Peter gave us a good summary of the 'hearing' issue.

Some insects have true hearing, much like ours.  I did my Ph.D. research on grasshoppers.  They have a tympanum which is much like our own ear drum.
 
Bee specialists have gone from the notion that bees don't hear, to an understanding that bees have sensors that pick up vibrations - and we're slowly beginning to realize that this may involve substrate and airborne sounds.  However, the key point is that bees recognize what we'd call sounds.   Consider queen piping, and Tom Seeley has observed bee head butting accompanied by a recordable sound - the two signals apparently convey a STOP message to a dancing bee.   We learned years ago that  we can STOP all of the bees on as many as 5 frames and hold them in that mode for up to two minutes using sound.  So they obviously 'receive' and recognize some sounds (whether airborne or substrate borne).
 
Many of you know that about a dozen years ago, we started studying the sounds produced by colonies in response to air-borne, toxic chemicals (for the  US military).   After 6 years, we had a system for using the sounds that bees produce to warn of the presence of poisonous chemicals in less than a minute - often instantly, most in less than 15 seconds.  Moreover, we could tell by the sonograms the type of chemical - different categories of volatile organic chemicals induced unique sounds.  We were honored to receive a nomination for an Innovation Award from the military.  Didn't get the award, but honored to be nominated.
 
About 7 years ago, we extended our honey bee sound research to  looking at the sounds colonies produce when they are queenless,  ready to swarm, infested with varroa (and the level of infestation), infested with hive beetles, infected with Foul Brood (and the levels of infection), infested with Nosema ceranae, as well as the race or species of bees (AHB, Apis cerana, possibly Cape Bee).  Last January we finally had a hand-held scanner that we started to Beta-test.   As is typical of Beta testing - the beekeepers found all the ways that the scanner could be broken (can't survive being run over by the truck), what happens if one doesn't charge a lithium battery (its not pretty), as well as the challenges of using the scanner with  a variety of different hive configurations (pallets, hive stands, screen bottoms, entrances on the inside of pallets, etc.  
 
We did learn that bees in New Zealand have an accent.  We learned that we need to have more hands-on with respect to calibration of the units -  it looks like for large commercial beekeeping operations, bees may have their own 'signature' sounds, much llke my crew members from the south, east, and western states, we all sound a bit different.  We discovered that queens produce sounds in frequency ranges outside of those typically reported.  So, this year we're going to stage two of Beta testing.  We can't fix the 'run-over by the truck problem', but we did change the battery charging system, we improved the graphical interface, and we're revising how we're getting data to calibrate these units.  
 
That's the key issue - we made changes a year ago in terms of the audible range that we're scanning.  This necessitated re-calibration of the units.  
 
Our USDA funding ended a year ago, and USDA doesn't provide funds for commercialization.   As such, the calibration phase is going slower than I'd like.
 
It costs us about $1000 to hand-build the Beta scanner prototypes.  We're a small company, so we've not the resources to build a bunch of these and get them calibrated (which has to happen in real beekeeping operations).  Our software programs LEARN sounds.  If you are one of our testers, and if you have colonies with varroa mites at different levels, you'd want to record these and send back to us the  recordings along with a log of the mite levels, etc.  Same for queenless versus queen-right. situation, and each of the other parameters.  And, if you want to calibrate for sounds that we haven't explored (one group is looking at different lines of breeder queens), you can conduct an experiment, get the recordings, send them to us, and we can see if the  system can detect a difference (be calibrated to your use).
 
Our scanners can process sounds on the hand-held device to give you a report on each hive (i.e., detection of  different pests and pathogens and probability of detection (e.g., 10%, 50%, 90% likely that the hive has mites, is queenless, etc.).  However, the device itself can't learn.   Our testers have to send back the data - either via an Internet upload, or simply by mailing back the SD card from the scanner.   As we get recordings back with data describing what was recorded, we add the sounds to our libraries.  Bob then re-calibrates the scanner program for best fit.  Re-calibrations are sent back  to the testers.  As we go, the system should get better and better.
 
Our initial work using the audible range of sounds resulted in about 86-98% accuracy, depending on the variable of concern.  It did nearly 98% on AHB.  Its too bad we're not in the UK.  According to press releases a few months ago, the Europeans now reportedly have some serious money to explore use of acoustics in beehives.  Unless they're a lot better than we are, it will take them some time  to pull this off.  We've been working on this for more than 12 years, and we had reasonable levels of funding - although not as much as has been reported is being spent in the UK led effort.  We made a lot of mistakes in the beginning, but  got better as we went.
 
Even a part of the UK/European investment would get our systems fully calibrated and ready to use .  Unfortunately, they didn't invite us to partner with them in the research, which we would have been very interested in doing.
 
So, as per usual, we're plodding along, trying to get this done.  We have the knowledge, we have the base libraries, we have the methods, and we have the hand-held scanners.  Anybody want to help?    Anyone have any suggestions how we move this along,  FASTER?  So far, we've been selling hand-helds at cost to a limited number of beekeepers and bee specialists, and letting them use them in their bee operations.  However, we're not getting back much data - as always, people get busy  with the business of managing bees, going to almonds, fixing the broken truck.  This could be a good project for bee associations.  Imagine if every bee association had a scanner and had someone or members who'd commit to getting specific sound recordings for the bees in their own geographical and climate settings, as well as for colonies from different queen lines.  Even at a state level  - 50 scanning units would provide a lot of data, and our Pattern Matching software would have lots of information for the training and testing, resulting in vastly improved calibration libraries.   May Project Apis  m or a Costco or Hagen Daz,  or a new Corporation - how about Kraft, General Mills, Post or any other distributor of honey flavored cereals?  Imagine a blurb on the box for kids describing fun facts about the bees that produce the honey, how a percentage of sales would go to this innovation and unique manner of improving bee health?
 
As per bee 'hearing' all of this makes me wonder.  I can understand that if a bee is exposed to a pesticide, its likely to change its sound in characteristic ways  simply because many pesticides are neural poisons.  Bees get drunk on alcohol, and they 'slur' the buzz (sorry, couldn't resist).  If I were a bee, and something as big as a varroa mite was biting me, I'd probably make a protest sound - ouch, or something four-letter.  But, the surprise to me was when we detected AFB and the levels of AFB.  Far as I know, the adult  bees are survivors.  It's the larvae that get sick, and I don't think  larvae make sound, or at least, not a very detectable sound.  I'm left with the conclusion that the adult bees are making a different sound in a colony with AFB.  If so, does this serve a purpose? Is it a form of communication?  Do bees 'talk' to each other.  Tom says they signal STOP, what else do they communicate via sound?
 
Jerry
 
 

 
 

             ***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software.  For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html