"And in our selective breeding for resistance, we've so far seen little
heritability"

I disagree based on the data Randy has shared with me.  But the difference in our opinions is due to expectations.  Following I list some real world breeding results.

Corn - Corn yields have gone from about 35 bushels/ acre to a bit over 200 in 85 years.  Corn breeders get three generations a year.  Thus the breeders have seen improvement of 0.75%/generation  assuming all the increase is genetic which is not the case.

Soy beans - 35 bushels/acre to 60 bushels/acre in 40 years and again three generations/ year.  This is an improvement of 0.5% per generation and is mainly or entirely genetic.

Racing homers - Distance a young bird will have a decent chance of flying home the day of release has gone up by a factor of 10 in the last 120 years.  Somewhat less than one generation/year.  The improvement has been a 2% increase/generation and is entirely genetic.  But pigeons are easy to breed as both maternity and paternity are easy to contol.

In the above examples improvements have been slow and steady without sudden big improvements.  In all three cases the populations to select from are close to infinite compared to honey bee breeding programs.  In the corn case some is due to so called hybrid vigor but remember we really have no clue what causes hybrid vigor and most people think it is far more universal than it really is.  The chance of seeing it in any honey bee breeding program is just about zero.  In fact it often fails entirely to show up even in corn.

So, how has Randy's mite resistant program gone by comparison?  He has seen a factor of 6 or a bit better improvement in terms of % of stock that proves resistant in about 10 generations.  That is a 10% improvement per generation. Far better than any of the above results.  Way better than homers which were improved following exactly the same breeding model.

In general the expectation of fast improvement is a pipe dream.  The problem is all us old folks learned genetics from the Mendel view point which most of the time in the real world is total nonsense.  Very few valuable traits are one gene.  Even fewer are either really recessive or dominate.  Most valuable traits are 10 or 25 or 50 genes.  Or even more.  And a bunch of them are supplemental mutants.  Until about the 1980s the one gene idea dominated most western thinking about genetics in spite of the fact that this idea had been proven wrong clear back in 1875.  The Mendel idea was so simple it was hard to abandon.  Even high school kids could easy understand that model and were never taught it was a massive over simplification.  Then we went thru the phase of thinking all genes coded for proteins and we were starting to get counts on the number of protein coding genes and there were no place close to enough of them.  Randy is likely not selecting for even one protein coding gene.  In general those are too important to mess with very much.  It really has only been in the last 15 years that ideas like this have come to dominate the thinking of the professionals in the field.  My guess is Randy is currently selecting for probably 10 or 15 non protein coding genes in his mite program.  Almost for sure some are supplementals.  I have not tried to calculate the number as his data is really too thin at this point to do a calculation with any accuracy.  But, the extremely rapid progress he has made says it is a small number.  Not something like 50. His rate of progress dwarfs programs like the Purdue mite biters as they have far more generations and are doing II to boot.  In fact their rate of progress is so slow in my opinion the program has out lived its reasonable life time unless they change it and go to marker assisted selection.  Doing what they are doing is not even giving the slave labor the quality of research education they are paying for at this point. Even if they could develop a pure strain of mite biters one out cross is going to go back to little or no performance so we would all be 100% dependent on purchased queens.  That means replacing every swarm queen or supersedure queen with a purchased queen immediately unless you have so many hives you can effectively drone flood your whole mating area for a circle 15km in diameter.

Dick

             ***********************************************
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