Hi Cameron

This is a fun question. I do not work with fish or neuroacoustics so all I can venture is a guess. So for what it is worth

Looking at the lateral line info on wikipedia, fish uses hair that pick up acoustic velocity and/or acceleration. 

https://en.wikipedia.org/wiki/Lateral_line 

As you state, the fish bladder senses pressure & the two can potentially sense acoustic intensity.

However, to the best of my knowledge, intensity is defined only as a time averaged quantity and it then has both a real part and an imaginary part, where the real part shows propagating wave intensity, i.e.  I am not sure if there is such a thing as momentary sound intensity.

I believe the bladder is used to control boyancy and that sensitivity to noise may be a side effect, e.g. lorcas use a chirp (fast sine sweep) to excite fish bladders and momentarily stun the fish, i.e. a combined use of velocity and pressure sense is a possibility but not necessarily a must. (Humans sense acoustic infrasound velocity with out neck hairs which is used in theme parks to create a spooky feeling and this is very much a side effect.)

A fish would be interested in detecting sudden disturbances as these are potential threats. One may speculate that the lateral line may be utilised instead as a sensor array to detect from where a threat is originating and it may do so withhout  the acoustic pressure information, i.e. a bit like this.

http://www.acoular.org/ 

From an acoustic perspective, it makes better sense to use velocity than pressure in shallow waters as the acoustic pressure is at minimum at the water-air surface (where the reflection factor is close to -1), i.e. there is not much pressure near the surface to sense.  

Sincerely

Claes

Claes,

I agree exactly with the threat/prey model you described. my thought was, you have an anisotropic noise profile presented by the wind, rain and shipping noise. if an object with excellent laminar flow (a predator) comes up quietly, the background noise will still create a reactive flow across the predator. so if a prey fish is listening with directivity toward the back (in shallows this makes sense, deep they may listen more omnidirectionally), then the predator will become obvious in the reactive portions of the intensity, particularly as the predator gets close.

intensity averaging is just a matter of convenience. quite a bit of relevant data is lost in the trade off. I did find an article on Rg that considers the subject, see attached: https://www.researchgate.net/publication/258248605_Measurement_of_complex_acoustic_intensity_in_an_acoustic_waveguide

Article Measurement of complex acoustic intensity in an acoustic waveguide

here is a better reference, since this is the type of sensor I typically work with. this is analogous to what I am curious the fish can (possibly) perform:

http://scitation.aip.org/content/asa/journal/jasa/138/3/10.1121/1.4933587

I was interested to see your discussion on intensity.  The paper you cite was part of a project we did on complex intensity where we managed to measure both real and imaginary parts.  But complex intensity proved to be very difficult to measure accurately, at least with the equipment we had at our disposal.

Complex intensity separates the flow of energy into a component that quantifies the transport or propagation of energy (the real part), and a component relating to localised oscillations of energy (the imaginary part).  The local oscillations occur in regions where scattering or resonance occurs, often induced by impedance discontinuities and/or multiple sound sources interfering with one another.  That is, the imaginary component of intensity is normally relevant only in what one normally terms the acoustic “near field” where something complex is going on. 

Now I know nothing about fish, or how they process or sense sound waves, but I will have a go at trying to interpret the discussion above.  Background noise is likely to be largely real valued if we assume that a fish is a reasonable distance from the source of background noise; that is, the fish is in the acoustic far field of the background noise.  However, if this background noise interacts with an object such as a predator, then it is possible that scattering from the predator will create some imaginary components in the acoustic near field of the predator.  Thus, if the fish is close to the predator (in the same acoustic near field) then it may be able to sense/process the imaginary component of intensity caused by this scattering.  Alternatively, the predator itself may produce sound (if this does indeed happen), in which case this is also likely to generate imaginary components of intensity in the near field surrounding the predator.

In both cases this complex intensity field is likely to be very complicated, but it will be unique to a particular animal (on the basis that the imaginary part is a function of the components generating the intensity field) and so it may be possible that a fish could use the imaginary part of intensity to decide the type of predator that is close by.  Of course, this will depend on very good velocity sensors and the ability to process and interpret what may be very complicated sound fields. 

I think they'd be close in most situations. The lateral line would feel the wake flow and presumably see some shift in the ambient noise relative to it. Thanks for corroborating the theory, I think I'll take a look at it now.

Dear all, 

Sorry for late reply. I have been away & I have not hade the time to scan you papers.Promise to take a look later though as this is a fun topic.

I know nothing about fish either but I am happy if my answer makes sense also from other aspects. Inspired from the ab ove, I will have another stab at reasoning from purely acoustic aspects and then throw in some speculatory remarks from shear ignorance on the topic. So, for what it is worth. 

Active intensity is the way energy flows from source to consumer (dissipator). The largest dissipator underwater would be transmission as dissipation is weak in water for a long list of reasons. 

Reactive intensity is not that much discussed as it hard to interpret and as it can be many things, e.g. standing waves in a room, a near field close to a source and so on. 

I imagine that if you measure velocity and are close to the water surface which heave and have surface waves, i.e. very strong velocity excitation, there should be both a very strong reactive near field and a decent real valued instensity component. In shallow waters, this would be mostly in the plane directions. 

If we measure pressure in a strong reactive field, we do not sense much when we are far away from the source, i.e. further away than one quarter wavelength. For a (strongly active) field with a clear direction, we can sense if a source is reflected or shielded. 

If we measure velocity, I guess, the situation would reverse, i.e. that what applies for the active field and pressure applies for velocity and the reactive field as velocity is 90 degrees phase lagged wrt pressure.

However, there is a fundamental difference between pressure, which is scalar, and acoustic particle velocity in that the latter is a XYZ vector and in that it does not transport mass but it does oscillate mass. (We know this from flow acoustics. ) The lower the frequency, the larger the velocity. 

I read on wikipedia that some fish have hair, a bit like human neck hair which senses particle velocity. I imagine that such hair would make an excellent flow change detector (antenna) when the reactive oscillating flow is modulated, weakened or simply changes direction. The hair wouild then sense a change of oscillating drag at least wrt direction and strength. 

So, in this sense, what you write about a predator being detected from reflection or shielding seems to make sense. Doing it at low frequency would enable detection at large distance. Thinking about it, the fish hair sensitivity should be tuned to typical frequencies associated with predator swimming. 

A few years back, I talked with a very senior guy on infrasound and he told me that once you start measuring and thinking about it, you train your sense to pick it up. He told me - I readily sense infrasound from bridge gaps (where the gap is compressed by a rubber cover when loaded by vehicles) that are located very far away, as well as when bus drive over a speed bump. I believe him, he was talking in earnest and knew most listeners would be sceptic. 

In any case, to sum things up. Velocity yes, reactive, yes - using sound to detect frequency the way our ear does - not necessarily, as I am guessing that would be overengineering the task and therefore, intensity, makes me a bit sceptic.  It sounds elaborate as compared with using an array of hair cells. 

Perhaps this is something one can set up a small experiment in a tank and make a horror show for the fish the same way we get a spooky feeling when our neck hair senses strong acoustic particle velocity at infrasound frequencies? Tap into the fish neuron's and measure its firing when sensing a change in the reactive intensity should be doable in a tank using speaker excitation. 

https://www.youtube.com/watch?v=E7Bkrypxzs4

SIncerely

Claes

 Hi again

I guess, a flexible straw fitted with piezoelectric film that pick up its bending in various directions should be a way to mimic a fish hair? I read on wikipedia that the fish har has feedback to counteract flow (from actual net transport of matter) which could be thrown in there as well if one would want to. Just a thought. 

The above listed youtube link shows an ainfrasound speaker for use in air. We sense pressure variation from infrasound but not its tonality + we sense particle velocity using our neck hair.

Why would fish not do something similar, in particular, if it is a matter closely realted to survival? 

/Claes

Dear fellows,

I've been watching your discussion related to "Do fish measure reactive intensity?" and found it interesting when talking about "total sound intensity from a pure acoustic point of view". When it comes to sound reception/sensing in fish you've forgot the "most" important sensor beside the swim bladder and the lateral line, i.e. the otoliths, which act like an accelerometer. There is some literature related to this which may clarify/elucidate sound reception in fish in relation to total sound intensity. I'll like to see you++ continue with this discussion.

Hi John

The bladder would be a poor receptor at shallow waters as sound pressure is at its minimum near the sea surface. 

I am no fish expert but wikipedia tells me that the lateral line contains hair cells. My thought was that these would be quite a good sensors for picking up acoustic particle velocity, i.e. the oscillating flow associated with a reactive acoustic field of the kind one would expect for shallow water and wave motion.  

The otoliths is new to me. Wikipedia tells me it senses gravitation and acceleration. If so, its response should be 180 degrees phase shifted to acoustic pressure, i.e. as poor a choice as the bladder at shallow water conditions, but all the more important at greater depths, i.e. well away (a number of wavelengths, say 3-4) from the sea surface. Also, the sensor is located inside the fish head which suggests it would be poor at picking up anything related to particle velocity. Again, purely from acoustic reasoning.

If the otholith is used for balance, I imagine being insensitive to particle velocity would be a positive feature at shallow water conditions.  

Taking a 2nd look at the lateral line, the hair cells are located in canals that connect via multiple holes to the fish exterior. My guess is that this system serves the same kind of function as the fluff does around a microphone, i.e. getting rid of the influence from wind/transport of matter. Also, I see at wikipedia that hair cells sensitive to velocity react mainly at frequencies below 30 Hz, which make sense when trying to detect motion made from swimming. 

/Claes

Yes, but the total hearing system may be necessary to understand the relevance of what the fish is actually capable of interpreting. Think of it as a human ear that stages it's linear/non-linear filtering. Then consider how real and imaginary shift might fuse to provide a single picture of intensity. Just saying for consideration. This would be a bio and env problem simultaneously.