Florida Extension Beekeeping Newsletter
Apis--Apicultural Information and Issues (ISSN 0889-3764)
Volume 13, Number 8, August 1995
Copyright (c) 1995 M.T. Sanford "All Rights Reserved"
SAMPLING FOR VARROA
In the July 1995 issue of Bee Culture, Dr. Mark Winston writes
that the possibility of creating Varroa mites that are resistant to
the one product now labeled for their control is very real (also
see November 1994 and February 1995 issues of APIS). One way to
hold off the inevitable as long as possible is not to treat with
Apistan(R) until absolutely necessary. Unfortunately, the number
of mites it takes to do irreparable damage to a colony is not
known. And sampling techniques that determine the actual number of
mites in a beehive are extremely variable in their effectiveness.
In Florida, the "ether roll" sampling technique continues to
be used almost exclusively by bee inspectors in the Department of
Agriculture and others. This technique is well spelled out in the
video production, "Varroa Mite Detection," VT 249, available from
this office by mailing a blank VHS tape to my attention. It is
used so much because it requires only a single visit to a colony
and the results are immediately available. However, it only
samples a portion of a colony and destroys valuable bees in the
Many Florida bee inspectors now urge beekeepers to think about
treating soon (within 10 days) when one or two mites are found in
an ether roll. More than that number is considered a signal to put
Apistan(R) strips in immediately. This recommendation, however,
will vary depending on location. Beekeepers should always check
with their local inspectors for the latest information on treatment
Several other methods for sampling Varroa mites besides the
ether roll continue to be used. These include counting mites that
are either killed/dislodged from bees using the product labeled for
control itself [Apistan(R)] or alternative fumigants like formic
acid and tobacco smoke. In addition, one can count Varroa present
in uncapped drone brood or those on the bottomboard that drop from
natural causes or may be groomed/knocked off by the bees.
In one of few studies to determine Varroa levels in honey
bees, M.D. Ellis and F.P. Baxdale reported on the efficiency of
five sampling techniques from 44 overwintered two-story colonies in
BeeScience, "Comparison of Formic Acid Sampling With Other Methods
Used to Detect Varroa Mites (Varroa Jacobsoni Oud.) and Mite
Distribution Within Colonies in Nebraska," (Vol. 3) 1994: pp. 139-
144, 1994. As expected, Apistan(R) recovered many more mites than
any of the other methods. Formic acid was less effective than
Apistan(R), but more so than ether roll, tobacco smoke fumigation,
debris sampling (natural knockdown) or decapping drone brood, none
of them being statistically significantly different from each
other. The authors saw evidence that formic acid might rival the
results of Apistan(R), but this would take more than the 24-hour
exposure allotted in that particular study.
To determine the efficiency of the ether roll, the authors
compared the number of mites found in two rolls (10-second shakes
for each) to those recovered by washing the sample in alcohol for
20 minutes. The alcohol wash detected about 1.7 mites for every
one found by ether roll. Earlier studies cited by the authors,
however, resulted in a two- to almost three-mite difference. The
authors concluded that because ether rolls are inherently variable,
beekeepers should develop a baseline standardized figure for their
own particular technique by comparing ether roll with alcohol wash.
The authors also described the mite distribution found in
colonies during the study period (July and August 1992). Drone
brood had over seven times as many mites as adult worker bees. And
there were four times the number of mites in drone brood compared
to worker brood. The authors stated that while only 10 percent of
worker bees in their study had mites, other investigators have
estimated as high as 20 percent. The time of year is also an
important issue. This is a period characterized by a reduction in
drone brood and adult drones will soon be ejected from colonies. It
is the time when the switch from drone brood to worker brood is
probably occurring (see October 1994 APIS). Thus, percentages
found in the study may radically change as more mites begin to
parasitize worker brood and adults.
Two other factors also may influence density of mite
populations in bee colonies. One is the effect mite population
density has on other mites' reproduction (see June 1995 APIS).
Another, discussed elsewhere in this newsletter, is the infertility
of female mites, which may depend on the host (race of bee) they
are found on.
A final factor, according to the authors, is the length of
time adult Varroa stay outside brood cells; there is evidence that
this is shorter in colder areas. Once sealed inside cells, mites
are less susceptible to most of the sampling methods described in
the study. The authors, therefore, recommended that regional and
seasonal studies are needed to effectively use mite population
The authors concluded that although beekeepers will not want
to treat colonies every time they find a mite, they need a sound
basis to decide when or if such treatment can be delayed.
Unfortunately, the above study, as most do, only raises more
questions about the effectiveness of sampling for mites. We must
wait for many more before the complex interactions between host and
parasite can be better understood. Only then will it be possible
to estimate a mite population and determine a definitive treatment
level. Even this, however, may not be enough and studies of this
nature may go for naught.
Varroa has been implicated in transmitting viral infections.
This may result in what is now being called parasitic mite syndrome
(see December 1994 APIS). In situations where this possibility
exists, the number of mites feeding on bees becomes less relevant.
Unfortunately, less is known about the viral connection than the
damage done by physical parasitism. In the absence of this kind of
information, beekeepers are left to their own devices. This
inevitably results in one philosophy: treating sooner rather than
Evidence corroborating that the Varroa situation can be quite
different based on climate and geography continues to accumulate.
These differences are evident within a single country, or as the
mite population shifts over time, they might even be observed in a
single colony (see June 1995 APIS).
For those who have had enough of Varroa, perhaps Brazil is the
best place to go. G. Moretto and colleagues published "Reduction
of Varroa Infestations in the State of Santa Catarina, in Southern
Brazil," American Bee Journal, Vol. 135, No. 7, pp. 498-500 in July
1995. The two-year study sampled bees in 19 locations using the
alcohol shake method. The authors found that over the course of
about 10 years, Varroa infestation rates have dropped in the state
of Santa Catarina. An average of about 17 mites per 100 adult bees
was found right after Varroa's introduction. It is now at two per
100 adults, the predominant level in the rest of the country.
The authors stated that this relatively low mite level is not
damaging to colonies in their country. Thus, honey bees in Brazil
do not require any Varroa treatment, despite the mite being present
since 1972. A major reason for this, the authors believe, is the
ability of Africanized honey bees to remove the mites from their
bodies. Climate may also be responsible. But this fails to
explain why honey bees in Santa Catarina state can survive without
treatment, whereas those in Florida with a similar temperature and
rainfall regime cannot. The authors concluded that the decreasing
infestation rates in Santa Catarina state demonstrate an adaptive
process has taken place in the host-parasite relationship.
Another study by P. Rosenkranz and W. Engels, "Infertility of
Varroa jacobsoni Females in Apis mellifera Worker Brood as a
Tolerance Factor Against Varroatosis," Apidologie, Vol. 25, pp 402-
411, 1994, compared European bees in Germany (Carniolan) with
Africanized bees in Brazil. In Brazil, over 40 percent of the
female mites found in worker brood were infertile. In Carniolan
bees, 10 to 20 percent were infertile.
These infertility differences are not dependent on either
broodnest or ambient (air) temperature, according to the authors.
And in both European and Africanized bees, if fertile, female mites
produced the same number of eggs. They also said this low
fertility rate in Brazil compares with that found in the original
host of Varroa, the Asian honey bee (Apis cerana). They
concluded: "Since host-dependent fertility evidently has a strong
influence on the population dynamics of this parasite (Varroa), it
should be properly considered in future selection and breeding
efforts in order to produce European bee strains tolerant against
And then there is the New Guinea enigma. A study by D.
Anderson, "Non-reproduction of Varroa jacobsoni in Apis mellifera
Colonies in Papua New Guinea and Indonesia," Apidologie, Vol. 25,
pp: 412-421, 1994, shows some surprising results. While comparing
Varroa populations on both Apis cerana and Apis mellifera, the
author found that mites were not reproducing on European bees.
They were only present by virtue of being spread from nearby Apis
Extremely low infertility, as observed in the above Brazilian
study, is ruled out by the author as a reason for no reproduction.
He concluded there should have been some successfully reproducing
mites seen in the large number of Varroa-infested cells examined
(about 1,500 out of 68,000 total cells examined). In addition,
there was no evidence that competition between Varroa and another
mite, Tropilaelaps clareae was responsible. The latter mite was
only observed in 35 percent of colonies examined. Environmental
factors were also ruled out because reproduction was not impaired
in Apis cerana colonies, either in the tropical lowlands or more
The author suggested the answer may come from two areas of
study. The first is mite taxonomy or classification. It could be
that the mites are different than those seen elsewhere in Asia or
Europe and the Americas. Studies using DNA could determine this.
The other area is mite reproductive biology. The fact that
female mites infested both worker and drone brood, and were
proceeding toward normal reproduction, but then were unable to lay
eggs, the author suggested, may be because some crucial chemical
factor was lacking. Given the results of this study, the most
important experiment to run, he said, is to introduce colonies
already found susceptible to Varroa in Europe into Papua New Guinea
and see what happens to them. On a broader note, he concluded:
"...the results of this study have shown the uncertainty that may
be associated with predicting likely impacts of pest species in
particular localities based on what is known about these pests in
And what about the viral connection associated with parasitic
mite syndrome as noted elsewhere? This possibility is not
mentioned in any of these studies, but until more is known about
the honey bee-Varroa connection, it should not be ignored.
PANHANDLE BEE SEMINAR
The 1995 Florida Panhandle Bee Seminar will be held September
8 and 9, 1995 at the Southport Mens Club, in Southport, FL. The
meeting runs all day Friday and Saturday. Dr. Clarence Collison,
Professor and Head of Entomology and Plant Pathology, Mississippi
State University will be the featured speaker. The cost is $35.00
per person for both days if registering early. After August 25,
the fee will be $40.00 per person. For more information, contact
Sylvia Bullard, Escarosa Beekeeping Association, 904/478-7690 or
Vernon Gwaltney, Tupelo Beekeepers Assocation, 904/476-4278.
Malcolm T. Sanford
Bldg 970, Box 110620
University of Florida
Gainesville, FL 32611-0620
Phone (904) 392-1801, Ext. 143
BITNET Address: MTS@IFASGNV; INTERNET Address: MTS@GNV.IFAS.UFL.EDU
APIS on the World Wide Web--
Copyright (c) M.T. Sanford 1995 "All Rights Reserved"