Bat echolocation is all about catching a meal.
After all, if bats could safely feed during the day, they wouldn't bother to stumble around in the dark of night.
So its good to catch the occasional ultrasonic recording of bats catching their prey.
This is one of my best recordings so far of "the feeding buzz" of a common pipistrelle.
From the spectrogram below (lower trace) we can see that each echolocation call covers a frequency band with a dominant frequency (known as FmaxE) of approximately 46kHz. This is denoted by the whitest section of each call.
As the bat locates some kind of flying insect, the spacing between calls reduces from around 40ms to less than 10ms. This is the transition from the "search phase" to the "approach/tracking phase".
The frequency initially rises, with a typical FmaxE just above 50kHz. I don't know whether the signal level is reduced because the bat is moving or turning away, or whether the bat is reducing the ultrasound level so as not to alert the prey (...I strongly suspect the latter).
When the bat is really close to its prey, the bat transitions from "approach" to "attack" phase, and the frequency drops and the ultrasound level is restored. FmaxE for the final 4 "attack" calls is only 36kHz.
Was it a successful catch or a near miss? I've no idea.
The last 4 "attack" calls (i.e. those at 36kHz) occur in a 20ms period. If we assume that the bat is flying at (say) 4.5 metres per second, then these calls may represent the bat closing in from a distance of about 90mm from its target. (i.e. 4.5m/s x 0.02s = 0.09m). So the bat must know precisely where the insect is, right up to the last millisecond.
Echolocation theory states that the wavelength of the signal used must be less than the 'size' of the target. From reading several articles, it looks like 'size' relates to the circumference of a disc. And naturally, a hard (acoustically reflective) disc is better than a soft disc.
The circumference of a circle is equal to diameter x 3.142, and the speed of sound in air at 10'C is approximately 337m/s. So I calculate that a bat echolocating at 42kHz could detect an insect whose body encompasses a circular area with a diameter of 2.5mm. However, little flying bugs are not hard objects.
Apparently pipistrelles feed on creatures as small as mosquitos, so whatever the theory predicts, they get by.
From the spectrogram below (lower trace) we can see that each echolocation call covers a frequency band with a dominant frequency (known as FmaxE) of approximately 46kHz. This is denoted by the whitest section of each call.
 |
| plots of amplitude/time and frequency/time (spectrogram) |
As the bat locates some kind of flying insect, the spacing between calls reduces from around 40ms to less than 10ms. This is the transition from the "search phase" to the "approach/tracking phase".
 |
| this spectrogram illustrates the bat call hunting sequence: (1)Search, (2)Approach/Tracking, (3)Attack |
The frequency initially rises, with a typical FmaxE just above 50kHz. I don't know whether the signal level is reduced because the bat is moving or turning away, or whether the bat is reducing the ultrasound level so as not to alert the prey (...I strongly suspect the latter).
When the bat is really close to its prey, the bat transitions from "approach" to "attack" phase, and the frequency drops and the ultrasound level is restored. FmaxE for the final 4 "attack" calls is only 36kHz.
Was it a successful catch or a near miss? I've no idea.
The bat quickly resumes echolocation calls (within 100ms), but I guess they can chew and call at the same time.
The last 4 "attack" calls (i.e. those at 36kHz) occur in a 20ms period. If we assume that the bat is flying at (say) 4.5 metres per second, then these calls may represent the bat closing in from a distance of about 90mm from its target. (i.e. 4.5m/s x 0.02s = 0.09m). So the bat must know precisely where the insect is, right up to the last millisecond.
Note: 4.5m/s has been quoted as a typical average foraging speed for a common pipistrelle (Jones and Rayner, 1989)
a little theory
Echolocation theory states that the wavelength of the signal used must be less than the 'size' of the target. From reading several articles, it looks like 'size' relates to the circumference of a disc. And naturally, a hard (acoustically reflective) disc is better than a soft disc.
The circumference of a circle is equal to diameter x 3.142, and the speed of sound in air at 10'C is approximately 337m/s. So I calculate that a bat echolocating at 42kHz could detect an insect whose body encompasses a circular area with a diameter of 2.5mm. However, little flying bugs are not hard objects.
Apparently pipistrelles feed on creatures as small as mosquitos, so whatever the theory predicts, they get by.
No comments:
Post a Comment