resin inhibits the growth of potentially pathogenic bacteria and fungi in the nest.

Thatch ants, Formica obscuripes¹⁸ and Acromyrmex heyeri¹⁹, use plant fragments as a building material and arrange them in a thick compact surface layer called “thatch” which has a lower thermal diffusivity than the surrounding soil. The thatch prevents nest overheating by the incoming solar radiation and avoids losses of the accumulated heat into the cold air during night¹⁹.

Nest moisture can have two different and opposite thermoregulatory effects: 1) moisture can support microbial heat production (i.e. increase the temperature)^3,15; 2) it can decrease the insulating properties of nest material (i.e. decrease the temperature)¹¹. A study of the relationship between daily temperature regime and moisture in Formica polyctena nests revealed two possible thermoregulation strategies which differ between dry nests and wet nests¹⁵. Dry young nests are usually located in sunny open places¹². Solar radiation heats up the nest and keeps the nest material dry, with low heat capacity and conductivity. The thermal losses of dry nests are estimated to be 0.15–4.3 W per nest¹⁵. The temperatures of dry nests are usually the highest in the evening and they drop during the night. Thermoregulation in dry nests is based on a combination of metabolic heating from the ants, the insulating properties of the nest and solar heating¹⁵.

During forest succession, ant nests become more shaded and gradually switch to a wet thermoregulation regime¹². In the evening the temperature in these nests is low and it increases during the night. The high night temperature at the nest surface indicates substantial heat loss, about 24–30 W per nest. The wet nests have a high thermal capacity; increasing the temperature by 1°C requires a thermal input of 35 W¹⁵. This means that maintaining a sufficient temperature in a wet nest requires a heat source beyond the metabolic heat produced by ants and the heat obtained from solar radiation. The additional heat source in wet nests is provided by microbial activity^3,15. Both of these two types of thermoregulation strategies are applied in natural populations, but in different stages of nest development. The microbial community in F. polyctena nests differs from that in the surrounding soil in part because of differences in pH and food availability and quality¹³. Thermoregulation via microbial heating was first proposed in 1915 by Wasmann³. In 1980 the existence of microbial heating in F. polyctena mounds was confirmed in an experiment showing that in the absence of ants non-sterilized nest material generated a substantial amount of heat but sterilized nest material generated almost no heat³.

Microbial activity can be estimated by calculating the respiration rate of the nest material, which is used as a proxy for the respiration rate of the microbes living in it^3,15. There are detectable seasonal changes in the respiration rates of nest material with the highest rates found in summer. The mass-specific heat production of ants is higher than that of the nest material but, when considering the total mound volume, microbial heat production is more than seven times higher than the heat evolved by ants³. Ants can affect microbial activity via nest material aeration, supply of fresh plant material and their own metabolic heat production. Since the microbial activity of wet nest material depends strongly on temperature³, a temperature increase in some small parts of the nest (due to ant metabolism or sun radiation) results in an increase in microbial activity and consequently in a subsequent overall increase of nest temperature.

As first mentioned by Forel in the early 1920s the ant mound often serves as a solar collector²⁰. Solar energy can both increase the metabolism of ants and help heat the nest mound²¹. Compared to underground nests, mounds absorb heat more quickly both in the direct sun and in the shade⁹. Ants in the Northern Hemisphere usually remove shading grass from the south side of the mound so that the temperature increases quickly on that side. This creates a temperature gradient that many species use for brood displacement⁹. Other species decorate the mound surface with small pebbles or dead vegetation, which can work as heat collectors or as radiation reflectors²².

Mounds of some Formica, Solenopsis and Lasius species are asymmetric, with the main axis oriented in a south-north direction^23,24. In S. invicta the angle of the south slope of the mound is negatively correlated with the maximal sun angle²². Sun-influenced thermoregulation in termites has also been documented, for example in fungus-growing members of the genus Macrotermes²⁵. The amount of intercepted sunlight influences the shape of termite mounds and leads to great structural differences in nests in forests vs. savannahs²⁵. But the most admirable sun-induced nest shape differences can be seen in Australian “magnetic termites” Amitermes meridionalis²⁶. The nests are wedge-shaped with apparent north-south orientation which prevents overheating at noon and enables maintaining the warmth of nest in the evening.

Effective ventilation takes place as part of nest thermoregulation in many ant species, being regulated by the opening and closing of nest entrances (see below). A ventilation system in nests of the leaf-cutting ant genus Atta was described by Kleineidam et al.²⁷. There is not a thermal gradient big enough to generate thermal convection flow; rather the ventilation in Attini nests is driven by the wind. There are many openings on the nest surface, which are functionally divided into entry and exit tunnels. Wind flowing over the nest from any direction causes air to exit from the central tunnel and to enter tunnels at the nest periphery²⁷. This ensures exchange of respiratory gasses and optimal thermal conditions for symbiotic fungi which are damaged by temperatures higher than 30°C²⁸.

Behavioural reactions of ant workers

Behavioural reactions of ants are based on sensing temperature and temperature preferences⁵. To react to these gradients, ants have evolved inherent temperature preferences, which are the key element in thermoregulatory behaviour^4,29. Nurse workers are able to choose the optimal temperature for pupae production and sexual brood incubation and to move the brood along temperature and humidity gradients to achieve the best conditions for its development^5,29,30.

The optimal temperature range is variable for different groups of social insects, for example the brood of the honey bee Apis mellifera develops fastest at 35°C³¹. In Formica polyctena a temperature 29°C is preferred for pupal development¹⁴, whilst colonies of Solenopsis invicta can grow only between 24 and 36°C¹. In contrast the genus Myrmica is adapted to cold climates, M. rubra⁴ and M. punctiventris³² prefer temperatures between 19 and 21°C, about 8°C lower than the temperature preferred by other ants.

Temperature preferences can be affected by many factors including age and sex³³, working caste or feeding condition¹, or prior acclimation³⁴. Ant queens in Formica polyctena³⁵ and Solenopsis invicta¹ prefer slightly higher temperatures than workers, especially during the egg-laying phase; inactive queens may prefer cooler temperatures. Workers generally prefer lower temperatures, which decrease their metabolic rate and increase their lifespan^1,36. A decrease of 2°C can lengthen the worker lifespan in S. invicta by 14%³⁷. Preferences of nurse workers tend to be shifted towards the higher temperatures that favour brood development^1,4,21,29. Similar patterns have also been found in Apis mellifera^5,31.

Most ant species rely on brood translocation along temperature gradients as the main thermoregulatory strategy^{6,9,23,29,30,38}. The brood translocation usually has characteristic time rhythms, for example Camponotus mus follow a photoperiodic circadian rhythm. In the presence of a temperature gradient, nurse workers move the brood twice each day²⁹. The first displacement starts at 2 pm, when the brood is transported from the colder night location to a warmer day location. This movement occurs 6 h after sunrise, i.e., in the middle of photophase. At 10 pm (8 h after the first displacement and 2 h after nightfall), the brood is transported back to the night location. Under artificial light/dark day cycles the brood translocation rhythm changes according to the new photophase length²⁹. In Solenopsis invicta moving the brood up and down along temperature gradients does not seem to depend on the time of day or photoperiod⁹.

In response to temperature gradients, leaf-cutting ants from the genus Acromyrmex move not only the brood but also the symbiotic fungi which provide their food. The fungus requires high humidity and temperatures between 25 and 30°C²⁸. Acromyrmex ambiguus workers move the fungus garden according to humidity conditions, but they are also capable of nest humidity regulation by changing nest architecture. The flow of dry air into the colony is a signal for workers to plug ventilation tunnels to prevent nest from drying out³⁸. Similar behaviour has been proposed for termites³⁹.

When the nest interior becomes too hot, workers can reduce the inner temperature in several ways. In ants and termites nest cooling is usually achieved by changes in building behaviour. Workers of leaf-cutting ants in the genera Atta and Acromyrmex open tunnels to allow air circulation^27,30,38; this behaviour could be limited by a trade off for humidity control⁴⁰. Ants of the genus Formica can also partly remove the nest material, which reduces the wall thickness and increases heat dissipation¹⁶. In the genus Eciton, loosening of bivouacs´ structure (temporary nests similar to a honeybee swarm) is an effective way of cooling⁴¹.

Even more efficient ventilation systems providing both temperature regulation and respiratory gas exchange can be found in nests of the termite Mactorermes bellicose. These nests can have either externally or internally driven ventilation, depending on the habitat, nest shape (cathedral vs. dome shaped nest) or the time of day⁴². Cathedral shaped nests in open habitats are warmed by the sun, which creates a steep thermal gradient leading to convention currents in peripheral air tunnels. Dome shaped nests located in the forest rely more on internally driven ventilation. The same is true for cathedral nests at night⁴².

In contrast, bees, wasps and bumblebees cool their nests by wing-fanning and regurgitating water droplets. Water is spread over the brood comb surface enabling cooling through evaporation⁵. Cooling by water evaporation is very effective. Lindauer⁴³ placed a bee hive onto a lava plain where the surface temperature reached 70°C. Taking water ad libitum, the bees were able to maintain the hive temperature at the favoured 35°C.

Some social insect species have evolved highly elaborated systems of thermoregulation that enable to keep stable temperatures inside their nests. Bees, bumblebees and wasps are capable of direct incubation of selected pupae. Specialized workers sit on the surface of the brood cell and maintain their thoracic temperature over 35°C. Bees sitting inside empty brood cells can heat six brood cells at one time⁴⁴. Metabolic heat can also be used for protection, as shown in the interaction between the predatory hornet Vespa mandarina japonica and Japanese honeybees Apis cerana japonica⁴⁵. If the hornet attacks a honeybee nest it is surrounded by bee workers who increase their body temperature to a level which is lethal for the hornet but not for the honeybees⁴⁵.

Although ants cannot actively produce heat they are able to use indirect metabolic heat (i.e. heat produced as a by-product of metabolism) for ensuring brood development. This ability has so far been postulated for the Formica rufa group^3,15,46 and the army ant genus Eciton^41,47,48.

The Neotropical army ants Eciton hamatum and Eciton burchelli form temporary swarms called bivouacs, which can regulate temperature very precisely to ensure optimal conditions for developing brood⁴¹. According to Franks⁴⁸ bivouacs have a similar construction to a bee swarm; they can be divided into an outer mantel and inner core, which together maintain a stable temperature between 27.5 and 29.5°C regardless of the ambient temperature. On cold days the bivouacs change shape: they become more hemispherical to reduce the surface to volume ratio⁴⁸. Franks⁴⁸ postulates that all the heat required by the bivouac can be produced by ant metabolism.

Red wood ants of the Palearctic genus Formica (Formica rufa group) are supposed to use metabolic heat to maintain a heat core, an area with high and stable temperatures, in their nests. High temperatures are required for sexual brood development; nests producing sexual offspring always have higher temperatures than those producing only workers⁴⁶. The heat core’s position moves according to nest shape and size^3,15. The temperature inside nests of red wood ants begins to increase very early in the spring, even when the nest surface is covered by ice and snow⁴⁶. At this time some nests can contain larvae, pupae, and even some winged individuals, indicating that the inner heating must have started much earlier, because larval development cannot start in a cold nest and requires some weeks of constant warm temperature²¹. It is supposed that in large nests of F. rufa containing over 1 million workers, spring nest heating can start as an autocatalytic process⁴⁶ that relies on utilizing lipid reserves in young workers⁵⁰.

Another factor contributing to quick spring increase of nest temperature in red wood ants is a temperature intake by ant bodies. The Formica ants are dark colored so they heat up quickly when exposed to the sun during their outside-nest activities. In the spring ants are observed to create clusters on the mound surface as they bask in the sun⁵¹. Their bodies contain a substantial amount of water which has high thermal capacity making ant bodies an ideal medium for heat transfer. After getting hot enough the ants move inside the nest where the accumulated heat is released. This principle works throughout the year but in spring it is most obvious and supported by ant clustering on the nest surface^15,46.

Daily temperature fluctuations in the red wood ants nest seem to be correlated with temperature-dependent changes in ant density and ant aggregations in the nest. The highest nest temperatures usually occur in the afternoon or in the evening which corresponds with the return of foragers^15,46. This apparently results from the heat brought into the nest by returning workers (heat coming from absorbed solar energy) as well as the heat generated by worker metabolic heat production within the nest. In some nests the temperature drops slightly in the morning when ants leave the nest^15,16. Heat coming from the metabolism of foragers clustered in the nest center on cool days, when ambient temperature limits outdoor activities, could also explain a negative correlation between the inner nest temperature and the ambient temperature found occasionally in spring⁴⁶.

The seasonal fluctuations in the thermoregulation behaviour of Formica polyctena along a geographic gradient were studied by Frouz & Finer⁴⁹. Both in Finland and the Czech Republic the ant colonies maintained a high nest temperature (over 20°C) in spring and summer, for about 65–129 days. A rapid increase in the inner nest temperature in early spring was observed, mostly at the beginning of April. Annual nest temperature peaked in June in both locations and decreased gradually from August to November. Nest temperatures fluctuated more in the Czech Republic than in Finland, possibly because of greater differences between day and night ambient temperatures⁴⁹.

An interesting question is why F. polyctena maintain stable nest temperatures for the same period in Finland and the Czech Republic even though the length of the vegetation season and the ambient temperatures in Finland and Czech Republic are different. This might be explained by the reproduction cycle of the queen. Regular shifts between reproduction and diapauses has been documented in F. polyctena queens. The queen enters diapauses after 100 days of reproduction even at a constant temperature and photoperiod^35,52. It was postulated that ants maintain high temperatures only during the queen’s reproduction phase⁴⁹. After that the nest temperature drops, despite the fact that the outside temperature usually does not limit foraging and the ants are still active.

Use of metabolic heat for nest thermoregulation in ants

Maintenance of a high inner nest temperature has been observed in ants, especially in species which build above-ground nests or inhabit tree hollows^15,46,53. Heating of these structures is much easier than heating underground nests, because the surrounding soil has a large heat capacity and conductivity.

In moderate climates most ants build nests in the soil where the temperature is quite stable (Table 1) or on the soil surface under a layer of leaf litter where the temperature can be buffered by the insulating properties of the nest material. Many species in the Northern Hemisphere also nest under rocks or stones which serve as heat collectors (Table 1). In the tropics only a few species nest in soil and the majority of species inhabit small pieces of rotting wood⁵. More precise microclimate regulation is achieved in the mound-building species of the genera Atta, Acromyrmex, Myrmicaria, Pogonomyrmex, Solenopsis, Iridomyrmex, Formica, and Lasius (Table 1).

The maintenance of a stable temperature in red wood ant genus Formica nests during spring and summer is widely known and has been the subject of many studies^{12,16,46,49,53}. These ants are able to maintain thermal homeostasis in the nest because of the insulation and heat storage provided by the nest material^15,20. Mound size is generally correlated with the number of inhabitants^54,55. In small nests located in sunny areas the solar radiation and the insulation properties of the nest material are thought to be the key elements in nest thermoregulation^11,12. In larger nests there are internal sources of heat production, such as ant and microbial metabolic heat, that enable the maintenance of high inner temperatures even in a permanently cold environment^46,49.

Many other ant species (see Table 1) build hill-shaped nests from soil. These nests show large spatiotemporal variations in temperature and ants select the optimal temperature for brood development by brood displacement¹. These nests, however, do not have a stable heat core. The heat core where the temperature is stable and higher than ambient temperature for an extended period of time (several months) can be found only in a minority of ant species. Why is this the case?

The heat core exists in the nests/hives of winged insects that are capable of active heating, i.e. bees, wasps, bumblebees (thermogenesis in flying muscles), and some ants, especially the genus Formica^3,15,46, the army ant genus Eciton^41,47,48 and probably in fungus-growing termites⁴². But there is no evidence of a stable heat core in Solenopsis invicta or in Acromyrmex heyeri, which also build upper-ground hills from materials with suitable insulation properties such as soil or dead vegetation. As discussed earlier, thermoregulation can also be the result of nest architecture, and only those species that build nests with a low thermal heat capacity and a low thermal conductivity are likely to maintain a heat core. In summary, the use of metabolic heat production for maintenance of a heat core depends on the insulation properties of the nest and the size of the individual workers and of the entire colony.

The latest review about nest thermoregulation in social insects⁶ distinguished three types of ant thermoregulation strategies: “First, like many social bees, some ant species rely on protection from a cavity, such as a tree stump or underground burrows⁵⁶. Second some migrate their nest frequently, varying the amount of cover they select, depending on the temperature and season^57,58. Third some others move their brood to areas of optimal temperature within the same nest”²⁹. We suggest dividing the last category in the following way: a) ants that move the brood in daily cycles to places with optimal temperatures within the whole nest structure, for example Solenopsis invicta⁹ or Camponotus mus²⁹ b) ants that keep a stable heat core inside their nest and do not move the brood from the nest interior, including the ants from Formica rufa group^3,15.