Charles said:
"I can test MANYhybrids.   Putting bees on every field and seeing which ones they work is notfeasible.Evaluating nectar after the bee consumes it  doesn't tell meanything about the flora source volume."


Ok, you've asked a question with multiple parts.  For DoD we built an upgraded Proboscis Extension Assay System (PERS), similar to the approach taken by Inscentinel in the U.K.  In fact, they worked with us for a summer here in Missoula.  Their system was very expensive, ours is simpler.  

This assay allows one to test olfactory attractiveness on a benchtop. Bees are placed in holders and then conditioned to a reward of sugar syrup  IF they extend their proboscis when an odor is presented, that's a clear sign that they perceive an odor and that it isn't repellant to them.  

For soybeans, the question was  - could they be conditioned to the scent from each new variety.  For most, the answer was no.  Either they couldn't perceive any discriminable scent, or they had a negative reaction to the scent of the variety.

PERS is a OLD method, dating back to 1800s, using bees captured with their heads above index cards with V's cut into them to use as clamps, and manual presentation of odors, syrup, etc.  The problem with this test is that if the bees can see the person conducting the test and rewarding them, they quickly learn to key in on the observer, rather than the odor.  

Inscentinel built an isolation bee container with video imaging.  First a benchtop system, then a hand-held unit that looked like a deluxe Dust Buster vac.   Their business plan was to sell clients a bee hotel, and then they expected their main income to come from them providing the client with bees trained to different scents.  They projected using this approach for everything  from strawberries going bad in grocery stores, to  human diseases,etc.

Our PERS approach is different, we are primarily a contract research lab, providing custom tools (i.e., a PERS assay system designed for the needs of the client) and research services.  We're more interested in the questions that can be answered than production of test units, and we don't see airports and stores using trained bees, delivered weekly, as a sensor system.  (Conditioned bees usually retain their training for about a week max, and the length of their survival living in any form of cage outside of a colony isn't usually more than 10 days.


For our own interests and those of our clients, we  mainly  want to be able to quickly ascertain whether bees respond to specific odors - or to provide others with systems that can.  So we build  small knock-down  hoods that can be set up anywhere - even a hotel room.  Bees are held in plastic holder with heads and antennae sticky out.  The hood has an air evacuation system to pull the presented odor past the bee and to evacuate it from the chamber.  The chamber is build of a hard, easy to clean plastic.  We don't want to odor absorbing or adsorbing onto the hood surfaces, and we want to be able to clean it. We also provide a high end flow meter - the air flow rates are critical to making this assay work. PERS requires puffs of air, not a continuous flow, in order to elicit a response from the bee.  Too much air and the bee hunkers down and the odor is too dispersed.  Too little air flow, and the odor perception thresholds drop.  Bees don't do well when air flow is strong or air isn't moving past their antennal receptors.  So, a high quality air flow meter is a must for reliable, reproducible results.

The main thing that we do is to build a computer-controlled, custom air puff injection system.  A common aquarium pump is sufficient  for an air supply, or an air bottle can be used.  What is important is that two air lines converge just in front of the head of the bee.  One line is for clean air, the other for air containing the test odor.  A red plexiglass shield btw the observer and the bee keep the bee from watching the tester.  The computer randomly puffs clean and test air.  Neither the bee nor the observer knows whether the puff contains an odor, and if it contains an odor, which odor.  That keeps the bees from taking their cues from the observer, and it keeps the observer from being biased by knowing whether the bee should or should not be responding.  The final scores are provided by the computer in an Excel Spreadsheet report.

The computer not only determines the sequence of blank versus test odor, but  it also triggers miniature solenoids that open and close the air lines.

The beauty of this approach, one can test almost anything for attractiveness and/or for the odor perception threshold in the absence of other confounding factors - it tests the olfactory perception of the bee.  It will also flag those odd chemicals that actually seem to innately repel bees.  Some drugs are a real turn off to bees - they scrub their antennae vigorously.  Some pesticides appear to block perception of plant odors.

But, for most chemicals, if bees can detect them, and if a reward is associated with them, they'll respond in a positive way. That's the essay our colleague was using with his soybean varieties.  His grad student was using a GC to see what the chemicals were in the varieties of soybean that proved to be attractive to bees.

From my perspective, that a key first step.  If the plant is not producing volatiles attractive to bees, they'll only find out whether the plant is yielding nectar or pollen by visiting the plant - but that's less likely if the  plant is not  producing an attractant.  

We've also found for onions, carrots, and other crops that bees can be persuaded to focus on pollinating specific crops that are not normally attractive to bees or that may even be repellant.  We did this for onions in the early 2000s for a US Seed company that was later sold to a Dutch Seed Company, a part of Bayer's European Seed Production Division.

Currently, we're continuing this type of investigation for an overseas client.  They have one of our PERS assay  systems and their own biochem lab working on the chemicals; we are training them in methods for improving bee visitations in the field.  So far, we've helped them look at several applications, and we are starting to help them use bees for biosecurity.

One may argue that scent isn't an issue - it's a plant, there are visible flowers, so obviously bees will visit them.  That's may be true if we condition the bees to the visible markers of the plant, but on their own, in the absence of attractive odorants, we don't see many scouts checking out the plants.  Bees can be conditioned to visit shapes, colors, odors, and combinations - but what they do on their own is often mediated primarily by the scent (my bias, based on years of work conditioning bees).   No or weak odor attraction, low visitation.

Finally, have any of you tried Bee Boost and other pheromone based products and wondered why they sometimes work and sometimes don't?  We find that we can change that result - the error is that we expect them to be inherently attracted - but other factors control whether bees are in a receptive mode.  Combine Bee Boos and the appropriate conditioning  and the result will change.

And in this case, I mean that we have had multiple successes in conditioning tens to hundreds of colonies to focus on a specific crop.  I'm also about 95% sure that we can condition  them to avoid certain crops, but the seedless orange growers didn't believe us - they called one of my peers in CA and he told them it wouldn't work, so we  couldn't even get them to try.  SO now they build tents over entire fields.

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