Signal generation - hearing and processing Flashcards
Insect ears - the basics
Insects have tiny bodies, so can’t use interaural time differences for directionality.
Instead, some have one connected ‘pressure difference receiver’, with a tympanum on either side of their body, so it’s inherently directional.
Mechanosensory neurons arranged in scolopidial sensilla carry deflection information from thin elongated hairs on e.g. cerci of crickets or antennae of mosquitoes. These hairs can detect particle movements from sound pressure waves. These movements are vectorial, so these mechanosensors are inherently directional
Note that pressure and velocity maxima are out of phase - peak response from afferents contacting tympanum will be out of phase with peak response from afferents contacting hairs on cercae or antennae.
Frequency discrimination - brief comparison of grasshoppers and crickets
Grasshoppers have a sort of tonotopic map on the tympanum itself - different frequencies will cause place-specific oscillations in the membrane, which are picked up by one of an array of neurons contacting it.
Crickets make use of a travelling wave in the crista acustica, attached to the auditory trachea, with neurons tuned to particular frequencies.
Common feature: it’s the peripheral biomechanical processing that does the discriminating
Grasshopper ears in more detail - where are they, how do they work, what’s the processing pathway?
Sit at abdominal segment A1, connected through the body by air sacs.
Only low frequencies 4-7kHz can pass through the body, and the ears act as pressure difference receivers then
Above 10kHz, sound is heavily damped as it passes through the body, so the ears act as pressure receivers
The first stage of processing is the auditory ganglion, or ‘Muller’s Organ’, which sits on the tympanum and consists of 60-80 scolopidial sensilla, aka auditory afferents. The dendrites of these different afferents (a+b, c- and d-cells) contact different places on the tympanum, so when a part oscillates due to a particular pressure, the dendrites bend and fire an AP.
Travelling waves through the tympanum are frequency-dependent. Different parts of membrane respond best to different frequencies, which correlates with tuning curve of afferent contacting that part
Afferents go to the auditory neuropil in the metathoracic ganglion, where they contact ascending interneurons.
Cricket ears in more detail - where are they, how do they work, what’s the processing pathway?
Cricket auditory organs are on the tibia of the prothoracic legs. Having them far apart like this aids localisation.
There is a tympanum, but afferents do not attach to it. Instead they attach to the auditory trachea - sound enters the auditory trachea via auditory spiracles, high up on legs to catch more sound.
Both legs are connected, so inside and outside pressures interact to allow pressure difference receiver behaviour.
Instead of Muller’s organ, there is the crista acoustica, where 40-60 auditory afferents are linearly arranged.
Tonotopic arrangement established in crista acustica, continues in auditory neuropil in prothoracic ganglion. Those responding to highest frequencies are furthest from the ear, lowest are closest. Detectable range is 2-30kHz, but most are tuned to about 4.8kHz, the frequency of the cricket calling song.
Travelling wave will activate those afferents tuned to its frequency, which go on to the prothoracic ganglion and connect to ascending interneurons
E.g. AN1 responds to the calling song (tuned to 4.8kHz)
E.g. AN2 responds to the high frequency calls of bats
Processing of directionality - how’s it processed? How well can grasshoppers do it? How’s it processed? What’s their response like?
Grasshoppers use pressure difference receivers - so sound travels through the body and acts on the opposite tympanum, causing ‘destructive interference’ dependent on frequency.
Determining directionality in grasshoppers is normally based on differences in firing patterns of neurons on each side (spike number, latency and recruitment)
Maximum left-right difference, when song is from 90°, corresponds to 8dB (i.e. you’d need to increase song by 8dB to get that response at the other ear).
Grasshoppers are good at detecting directionality (and can tell distinguish two songs presented one on each side when they’re only 1dB amplitude [90% correct turns] or 0.5ms arrival time different)
Auditory interneurons are stimulated by ipsilateral stimulation and inhibited by contralateral
Auditory afferents are all excitatory; lateralisation is enhanced by reciprocal inhibition
Grasshoppers are not so good at detecting angle of directionality - no correlation between angle of target and angle of turn when female call is limited to 400msec. But in the wild, female call is 1s, and male corrects overshoot whilst the call’s still happening. They tend to jump when sound is nearly directly in front or behind.
So phonotaxis is a turning phase, then a short walking phase, until they hear another response (remember calling is bidirectional in grasshoppers)
Primitive (pneumorid) grasshopper ears
No clear tympanum, afferents attach directly to body wall
Threshold of 15dB at 4kHz, lowest threshold found in insects
Chordotonal organ instead of Muller’s organ
2000 afferents, instead of 60-80!
Using frequency discrimination for sex-specific recognition - what do female vs male songs sound like? what are female vs male song preferences?
Female grasshoppers’ songs have only a low frequency peak, at 5-10kHz.
Male grasshoppers have 5-10kHz peak, and 15-30 kHz peak.
Female sound preference is highest for songs with both high and low frequency components
Male sound preference is highest for songs with only low frequency components
Functional significance points to note
In crickets, there’s a narrow frequency band that they need to pay attention to, so hearing is tuned both in sensitivity and directionality to the frequency range of the calling song. Note that this frequency range is separate from the frequency ranges of predators - frogs and toads have lower frequencies, bats have higher - allowing labelled lines and the appropriate behavioural response (i.e. the IRM won’t be released when it hears a predator)
Grashoppers are mostly non-resonant, so have a broad frequency range
Simply having sound arrive at both sides of the tympanum does not guarantee directional hearing, they must be in the correct phase relationship, as determined by the anatomy of the insect and the frequency of the sound.
The connection between the two ears that goes through the body can be modified, e.g. by air sacs in grasshoppers, by a double membrane at the midline or acoustic vesicles in crickets. These are important for altering gain, delay, amplitude and phase relationship. They’re tuning devices.
Ears in insects - who has them?
Romer 2015:
Insects that communicate with acoustic signals have ears, for conspecific communication and recognition, and as a barrier to hybridisation.
But insects that do not generate acoustic signals have also developed ears, suggesting an importance in avoiding predators.
Ears are thought to have evolved 19 different times in insects, on a variety of body parts, including legs, abdomen, mouth parts and wings.
Three dimensional directionality
Tree crickets and katydids need to localise sounds from above and below, too.
Ofner et al 2007 - Katydids localise accurately to angles 60° above and below the azimuth, but with less accuracy as the angle increases, meandering more and going round in circles.
It’s not known how they do this, but some tilt their body up and down, perhaps to bring the angle closer to the azimuth for better localisation
Bush crickets, and lateral inhibition
Bush crickets have ears on prothoracic leg, with the trachea extending into the body a bit. Two symmetrical tympani (unlike cricket), and spiracles on the mesothorax. The tracheae are not connected (unlike cricket and grasshopper), so all directionality is via comparisons. Still have crista acustica with travelling wave and tonotopic map.
Reciprocal inhibition between Omega Neurons (just like in crickets) allows interaural intensity differences of 1dB to be detected, and separation into two hemispheres of sound landscape. ON also exhibits an intensity-dependent inhibition over seconds which leads to preference for the louder signal, to aid interaural intensity differences, though it’s not known how this happens. Crickets do this too.
AN1 is also inhibited by the contralateral ON, which aids directionality but also acts as gain control in neurons with a saturating intensity-response curve. AN1 is tuned to carrier frequency, and tuning is sharpened by frequency-dependent inhibition (lateral inhibition?).
Frequency-dependent inhibition means that the intensity-response curve is different for different frequencies, which may mean the neuron stops firing at v high intensities of a conspecific’s call, e.g. when the female comes very close. At this distance, vibratory components may be more important.
