Mark  I don't have a magic number to give anyone in terms of comb replacement.  The comb stores all - food, brood,  chemicals and many pathogens.  Entombed cells are nothing new - but I don't think most of us old timers asked why?  And I used to see more of it before we had widespread use of  antibiotics, so maybe bees have yet another form of hygienic behavior.  Foragers are known to even sleep in cohorts in adjacent cells.  Lots going on in the comb.

Chemicals and pathogens in cells and in wax can be accumulative.  Yes, move a colony into a contaminated area and the wax RAPIDLY looks like the wax in resident colonies.   But, if you move to a new location with different contaminate concentrations or different chemicals, you  are likely to add new to old contaminants; and the same goes for pathogens.  There's an ebb and flow, but OLD comb in brood nest tends to end up with layers of chemicals and spores, one on top of the other, until TEAR DOWN by the bees.  If your bees are getting small in size, that's a warning that the comb is old, and the cells reduced in size, and you can expect someday to see the bees remodel - and that's when you don't want  all of that junk coming back out.

Yes, if you put in clean comb, it will pick things up - you can't avoid it.  But, I am a strong proponent of comb rotation.  In Chile, I saw some of the prettiest comb I've seen in years, and I saw good  looking colonies.  Some of their beekeepers claimed very high replacement rates for comb - but I'll let our friends from Chile address the issue of how many combs on what intervals.

I learned years ago that 'gifts' of comb were no gift at all.  As a researcher, I always started from foundation.  Never had much of any problem until the first gift - of comb, of Foul Brood, of wax moth.  Even if it looked good, it still bit me in the butt.  Clearly, when and where problems started could be traced to the gifts.  So, I stopped accepting  gifts, got rid of it all, and went back to better, healthier bees.  Then, CCD occurred, and I started bringing in hives to look at this, and since then I've been keeping hives for Nosema, virus, and pesticide work.   We have to isolate those yards to keep the rest healthy.  Now I have to either destroy everything after a trial OR I analyze to see if its 'clean'.

One of the presentations at ABRC in San Antonio surprised me at the high % of hives where  ABF was found in one of the surveys.  That makes me wonder about the beekeepers who swear by a  certain Pollen Supplement with a mix of essential items.  Bees look great, best in years.  

Except, that Pollen Sub has an essential oil that isn't an essential oil, its Tylosil (Tylosin) - which is illegal, as well as a fast path forward for developing resistant AFB.  Our work indicates that the majority of essential oils in that particular  mix shortened bee life and they didn't help at all with lowering viruses levels.  Two oils may have value - we're still working on that issue.  But, I suspect that the Bees Doing Great result may be a function of controlling undetected Foul Brood.  And, because I'm tired of being threatened with law suits for speaking my mind, I'm not naming the product, just ask your supplier if it contains Tylosin.

So, what to do?  I know Randy doesn't see a reason to rotate, but he also says  his bees are in areas with low pesticide use, and he closely monitors his bees for lots of things that can affect their health.  So, I suspect his combs are relatively free of chemical build up and pathogens.

But, if I were a migratory beekeeper or if I kept bees in industrial areas or placed bees in areas of high pesticide usage or had lots of neighboring beekeepers who were into sharing pests and pathogens, I'd rotate comb.

But first, I'd send a few of the worst looking combs off for chemical and for pathogen analysis to establish a benchmark.  Send the whole comb - you're not going to want to use chopped up combs anyway.    If the results come back negative, then there's no reason to remove and replace combs.

With respect to replacement, you will ask Which combs, how many?  I'd focus on brood combs, since they get everything.  I'd prioritize old, small cell combs.  I'd  get rid of mouse chewed combs.  Most importantly, I'd set up at least one trial beeyard, preferably at least three yards.  In each yard, put identifiers on the hives.  Now, randomly choose 1/2 the colonies and don't replace any combs in those colonies.  Replace a fixed percentage of the combs in the other 1/2 of the colonies.  Best, run this for 3-5 years.  Keep records on bee population size, overall health, and honey production.  A bit more work than NOT experimenting, but you'll have data, not guesses on which to make a decision as to whether this is worthwhile or not.

Please note, as a scientist who intends to publish experiments, palletized operations pose an issue.  In theory, randomized and assignment of colonies to treatments should mean that you start with all of the colonies at a location and randomly assign each hive to one or the other treatment (in this example, assume control and replaced comb as the two treatments).  With pallets, there are concerns about drift, and its easy to accidentally mix and match equipment on the pallet.  So, the alternative is to assign treatments by pallets.  Each pallet of 4 or 6 hives get the same treatment.  You  randomize which pallets get which treatment, not the colonies on a pallet.

Why randomize?  Simply stated, there may be position effects - drift, sun, shade, moisture, within any  yard.   So you want to equalize the environmental factors by making sure equal numbers of colonies for each treatment experience each of these  factors.  In other words, don't split the yard front to back or side to side, where all of the hive in one treatment are in the sun and all of the ones in the other treatment are back in the shade.

I'm assuming a yard of 20-30 colonies.  Variability of performance in colonies is such that as a rule of thumb, if I can, I make sure I've at least 12 colonies in a treatment, that will help a lot when you get to comparing the data.  Never have less than 3 colonies in a treatment - one  colony alone has no replication, two colonies may leave you with a high and a low value, and three colonies at least gives you a tie breaker.

You  can have more than two treatments - control versus say 30% percent replaced combs, but I'm trying to give you a simple approach.  What will be difficult, if not impossible to assess the data, is if you use too many treatments and not enough repeats (replication) of each treatment.
We make these decision all of the time when designing trials.  There's the ideal, and then there's the practical given available time and money.

Replications can occur within a yard.  For 30 colonies, you could have 15 control, and 15 colonies with a given number of replacement frames.  For this illustration, let's say 30% comb replacement.  So 15 colonies at 0 % replacement, 15 at 30%.  Or, you might have 10 at 0%, 10 at 15%, and 10 at 30%.  Or 3  at 0%, 3 at 5%, 3 at 10%, 3 at 15%, 3 at 30%, 3 at 45%, 3 at 60%, 3 at 75%, 3 at 90%, 3 at 100% - but that would be a LOT of work.  From a statistical perspective, keep the numbers of colonies per treatment balanced (equal); it makes the statistical analysis easier.

The other approach, using this example, is to use 30 locations, with one hive at each location representing one of the treatments.  That makes for a strong data set, but it generally greatly increases the logistics, costs, times.  Finally do NOT put all of one treatment in one yard, all of the next treatment in another yard.  That gets into a problem called pseudoreplication.  Any effect may be a function of the location, not the treatment.  So, if you spread treatments across apiaries, again use a balanced set of treatments in each yard.   Back to our 0% and 30% replacement example.  You could have pairs of 1 0% and 1 30% in each of 30 yards.  Or you could have 3 0% and 3 30% in each of 10 yards.  Or any other combination where the total number of hives per treatment is the same when added up over all of the test apiaries, and the each of the treatments occurs with the same number of colonies in each apiary.

Minimally, split one apiary, use 0%  in half the colonies, and one replacement% in the other half.  If the yard is large enough, say 30, I'd suggest 10 colonies at 0%, 10 at a low % replacement, and 10 at a higher % and consider doing the same in a second yard.

Its your choice as to what levels of replacement you want to compare - it all depends on how many years it will take for complete rotation.  33% takes 3 years, 50% two, 100% in one.

Finally, if you understand statistics, and have either a spreadsheet like Excel or a stats package, have fun.  If the differences are large, you  may not need statistics.  But, if as usually happens, it looks like you may have a response and if you aren't able or comfortable doing your own stats, hire a statistician.  Just like Building Inspectors, Electricians, CPAs, statisticians are available for hire.  In fact your, best approach is to hire a statistician  BEFORE you start the trial.  That way you  won't have to try to FIX what you should have done.  

Its just like taxes.  Keep your records in a shoebox and drop them on an accountants desk 5 years later, when the IRS plans to audit, and you've a mess.  Keep good records and hire a good accountant and things should go much better, with far less chance of getting nasty letters from the IRS.  

Finally, there are statistical specialists.  The statistician in an engineering firm or in a mathematics department may be just fine for your needs, maybe even a high school or college student. 

Just keep in mind, for biological statistics, especially for field trials, the application of statistics to experimental design and data analysis is a specialty unto itself.  Field trials and biology are messy systems.  Ideally, you want a quantitative ecologist or biometrics specialist.   Ask aboutbackground, publications.  

There's an age old split in academia regarding stats courses.  The biologists often find that math departments have a different viewpoint  - and I'm not putting down mathematicians.  But, often, a math department focuses on the theory and assumptions underlying the stats - the math, but fails to provide good guidance with respect to application. Now, if we were all mathematicians, then I'd agree that if you understand the math, you could readily apply it to all sorts of issues.  But, many of us are biologists, not mathematicians.  Some of us will never really understand the math that well.  So, one often finds biology departments (life sciences) offering biometrics or stats courses to biology students, with a focus on which stats to use for which problems, and HOW  to apply.  We're content to leave it to our colleagues in math to determine whether the underlying concepts are valid.

Jerry

J.J. Bromenshenk
Bee Alert
Missoula, Mt


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