How Pangolins Choose Which Termite Mounds to Raid

A pangolin moving through savanna or forest on a foraging night does not dig indiscriminately. It pauses, sniffs, moves on, pauses again, and eventually commits — claws driving into a mound or fallen log with extraordinary force. The choice of which mound to excavate and which to pass is a foraging decision shaped by millions of years of co-evolution with social insects. Understanding how pangolins select prey colonies reveals as much about cognition and sensory ecology as it does about nutrition.

The Foraging Challenge

All eight pangolin species are obligate myrmecophages — they eat almost exclusively ants and termites. This dietary specialisation is extreme: pangolins lack teeth entirely, cannot process plant material or vertebrate prey, and have digestive systems tuned specifically to chitin-rich insect prey. Their survival depends entirely on reliably locating and extracting sufficient quantities of colonial insects night after night.

This creates a specific optimisation problem. Termite and ant colonies vary enormously in size, in the biomass of edible workers and soldiers per unit volume of mound, in the hardness and architecture of the defensive structure surrounding them, and in the chemical defences that workers can deploy against intruders. Not all mounds are worth digging. A pangolin that expends significant energy excavating a hardened mound for a depleted or chemically-defended colony may achieve a net energy loss from that foraging bout.

Field researchers using GPS telemetry have documented that pangolins typically visit between 3 and 30 foraging sites per night, spending highly variable time at each. Some sites receive only a brief olfactory inspection before the animal moves on. Others receive sustained excavation lasting 15–40 minutes. The selection process — what determines which mounds get dug — is driven primarily by chemosensory assessment.

Olfactory Assessment Before Excavation

Before committing to excavation, a pangolin performs a characteristic sniffing sequence. It approaches the mound’s base, places its snout close to the surface, and samples volatile chemical compounds emitting from colony activity below. The assessment may last from a few seconds to several minutes.

What the pangolin is detecting is not well characterised at the molecular level for most species, but the behavioural evidence points strongly to colony activity level as the key variable. Active termite mounds produce a continuous flux of volatile organic compounds (VOCs) associated with worker respiration, brood pheromones, and fungal cultivation in Macrotermes and other fungus-growing termite species. A dormant or depleted colony produces substantially less chemical signal.

Field observations in southern Africa have noted that pangolins consistently reject recently-raided mounds — those that have been partially excavated within the previous few nights — even when no visible external damage marks the location of the prior raid. This suggests the animal is detecting reduced colony activity through olfactory cues rather than relying on memory of where it has foraged (though memory-based spatial avoidance may also contribute).

The pangolin’s olfactory apparatus is correspondingly well-developed. Post-mortem neuroanatomical studies confirm large olfactory bulbs relative to overall brain size, and the nasal turbinate bones — the scrolled structures that increase olfactory epithelium surface area — are complex and extensive. The vomeronasal organ (Jacobson’s organ), which processes non-volatile chemical signals in contact chemoreception, may also contribute to ground-contact assessment at mound bases.

Prey Selection by Colony Type and Species

Pangolins are not indifferent to colony type. Diet studies using prey remains and gut content analysis have documented species-level prey preferences that vary by pangolin species, geographic range, and season.

In South Africa, Temminck’s ground pangolins (Smutsia temminckii) have been documented feeding predominantly on one or a few ant genera — particularly Anoplolepis, Camponotus, and Carebara — rather than the taxonomically diverse insect communities available in the same habitats. This selectivity is inconsistent with purely random exploitation and suggests active assessment of prey quality.

Termite versus ant choice varies by season and habitat. In dry season conditions, when ant activity is reduced near the surface, pangolins show increased reliance on termite mounds, including the hard, permanent mounds of Macrotermes and Trinervitermes species that require more excavation effort but contain large biomass. In wet season conditions, when ant colonies are highly active and surface-accessible, pangolins shift to shallower foraging in grass root zones and leaf litter, reducing excavation time per foraging bout.

The size of the colony matters as well. Larger colonies — identifiable externally by mound height and surface area in mound-building species — contain more individuals per unit depth and allow a pangolin to extract more biomass per minute of excavation before colony defence mechanisms force withdrawal. Pangolins appear to preferentially raid large, active mounds rather than small satellite structures, consistent with energy-maximising foraging theory.

Dealing With Chemical Defences

Ants and termites do not passively accept being eaten. Many species deploy potent chemical defences: formic acid in ants of the subfamily Formicinae, quinones and other compounds from glands in termite soldiers, and physical biting and stinging in aggressive species. Pangolins have anatomical and behavioural adaptations to manage these defences.

The most important anatomical defence is the eyelid system. Pangolins close their thick, reinforced eyelids during active excavation and feeding, protecting the eyes from formic acid spray and mechanical damage. Ear openings can be contracted. The nostrils close between tongue insertions. These sequential closures allow the animal to protect sensory organs while still feeding at depth inside an active colony.

Pangolins also manage exposure duration. At most foraging bouts, the animal extracts insects for a period of one to several minutes per entry point, then withdraws. This intermittent extraction — observed across African and Asian species — is thought to reduce the accumulation of chemical exposure and also to allow the tongue to return insects to the mouth (pangolins cannot chew at the insertion point; they withdraw the tongue loaded with prey and swallow at the surface).

Some ant species — particularly aggressive, biting Ponerine ants — appear to be avoided entirely. Gut content and faecal analysis from wild-caught pangolins consistently underrepresents these species relative to their habitat abundance. Whether this reflects active olfactory rejection prior to excavation or learned avoidance is not yet established experimentally.

Spatial Memory and Mound Rotation

Termite and ant colonies can recover from pangolin raids, but require time. Worker populations depleted by a significant feeding event may require weeks to restore normal activity levels in the excavated zone. Repeated excavation of the same mound at short intervals risks creating a depleted resource that yields diminishing returns.

GPS telemetry studies in Africa have documented that pangolins use spatially distributed foraging circuits within their home ranges, typically avoiding recently-visited sites. The mechanism — pure olfactory assessment of current colony state versus explicit spatial memory of foraging history — cannot be resolved from field observations alone. But the outcome is consistent with an optimal foraging strategy of distributed exploitation that allows recovery time.

Colony recovery rates depend on colony size and species. Large Macrotermes mounds, with millions of workers, recover surface-zone worker density relatively quickly — possibly within days. Small ant colonies may require weeks. This difference may partly explain why pangolins in termite-rich habitats show faster revisit rates to individual mounds than those in ant-dominated habitats.

Implications for Captive Management

Understanding wild foraging selection is directly relevant to the persistent problem of maintaining pangolins in captivity. Wild-caught pangolins frequently refuse food in captivity, and even those that begin eating substitute diets often develop nutritional deficiencies and abnormal foraging behaviours. Part of the problem may be that captive feeding presents food in ways that bypass the assessment behaviours the animal relies on.

Research groups working on pangolin rehabilitation have experimented with presenting food in conditions that require olfactory assessment and physical foraging work — burying larvae and insects in substrate, hiding prey in wooden structures that require manipulation, and presenting freshly active live ant and termite colonies where available. Preliminary observations suggest improved feeding engagement and reduced stereotypies compared to direct food presentation. The underlying principle is that a pangolin optimised by evolution for detecting and extracting insects from complex structures may be poorly stimulated — and may not recognise food as food — when it is presented without the chemical and structural cues that normally trigger feeding behaviour.

Frequently Asked Questions