Tom Fink –
Lichuan Gui, Yong Wang, Zhonghua Cao, Adarsh Jaiswal, Orwa Tahaineh, Vijay Ramalingam, Roger Hasse, *Alan Lax, John Seiner
National Center for Physical Acoustics
1 Coliseum Drive
University of Mississippi
University, MS 38677

1100 Robert E Lee Blvd.
New Orleans, LA 70124

Popular version of poster paper 5aAB7 presented at the 152nd ASA Meeting in Honolulu, Hawaii.


Figure 1. Formosan Subterranean termites: Soldiers have the dark amber colored heads and dark pointy mandibles. The rest of the termites seen here are workers.

Subterranean termites are very weak and fragile insects.  When placed in a small open petri plate termites can not escape and will  rapidly dry out and die.  Ants and other predators can easily overpower exposed termites.  To avoid these perils subterranean termites tunnel in the ground and wood and wall everything out.  Thus to a subterranean termite any breach of their tunnels or carton nest (a structure made of undigested cellulose, mud and termite saliva and punctuated by a maze of tunnel-like galleries) is a cause for alarm.  When the Formosan subterranean termite (FST) (Coptotermes formosanus) and the native subterranean termite (RF) (Reticulitermes flavipes) detect a potential breach, the soldiers will usually bang their heads apparently to attract other soldiers for defense and to recruit additional workers to repair any breach.  In the video below (Video 1) taken inside a Camphor tree infested with a large population of FST you will see a soldier on the right banging its head very hard on the carton nest substrate while an adjacent soldier on the left tremulates (rapidly moves forward and backward without moving its legs).

Video 1: Click above to see video of FST head-banging and tremulation in a Camphor tree.

Head-banging results in the strongest substrate vibrations/loudest sound that termites make.  The sound of a large number of FST soldiers head-banging in a large River Birch tree in New Orleans can be heard in Sound Clip 1This is a much more amplified vibration/sound than caused by chewing (see Sound Clip 2, from the same tree and recorded in the same way with the same settings).  Chewing in most trees and structures is usually much more reduced in amplitude than recorded in Sound Clip 2.

Head-Banging Details

In the second video below (Video 2) you will see a view inside an FST infested Black Willow tree.  The central channel is the drill hole that allowed entry of the borescope video camera, while to the sides is the carton nest.  A soldier can be seen vigorously head-banging for 2.5 minutes.  Near the end of the clip another soldier joins, even more vigorously.  Notice also that most of the soldiers are NOT head-banging.  Perhaps some soldiers are more inclined to head-banging than others.

Video 2 (without sound): Click the picture above to see video of FST head-banging in a Black Willow tree.

In Video 3 below you are looking into a natural carton nest in a Slash Pine tree heavily infested with FST.  No soldiers are seen head-banging in this video clip but three workers are (indicated by a red and yellow star).  While the soldiers (Videos 1 and 2 above) flatten their heads and hit the underside of their heads, workers hit the substrate with their mandibles as they do not flatten their heads at all.  Workers also only head-bang briefly.  The head-banging sound you hear in Video 3 is from soldiers elsewhere in the carton nest.

Video 3: Click the picture above to see video of FST workers head-banging in a Slash Pine tree.

In Video 4 two RF soldiers are observed to be  head-banging vigorously.  Despite having larger heads than FST, RF head-bangs faster.  RF soldiers will readily head-bang outside their nests and galleries when provoked, while FST rarely do so.

Video 4 (without sound): Click the picture above to see video of two RF soldiers head-banging.

In Video 5 below you will hear a single RF soldier head banging for a long time.  The average time between successive head-bangs in this soldier is 0.044 seconds (average of 31 inter head-bang periods).  RF soldiers typically head bang in bursts of 3-4 head bangs.  This pattern of 3-4 bangs in a burst is often easy to recognize with the ear.  FST usually do not head-bang in bursts but rather have a steady uniform head-banging pattern.

Video 5: Click the picture above to see video of the waveform of one RF soldier head-banging.

In the Video 6 below you will hear a single FST soldier head-banging (the clip will repeat four times).  The average time between head-bangs in this clip is 0.066 seconds.  Often the time between head-bangs in FST is longer.

Video 6: Click the picture above to see video of the waveform of one FST soldier head-banging.

Further Details of Head-Banging As Observed with High-Speed Video

Because FST and RF head-bang so fast the details can not be seen with the eye. Using high-speed video we can see the details as shown in this video of RF shown in slow motion (Video 7, original video taken at 10,000 frames per second, and shown here at 30 frames per second). In both species, RF and FST, the head is raised up by moving the prothorax up at its articulation with the mesothorax (the head is attached anteriorly to the prothorax, which possesses the first pair of legs, the mesothorax contains the second pair of legs and is attached to the prothorax by a movable junction). The legs are not used to raise the head. The head is then rapidly moved down like a battering ram and the head hits flat on the substrate. What can only be seen in slowed down views of high-speed video is that the head often rebounds, or bounces up after the initial impact and then hits the substrate again in about 4 milliseconds. This can be seen on head-bangs 1 and 3 in video 5. This rebound substrate hit produces an additional substrate vibration.

Video 7: Click the picture above to see video of one RF soldier head-banging.

While similar in body size (but larger in head size), the native termite generally raises its head higher (up to 2 mm) than the Formosan termite (about 1mm). In order for Reticulitermes to head-bang faster (shorter duration between successive head-bangs) it must accelerate its head faster and show a higher velocity than for FST specimens. Upward and downward head velocities in the FST specimen are respectively 100 and 240 mm/sec, while the values for Reticulitermes are over 200 and about 400 mm/sec. Similarly acceleration values for FST soldiers are about 400 m/sec2 vs up to 700 m/sec2 for Reticulitermes.


Practical Use of Head-Banging by Humans

Because termite head-banging results in the loudest and most diagnostic sound that termites make, it can be utilized for termite detection. The most difficult part of termite control is termite detection due to their subterranean and hidden tunneling habits. A useful sensor for termite vibration (due to head-banging and chewing) is a 1000 mv/g accelerometer (Figure 2). Termite vibrations are transmitted to the accelerometer sensor by a waveguide (bolt, screw, metal rod, etc.) which causes the piezoelectric material of the accelerometer to produce an electrical signal in proportion to the amplitude of the vibration. These electrical signals are then amplified by a signal conditioner and transduced into sound that can be heard through headphones or recorded.

Figure 2: Accelerometer, waveguide, signal conditioner and headphones used to detect head-banging of FST infesting the floor below the linoleum tiles.

Detection of subterranean termites by head-banging, especially FST (due to their higher percentage of soldiers per colony, approximately 10%, than RF at approximately 1%) is especially useful for trees. FST infested trees possess a carton nest at the base of the tree.  Placing a waveguide in a tree crotch gets close to the carton nest/galleries and induces head-banging.  A total of 1957 trees were examined in New Orleans City Park visually (termite mud tubes, wood damage, termite mud packs at tree base, observing termites while removing soil at the base of the tree) and acoustically (using a 1000 mv/g accelerometer and a 35 cm long metal wave guide inserted into all tree crotches if necessary).  Five hundred twenty-eight trees or 27% were infested with FST.  Fifty one percent were missed by visual methods alone while only 3% of trees were missed acoustically (but termites were found visually).  Head-banging is so distinctive and often loud that “listening” to termites in noisy urban environments is not usually a problem.  In Jackson Square, New Orleans head-banging in two Aristocratic Pear trees could easily be heard above intense urban noise including the sounds of street performers and their cheering crowds (see Figure 3 below and associated Sound Clip3).

Figure 3: Two heavily FST infested Aristocratic Pear trees (tree 86 is distal to 85 alongside the bench) in Jackson Square Park in New Orleans, November 2003. Both of these trees appeared not to be infested by visual methods but loud head-banging was detected acoustically.

Sound clip 3 plays termite head-banging in tree 85 along with background sound from the crowd cheering for the street performers

Selected Representative Literature

Connetable, S., A. Robert, F. Bouffault and C. Bordereau. 1999. Vibratory alarm
signals in two sympatric higher termite species: Pseudacanthotermes spiniger and P.
Militaris (Termitidae, Macrotermitinae). Journal of Insect Behavior 12: 329-342.

Kirchner, W.H., I. Broecker and J. Tautz. 1994. Vibrational alarm communication in the
Damp-wood termite Zootermopsis nevadensis. Physiological Entomology 19: 187-190.

Mankin, R.W., W.L. Osbrink, F.M. Oi and J.B. Anderson. Acoustic detection of termite infestations in urban trees. Journal of Economic Entomology 95: 981-988.