Beekeepers in this area have been telling us for some time that this has been a very bad year for Varroa mites.  By late summer we knew that one of our hives had an escalating Varroa mite problem, which we outlined in our last post.  The other three hives in the apiary still had low mite counts, but knowing the burden was high in one, we elected to treat all four hives in mid-August.

Varroa mites on drone pupae in July

As it was still technically summer when we started treatment, we elected to use a Thymol based product.

After two weeks of treatment, just for our own interest, we went ahead and did a mid-treatment mite count.  As expected, the mite population had not appreciably decreased (for an explanation as to why, see our previous post).  In fact, in three of four hives the mite populations had actually increased slightly.

At that time we weren’t sure if the increasing mite counts in those hives was due to increased mite drop, directly as a result of treatment (accelerating the drop in mites), or if the mites were managing to gain population despite treatment.

General Inspection Findings

This weekend we went back into the hives, primarily to check on food reserves, and remove the remaining drone frames, and then perform our end of treatment mite count, hoping to see a decline in mite population.

For all the hives, except Rosemary, the Thymomite pads were more or less where we left them.  We did note two things of significance though.  The first is that the Rosemary hive, for reasons known only to the bees, insisted that the Thymomite must be removed from the hive.

The bees in the Rosemary hive discarded this Thymomite strip on the ground outside the hive

At the start of this inspection we found one partially broken down pad on the ground in the apiary. We suspected this was from the Rosemary hive as we’d seen a pad at the hive entrance a few days before.  Indeed, the inspection showed that only the Rosemary hive was missing most of their Thymol strips.  I remember spitting out cough syrup as a child, so I can hardly blame them.

We thoroughly inspected the Salvia hive first, and as we were wrapping up that inspection, we noticed that Rosemary was wrestling yet another Thymol strip out of the hive.

It was remarkable to watch the bees attempt to pull this massive strip up and over the robbing the screen (left), but they gave up and dropped it (right), so we removed it

We watched as bees at the entrance, with apparent herculean strength, struggled to carry the Thymol pad OVER the robbing screen!  These bees clearly wanted this OUT of their hive!  As such, we wondered just how effective this treatment would prove to be for that colony.

The other point of note is that our feral colonies, Salvia, and Lavender, both heavily propolised around the Thymol pads on the top bars.

The feral colonies produced excess propolis around the edges of the Thymomite pads

Rather than drag the Thymol pads out of the hive like Rosemary, they just tried to wall it off. Quite literally.

A great wall of propolis around the Thymol pad in the Lavender hive

However, neither of our Italian bee colonies, Rosemary and Chamomile, put any propolis around the pads.  You can see propolis around the edge of the frames here in the Rosemary hive, but note there’s NONE around the Thymol pad at all.

The Thymol pads in the hives with the Italian Queens had no propolis around them

Apparently our Feral bees like to produce a lot more propolis than the Italians.

Also of note, despite not taking any honey this year, all the hives were low on honey and pollen stores.

The Salvia hive has consumed more than 10 medium frames of honey in the last four weeks

Most notably, all the hives were showing significantly more spotty brood patterns. Whether that is a result of treatment, or mite levels, isn’t clear, but we suspect it’s likely both.

The brood nests in all the hives have a spotty laying pattern, with various ages of larva at any given location

Treatment Results

The Thymol treatments appeared to be somewhat disruptive to the colonies, but what about the treatment effectiveness over the last month?  Has it knocked back the Varroa counts as much as we hoped?  This is where things get interesting.

The short answer is that at first glance the treatment seems to have failed.  However, I suggest you read on, as the answer isn’t really that simple.

The chart below graphically represents the pre-treatment, mid-treatment, and post-treatment mite drop counts for each hive.

Graph of mite levels in all hives pre, mid, and post-treatment (click to enlarge)

The actual numbers for the counts are in the table below.

24 hour mite drop counts pre, mid, and post-treatment

As you can see, the post treatment numbers, at least initially, look like the treatment was ineffective, the mite counts are well above recommended treatment thresholds [1].  However, after spending some time looking at the numbers, and really thinking about what’s going on, both with the bees, and the mites at this time of year, it may not all be as horrendous as it looks at first.  I won’t argue that it’s not bad, but it could be worse.

When I pulled the latest mite boards, my initial reaction was surprise, the post treatment numbers weren’t at all what I expected to see.

Surprise then quickly, and briefly, turned to panic, as fellow beekeeper Lisa will attest to, when I immediately fired off an email to her about her treatment experiences last fall.

The whole point of treating our bees in mid-August was to get the mite count down so the bees have the best chance of building a healthy last round of summer bees in late September, who will then raise the critical winter bee populations in October.

Healthy populations of late summer bees are critical to ensure the health of a hive that's preparing for winter

If these September nurse bees aren’t healthy, and there aren’t enough of them to rear a substantial population of winter bees, these hives will all be at significant risk of failing before the end of the year.

Graphing the data though allowed me to consider the numbers a little differently, and after staring at it for a while, I began to think about the numbers in context.  Why had Salvia’s mite count barely changed from the start of treatment?  Was Thymol that useless?  Most beekeepers we’ve spoken to swear by it, especially as a warm-weather treatment.  Why had Chamomile, that hardly had any mites at the start of treatment, and based on standard treatment thresholds didn’t even need to be treated, now find itself with mite counts that rival the Salvia colony!?

We know we can't reduce mite counts to zero, but we at least expected the numbers to decline

Let’s start by looking at the Salvia hive.  The mite burden in this hive at the start of treatment was the impetus for us deciding to treat the bees in our apiary in the first place.  Look at the pre-treatment bar in the graph above, compared to the other colonies.  Salvia was the first hive we set up this year, had been raising brood longer, had been producing a robust population of drones all season until just a couple of weeks ago, and all that translated into supporting a burgeoning mite population.  There was no question we needed to consider treating that colony.

I really questioned the rationale of treating all the hives at this point though, just based on mite counts, as the other hives were all below the approximately 30 mites/24 hours treatment threshold that had been previously recommended to us.

Now look at the red bars, the mid-treatment levels after two weeks of Thymol in the hives. There’s an ever so slight dip in Salvia, which could quite simply be sampling error, or a slight shift in mite count due to population change in the hive (decreased drone production mid-treatment), or ambient temperatures affecting grooming activity of the bees.  There are many variables, so who knows?  However, also note that the mite drop counts for the other three hives appeared to be slightly increased at that same point in time.

When adding more strips during the second half of treatment, the existing strips are left in place

At the 2 week treatment mark, we applied a second round of Thymomite strips to the hives. We’d treated toward the low end of the dose range for the first round.  After the mid-treatment count though, I elected to increase the dose in all of the hives.  Feeling like we’re fighting against the calendar, and wanting desperately to get the mite levels down, we hoped this would just be a little extra insurance.  As such the two smaller colonies received 1 whole strip, and the two large colonies received two strips during the second half of treatment.

So yesterday, the first post-treatment sticky board I pulled was from under the Chamomile hive, and I was horrified to see how many mites were stuck to the board.

The Chamomile hive had few mites at the start of treatment, I expected the board to be more or less clean after treatment. I was wrong, very wrong.

This hive has always had very few mites, but has also always been our smallest colony.  Although natural drop counts aren’t quantitative, multiple counts over time do demonstrate whether or not mite levels are increasing or decreasing within a particular hive.

The Chamomile bees were packaged Italian bees that arrived in May, and unlike their neighbor’s, Rosemary, who arrived the same day, Chamomile was very slow to build up population all through summer.

What the graph doesn’t show is that Chamomile has about 30% of the population of bees as Salvia, so although their post-treatment mite counts are now equivalent, that suggests Chamomile’s population of bees could now be considered to be roughly THREE TIMES MORE INFESTED than Salvia.  Their mite population seems to have exploded, with a 3300% increase in mites in the last four weeks, in the presence of Thymol!

Looking at the difference in mite counts before treatment, and afterwards, it's clear that Chamomile's count has skyrocketed...but Salvia's percent increase is MUCH less, but why?

Using an online Varroa Calculator, this returns an estimate of 5,400 mites in the Chamomile colony, well above a presumed tolerable pre-treatment limit of 2,000 mites per colony.

Note however, Salvia’s mite count on the other hand has held relatively steady, only increasing 20% in the same time period.  We’d prefer there was no increase, but this time of year is when Varroa mites really ramp up their reproduction.  Bees are also decreasing the sizes of their colonies in preparation for winter, and weaker colonies can quickly become overwhelmed by burgeoning mite populations.

Lavender and Rosemary both have similar numbers of total bees in their hives, and the mite counts have increased at a similar rate, just over 1200% since the start of treatment.

Trying to Make Sense of the Mite Counts

With the exception of Rosemary throwing the Thymomite out of the hive, all four hives received the same treatment, dose adjusted for population, at the same time, for the same period, under the same climatic conditions. So why the variability in increasing mite levels?

The primary differences between these four hives are total colony size/strength, with Salvia > Rosemary > Lavender > Chamomile, and the source of the Queens, and worker bees.

The Chamomile hive is comprised of very blonde Italian bees, with a pure Italian Queen, from one of our packages that arrived in May.

Two of colonies have Italian Queens, bred at commercial apiaries. The other two Queens are feral, locally mated Queens.

The Rosemary hive was also a package colony, from the same commercial apiary, comprised of Italian bees, but seemed to gain significant population from our large feral colony situated immediately next to them on the hive stand.  We noted a lot of drift into this hive from Salvia early on, and the Rosemary colony has been stronger all season as a result.  Looking in Rosemary’s hive, it’s easy to see a clear mix of very blonde Italian bees, and lots of dark stripey bees from Salvia.

The feral Salvia and Lavender hives were both sourced from the same colony, just across the creek from us.  The main difference between these two is that Salvia has a Queen that survived last winter, which swarmed in late March.   The Lavender hive swarmed with a virgin Queen in April, from the same colony.  The Lavender colony has been building up population for approximately 1 month less than Salvia.

Our feral Queens (red arrow) are much darker than the Italians

When we first were attending guild meetings and lectures, every beekeeper we met touted the value of ‘feral’, versus ‘package’ honey bee colonies.  Is it possible that part of what we’re seeing here is that the Salvia hive, despite a high mite burden for most of summer, and seemingly increasing mite populations in our hives, is actually, with a little help from the Thymol, simply better able as a colony to cope with their mite infestation?

Hygienic bees are known to remove diseased and infested pupae from the brood nest

It’s difficult to know for certain, because our data set simply isn’t big enough, however, it may help to explain some of the difference we now see in trending mite populations between Salvia, and Chamomile.  Our feral colonies, until arriving in our apiary, had never been treated, so it’s reasonable to expect they have adapted at least somewhat to the presence of Varroa in their environment, or they wouldn’t be here.

Even if all the hives are removing diseased pupae though, a weaker hive may simply not be able to keep up during the period of escalating mite populations in the fall.  As such, that would imply that this is very bad news for Chamomile.  This has been a relatively weak colony since its arrival, and looking at this colony as a whole, in conjunction with the current overwhelming mite counts, my prediction is this colony, left as is, is highly likely to fail in the coming weeks.

Although mite counts have increased for all of the colonies despite treatment, without treatment we’d expect the counts would have been much higher (this is where having a control colony in our apiary would have been informative).

If only the bees could talk

Salvia, the colony that survived last winter, ironically now seems to be the best off of all our hives, and yet at the start of treatment it was the colony we were most concerned about.  Lavender, although mite counts have increased, isn’t as bad off as our ‘pure’ Italian colony, and Rosemary, who has received an influx of feral workers, despite having an Italian Queen, seems to be holding its own as well as Lavender.  Are the feral bees conferring an advantage to the colonies that have them?  Are we simply seeing a colony’s ability to cope with mites based on colony strength, or origin?  Or could the increased mite counts in the smaller colonies be a consequence of robbing, and an influx of mites into those hives?  It’s likely a combination of all those things.  Interpreting the effects, and effectiveness of the Thymol treatment, would have been so much more straightforward if the mite counts had declined.

The Next Step?

The more important question now though, is where to go from here?  The mite counts are still too high.  We’re now trying to decide if perhaps we should try retreating again with the Thymol, as although it hasn’t decreased the mite counts, yet, it may at least be preventing them from escalating further (I shudder to think how high Chamomile’s count may have been if untreated).  Alternatively we could follow up with a different organic acid treatment, perhaps either Formic Acid, or Oxalic Acid a little later in the season.

For Chamomile, our weakest hive, as this hive has struggled relative to the other hives all season, and it’s mite population is exploding, we are seriously considering replacing their Queen with a mated Queen of a difference race of bee.  Perhaps either Russian, or Carniolan.  It’s too late to let this colony raise their own new Queen as drone season here is done, and she would risk a failed mating.

The bees are no longer building drone-sized comb in the empty drone frames

Requeening this colony now, while there’s still time to raise winter bees, may be the best hope for the Chamomile colony to survive, without having to bombard it excessively with treatments.

I wish we could fast forward to the end of February 2012 to see how this all turns out.  We’d hope that at least the Salvia colony would survive winter again this year, and any hive that proves its mettle, and survives this winter will be split at the beginning of the season next year.  Much of our Varroa management in the future we’d like to base on splits, and breaks in the brood cycle.  The consequence of that, we hope, is that we’ll also be selecting for stronger bees.  The question is, who will survive?

Very few full frames of capped honey remain in any of the hives. Fall feeding will help the bees as they raise their winter bee populations

While we decide which path to take next to best help our colonies, the more pressing issue now is colony nutrition.  The other horror in the hives this weekend was the discovery that our bees have consumed more than half of their honey stores in just the last month, and we’re squarely in the middle of a nectar dearth. How we’re handling fall feeding of our colonies though, and why, we’ll save for our next bee post.


[1] Oliver, Randy.  2006.  IPM 4, Fighting Varroa 4: Reconnaissance Mite Sampling Methods and Thresholds