V . .I.u(1.?l.::£c;a’
. . v s: Ix.loha..ul.'bb
a; 1:. 1.2... . .
I: .5! Pg 50‘ 0|
. ivovoA...:..0AI:..:.
11.)? z.
.A .0: a 4v:
. z! t.

. :52! .

u n... In.)

3. 1H1 at},
r:

_ . V . V . , V. . ., ..zb§.m...un§..:uflmu _
u . .. . ,. A V. ... . . _ V: .83.}. V l

.4. V 353... . , . . V . . , V . . V. . . V
.r .um.wwohnixfl 4—5. . .. . . . . .
VT rimmwfimch . .

. . . . 1.1%-.)
. . _ V . . . . .. V . V. .n. w . V V V anfinumuo
.uauiflxr. . V _ . . V V. . . . _. A . V V A. . . .V V . . . .

51TH”. 13

, I
p i All...
. . if... - ; £131”. .V ‘ I... .1
15:... 13.2:1T‘f51. ti: 8...! .1 2. ...
.. 5.32:13‘. I... 4.. . .
I: .11.!!! l
V V it.!&2¢viu|.fu
. afflictinKS
.5,

i. 4
it 5.
fly... V .JuL.._!..w....leV
. .394 PM. .3“..ch uIIJWUWW.
if. .I. . .... .V I n .
.3. .31.: Vfiushufimzu...’
.( .3... as. :1. fit.
.. 33.98.91 ... i..v...lf
1314...?le 11...; ,
.1de . . a.

. \...l...l;.1u I.
III. 5 list]: I .Ivl.v.
, V: .Sll. ‘7': Ill- 5
. . I, K fift‘v «v1.2.1! V
. AJOVtAv.\l!! I
.v slit-{$0.128
vu§ it'll Il...‘ . ,

It?! I.“
1);-g‘. 0“...
III.» :‘40'

v t: V

- u , 1i; .

V. ! {Hutu-II?”
. limit; '12-! . n .

r; .2". (V‘s...) V.

.39: 9.5 ...

Ha
I1:!.J.:...¢u.v .
K... . .. 9.. X . AA I...
1.1.3:!“ JV... :11!
a. ,. . to .V. n...ot1l| 1.. V .
{:3 .Ԥ..:iix;.
. .1. l.V 31.3.27}!
a 9-..: ..lu.ast.t..v.1.
. ., .57. .11. .v. . 1...): .0;
{Ju‘lthzrltvion V

.- 1.8!.AA .

Livintp.‘ 100’ ,

. runway.‘

153.; tn‘
g‘filr..lli‘l ll’!
8).... iltl‘90..lguhoul'Ll v 0|

 

ft htV‘ v.
1.9:.Jhl: !.I.(Ihflly!-: ..
.I V .a.‘ . .123! luv.» I I it: :10
L5¢ltnr litlt'ui) i..AIIA.'<OPu. A}... .IvI-flv4 .
J .1! (lit-lifl’n. 371.! ibis;

. V . I‘ll-(£2. .\. .
V . . . 4 I .V\ s 1‘ it'l‘13.v.c «5"
V . V . . V . . ‘ . VAIVlléi-VIPL .
. vcul‘é . V . V . I V V. .. . y . £3011“! : 113). u
.11.! . V . V. V . . 1y . lot! v. .
. . . V . . . v. V . . V .I A
V . . . . c:.! .7

.IAV. uVI'nBOJ-D‘

. :LALVrYLOLI A.y..l. V

 

TNESIS

IllllllllllllllllIlllllllllllllllllllllll

3 1293 01427 3753

This is to certify that the

dissertation entitled

THE INTERACTION OF APIS MELLIFERA AND VARROA JACOBSONI
POPULATION DYNAMICS IN MICHIGAN: SIMULATION MODELING
AND FIELD BIOLOGY

 

 

presented by

Ahmad A1 Ghamdi

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Entomology

 

 

@«4—W

Roger A. Hoopingarner

 

Major professor

Date June 13, 1997

 

MSU is an Affirmariw Action/Equal Opportunity Institution 0-12771

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

  

DATE DUE DATE DUE DATE DUE

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunlty Institution -

THE INTERACTION OF APIS MELLIFERA AND VARROA JA COBSONI
POPULATION DYNAMICS IN MICHIGAN: SIMULATION MODELING
AND FIELD BIOLOGY
By

Ahmad Al Ghamdi

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Entomology

l 997

ABSTRACT
THE INTERACTION OF APIS MELLIF ERA AND VARROA JA COBSONI
POPULATION DYNAMICS IN MICHIGAN: SIMULATION MODELING
AND FIELD BIOLOGY
By
Ahmad Al Ghamdi

This dissertation was divided into two components, computer simulation and field
biology. The computer simulation is interactive and demonstrates the relationship
between the life histories of Varroa jacobsoni and Apis mellifera. The field component
studied the development of the mite, its reproductive biology and some diagnostic
methods used for detection.

The simulation model was developed using a software package known as Stella II.
It generates population statistics at regular intervals throughout a designated time period
and outputs diagrams, equations, tables and graphs. The model is designed as a tool that
can help a user predict the best time for applying either a biological or chemical control,
or help a researcher decide which characteristic trait can have the greatest impact on the
mite population.

The field component of this project consisted of an experiment that had four
treatments (one control and the other three received 5, 10, and 25 mites, respectively)
with five replications. The mite population was monitored every other week from May
until October 1994. Two different methods of estimating the population were used; one
from mite downfall from sticky board and the other from live bee and brood populations.
Over the period of one summer, the mite population increased 81, 188, and l93-fold for

the groups that were infected with 25, 10 and 5 mites, respectively, from the sticky board

estimate. Mite population estimates were 2,032, 1,880, and 968 mites for the three
groups, respectively, from the sticky board. The mite estimate from the live bee
population was always larger than the one obtained from the sticky board. The sticky
board method was better than the adult bee and brood samples for the initial detection of
mite populations at low infestation levels. In addition, it was found that the mean number
of offspring reaching maturity are 1.82 and 2.69 offspring in worker and drone cells,
respectively (only the mothers that reproduced males and females were included). In
multiple infested cells, the average number of offspring was 1.26 and 2.03, respectively.
The study found that 82 and 90% of the mites were fertile in worker and drone cells,

respectively.

DEDICATION

To my wife and my sons Zyaid and Rami

Thank you for putting up with me during my study

iv

ACKNOWLEDGMENTS

I wish to thank my major advisor, Roger Hoopingarner. I think I was very lucky
to have him for my major professor. When I needed to talk to him he was always
available, and I appreciated the effort he made on my behalf. I would also like to thank
my guidance committee George Ayers, Stuart Gage, Fred Stehr, and Robert Husband for
their guidance during my exams, proposal writing and especially for encouraging me to
go and talk to them at any time. A special thanks to Robert Husband for his editing and
valuable comment on the first three chapters.

I would like also to than Camazine from Pennsylvania State University and Fries
from Sweden University for sending me their model which helped me a lot in my start.

I would like also to thank my graduate student colleagues for their friendship and
helping me in editing my writing, specially Becky Gore for her help in the modeling.

Finally, I would like to thank King Saud University for supporting my study in

Michigan.

TABLE OF CONTENTS

INTRODUCTION ...................................................... 1
LITERATURE REVIEW ................................................. 4
Life Cycle ....................................................... 5
Impact of Mite on Bee Populations ................................... 7

Direct Impact .............................................. 7

Indirect Impact ............................................ 7

Impact on the Bee Colony .................................... 9

Impact on Honey Production ................................. 9

Literature Cited ................................................. 10

DEVELOPMENT OF THE MII'E VARROA JA C OBS 0N1 IN THE HONEY
BEE COLONIES APIS MELLIFERA IN MICHIGAN AND COMPARISON
OF DIAGNOSTIC METHOD FOR DETECTION OF LOW INFESTATION

LEVEL IN THE COLONIES ............................................ 14
Abstract ....................................................... 14
Introduction .................................................... 15
Materials and Methods ............................................ 16
Results ........................................................ 18
Discussion ..................................................... 29
Literature Cited ................................................. 31

vi

REPRODUCTIVE BIOLOGY OF VARROA JA C OBS ON!
IN WORKER AND DRONE BROOD OF THE HONEY BEE

APIS MELLIF ERA UNDER MIDWEST CONDITIONS ....................... 35
Abstract ....................................................... 35
Introduction .................................................... 36
Materials and Methods ............................................ 37
Results ........................................................ 39
Discussion ..................................................... 46
Literature Cited ................................................. 50

MODELING OF HONEY BEE AND MlTE

POPULATION DYNAMICS ............................................ 53
Abstract ....................................................... 53
Introduction .................................................... 54
Materials and Methods ............................................ 62

Bee submodel ............................................. 62
Mite submodel ............................................ 70
Results and Discussion ........................................... 75
Literature Cited ................................................. 80

vii

USING MODEL SIMULATIONS TO PREDICT POPULATION RESPONSES
IN HONEY BEES AND MITES BY INTRODUCING BIOLOGICAL
CONTROL, CHEMICAL CONTROL, REINVASION, AND GENETICALLY

MANIPULATED CHARACTER TRAII‘S INTO THE SYSTEM ................ 86
Abstract ....................................................... 86
Introduction .................................................... 87
Biological and Chemical Control and the Effect of
the Reinvasion on Mite Population Development ...................... 88

Introduction .............................................. 88
Drone Removal ........................................... 89
Literature Review .................................... 89
Model Simulation ................................... 90
Trapping Comb Technique .................................. 93
Literature Review .................................... 93
Model Simulation ................................... 94
Hyperthermia ............................................. 99
Literature Review .................................... 99
Model Simulation .................................. 102
Chemical Control ........................................ 104
Literature Review ................................... 104
Model Simulation .................................. 104
Reinfestation ............................................ 107
Literature Review ................................... 107
Model Simulation .................................. 110

viii

Selection Toward Varroatosis Resistance ............................ 110

Introduction ............................................. 110
Grooming and Hygienic Behavior ............................ 112
Literature Review ................................... 112

Model Simulation ................................. 114

Infertility and Fecundity .................................... 114
Literature Review ................................... 114

Model Simulation for Fertility ........................ 117

Model Simulation for Fecundity ....................... 117

Postcapping Period ....................................... 120
Literature Review ................................... 120

Model Simulation .................................. 121

Reproductive Cycle ....................................... 121
Literature Review ................................... 121

Model Simulation .................................. 124

Phoretic Periods .......................................... 124
Literature Review ................................... 124

Model Simulation .................................. 126

Discussion ................................................. 126
Literature Review ............................................... 129
DISCUSSION ....................................................... 138

ix

LIST OF TABLES

DEVELOPMENT OF THE MITE VARROA JA C OBS ON!

IN THE HONEY BEE COLON IES APIS MELLIFERA

IN MICHIGAN AND COMPARISON OF DIAGNOSTIC METHOD

FOR DETECTION OF LOW INFESTATION LEVEL IN THE COLONIES

Table 1. Comparison between different sampling methods for
detection of Varroa jacobsoni at different infestation levels,
including their correlation coefficient (r) and the associated
level of significance (p) ..................................... 27

REPRODUCTIVE BIOLOGY OF VARROA JA C OBS ON!
IN WORKER AND DRONE BROOD OF THE HONEY BEE
APIS MELLIFERA UNDER MIDWEST CONDITIONS

Table 1. Varroa jacobsoni reproduction in Apis mellifera worker
cells under Michigan conditions .............................. 39

Table 2. Number of daughter mites produced per Varroa
jacobsoni mother mite in worker brood under Michigan
conditions ................................................ 40

Table 3. Varroa jacobsoni reproduction in Apis mellifera drone
cells under Michigan conditions .............................. 42

Table 4. Number of daughter mites produced per Varroa
jacobsoni mother mite in drone brood under
Michigan conditions. ....................................... 43

Table 5. Average number of offspring per mother mite taking
number of mother mites per cell into consideration ............... 46

MODELING OF HONEY BEE AND MITE

POPULATION DYNAMICS
Table 1. Parameters used for model simulation of Varroa jacobsoni and
honey bees Apis mellifera ................................... 63

xi

LIST OF FIGURES

DEVELOPMENT OF THE MITE VARROA JA C OBSONI IN THE HONEY

BEE COLONIES APIS MELLIF ERA IN MICHIGAN AND COMPARISON
OF DIAGNOSTIC METHOD FOR DETECTION OF LOW INFESTATION
LEVEL IN THE COLONIES

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Average number of varroa mites per treatment group (5, 10, 25
mites in initial inoculation), calculation on average daily
mite downfall (x 120, Liebig et al., 1984). The last sampling

date estimate has no brood in the colonies ...................

Average number of varroa mites per treatment group (5, 10, 25
mites in initial inoculation), calculation based on number of
mites per live bee * number of bees + number of mites per
brood cell * number of brood cells. Samples on 15th

of October from adult bees only ...........................

Average number of dead varroa mites per treatment group
(5, 10, 25 mites in initial inoculation) per three days collected

on the sticky board. .....................................

Average number of mites per 100 worker cells per treatment

per group (5, 10, 25 mites in initial inoculation) ...............

Average number of mites per 100 drone cells per treatment

group (5, 10, 25 mites in initial inoculation). .................

The percentage of adult bees infested by varroa mites per

treatment group (5, 10, 25 mites in initial inoculation) ..........

The average bee population per treatment group (5, 10, 25

mites in initial inoculation) ...............................

The average total brood p0pulation per treatment group

(5, 10, 25 mites in initial inoculation) .......................

xii

20

22

23

23

24

28

28

REPRODUCTIVE BIOLOGY OF VARROA JA C OBS ON]
IN WORKER AND DRONE BROOD OF THE HONEY BEE
APIS MELLIF ERA UNDER MIDWEST CONDITIONS

Figure 1. Average number of adult females produced by mother mites. ....... 45

MODELING OF HONEY BEE AND MITE

POPULATION DYNAMICS
Figure 1. Honey bee component of the modelling project. .................. 64
Figure 2. Mite component of the modelling project ....................... 65
Figure 3. Honey bee egg laying components of the modelling project ......... 66
Figure 4. Mite population components of the modelling project ............. 71
Figure 5. Adult bee population deve10pment when starting with

different initial populations in January ......................... 76
Figure 6. Worker and drone brood population development ................ 76
Figure 7. Mite population in correlation with worker and drone brood

development .............................................. 78
Figure 8. Mite population development in the first year when starting

in April with different initial infestations ....................... 78
Figure 9. Comparison between mite population development when

starting with different initial infestations in April until the

collapse of the colony the second year .......................... 79

USING MODEL SIMULATIONS TO PREDICT POPULATION RESPONSES
IN HONEY BEES AND MITES BY INTRODUCING BIOLOGICAL
CONTROL, CHEMICAL CONTROL, AND GENETICALLY MANIPULATED
CHARACTER TRAITS INTO THE SYSTEM

Figure 1. Mite and bee population deveIOpment when starting with

10 mites in April and 10,000 adult bees in January until the

colony collapses in the second year ............................ 91
Figure 2. Mite p0pulation development in the first year when drone

removal technique is used to control the mite population ........... 91

xiii

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Mite population development when using drone removal

technique to control the mite population: different percentages

from the whole brood is removed during the whole

season ................................................... 92

Mite and bee population development when using trapping
comb technique at the end of April the first year to
control Varroa population ................................... 95

Mite and bee population development when using trapping
comb technique at the end of July the first year to control
Varroa population ......................................... 95

Mite and bee population development in the first year when
using trapping comb technique at the end of April
to control mite populations .................................. 96

Mite and bee population development in the first year when
using trapping comb technique at the end of July to control
mite populations ........................................... 96

Mites and bee population development until the collapse of
the colony when trapping comb technique is used in the end
of April the first year and end of July the second year .............. 98

Mite and bee population development until the collapse of the
colony when using trapping comb technique in the end of
July the first year and end of July the second year ................. 98

Mite population development showing the effect of trapping
comb technique in first and second year ....................... 100

Mite populations for the first two years when using trapping
comb technique in the end of April or the end of July
the first year ............................................. 100

A comparison of mite populations when using the trapping

comb technique for three years 1—end of April the first year

& end of June the second year &. end of April the third

year 2-end of June the first year & end of June the second

year & end of April the third year ............................ 101

Mite and bee population development when using trapping
comb technique in the end of July first year and end July

second year and 99% drone removal in the third year ............. 101

xiv

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Comparison between mite populations in the first year with
and without hyperthermic control ............................ 103

Mite population development until the collapse of the colony
when hyperthermic technique is used to control the mite
p0pulation .............................................. 103

Mite and bee populations when using chemical control with
70% efficacy at the end of August for 65 days .................. 105

Mite and bee populations when using chemical control with 90%
efficacy at the end of August for 65 days ....................... 105

Mite and bee populations in one year when using chemical control
with 99% efficacy in the end of July for 7 days only .............. 106

Mite and bee populations until the collapse of the colony
when using chemical control with 99% efficacy at the end of
August for 7 days only ..................................... 106

Comparison of mite populations when using chemical control
in the end of August with 95% efficiency for 7 or 14 days ......... 108

Mite population development in the first year with and without
reinfestation ............................................. 1 l 1

Mite population development without re invasion or when 300
mites are introduced in one day by robber bees at the end of
August ................................................. 111

Mite populations in the first year when starting with 10 mites

in April and when having different percentages reduction of

offspring per worker and drone cell as the result of hygienic

behavior ................................................ 115

Mite populations when female mites have different fertility
rates in worker cells ....................................... 118

Mite population development until the colony collapse when
Varroa females have different fertility rates .................... 118

Mite population development in one year with different average
number of offspring ....................................... 1 19

XV

Figure 27.

Figure 28.

Figure 29.

Figure 30.

Figure 31.

Figure 32.

Mite population development in the first year when worker
brood have different post-capping periods ...................... 122

Mite population development until the colony collapse when
worker brood have different postcapping periods ................ 122

Mite and bee populations development when worker brood
have 10 day post-capping period ............................. 123

Mite population development when female mites have different
reproductive cycles ........................................ 125

Mite population development in the first year when female mites
have different phoretic periods .............................. 127

Mite p0pulations development until the colony collapse when
female mites have different phoretic periods .................... 127

xvi

INTRODUCTION

The mite Varroa jacobsoni is a parasite of Apis mellifera, and constitutes one of the
greatest dangers for modern beekeeping (Griffiths and Bowman, 1981; Boot et al., 1993).
The reduction of the beekeepers in the US. in the last five years from 211,000 to 125,000
is mainly attributed to the tracheal mite (Acarapis woodz) and Varroa mite. Since Varroa now

is the major problem, many scientists think the loss of bees in the coming years will be

greater.

Comparisons from throughout the world suggest that Varroa genotypes or an
interaction of Varroa genotype and the environment produce varied degrees of health
problem for colonies. Varroa in Europe was thought to have originated from Ussuria
(Ruttner, 1983) and is clearly a problem for European bees. Varroa mites in the US are
thought to have spread from South America (Delfinado-Baker and Houck, 1989). This same
genotype of Varroa, is believed to have been introduced to South America in 1971 from
Japan (De Jong et al., 1982). In Brazil, this mite does not seriously debilitate either African
honey bees or European Honey bees in Brazil (Moretto et al., 1991). In Europe, if the
population is not controlled, colonies infested with Varroa die in three to four years (Ritter,
1984; Fries, 1991). Deflinado-Baker (1989) suggested a lower virulence of mites from North

America compared to mites from Europe on the basis of the hypothesis of South American

2

origin of mites in the US. However, this has been proven wrong (Page and Kraus, 1995).

Initial observations in Michigan agree with Page and Kraus’s finding.

More than 174 chemicals have been used or tested for controlling Varroa (W ienands,
1988). Pyrethroid Fluvalinate (Apistan), the most common acaricide that has been used to
control Varroa all over the world, is the only one approved for controlling these mites in the
United States. Beekeepers are new pesticide users and they are using and misusing the
pesticide in the US. (J aycox, 1989). There are reports about Varroa mites developing
resistance to Apistan in Italy (Colombo et al., 1993; Lodesani et al., 1995) and the pesticide
residues in honey and wax from Apistan and other registered products ( Milani, 1994;
Hansen and Petersen, 1988; Barbina et al., 1990). Integrated pest management programs
have been used for control of Varroa in some countries and show good promise for small
scale beekeeping.

This study consist of two parts, one is building a comprehensive model for the life
history of the honey bee Apis mellifera and the bee mite Varroa jacobsoni and their
interaction. The second part is the study of the population development for honey bees and
Varroa mite and the reproductive biology of the mites in the field.

This dissertation is compilation of four ready-to—be-published papers. The first paper
is entitled , “Development of early infestation by the mite Varroa jacobsoni in honey bee
colonies Apis mellifera in Michigan”. This paper investigates the population growth of V.
jacobsoni under Midwest conditions and investigated methods of detecting mite infestation
at low rates and correlates these infestations to mite population levels in Michigan honey bee

colonies. The second paper is entitled, “A Reproductive biology of Varroa jacobsoni in

3

worker and drone brood of the honey bee Apis mellifera L. under Midwest conditions”. This
paper reported number of offspring produced by Varroa mites in drone and worker cells,
number of offspring in single infested cells verses multiple infested cells, fertility rates of
Varroa mites in worker and drone cells, percent of female mites that produce males only in
worker and drone cells, percent of mother mites that died in the brood cells.

The third paper is entitled, “Modeling of honey bee and mite population dynamics”.
This paper describes in detail the model building of the life history of the honey bee and
Varroa mites using a commercial software package known as Stella II. Also, it shows the
interaction between the bees and mite population.

The fourth paper is entitled, “Using model simulations to predict population
responses in honey bees and mites by introducing biological control, chemical control, and
genetically manipulated character traits into the system”. This paper illustrates the
importance of modeling in predicting the effect of applying chemical or biological control
on mite populations and it helps in defining when the best time in the season to treat. More
importantly, it shows how Varroa population dynamics are used as a tool to evaluate the
effect of separate resistance traits on the population growth of the mites and it helps to clarify

which resistance traits should preferably be selected.

LITERATURE REVIEW

Apis cerana, the Eastern honey bee, was the original host of Varroa jacobsoni.
Initially discovered on the island of Java by Jacobson in 1904, it was subsequently described
by the Dutch zoologist, Oudemans, and named after its discoverer (Oudemans, 1904). The
shift of Varroa jacobsoni from Apis cerana to Apis mellifera probably occurred in the 1940’s
in the USSR. However, the first reports of this event were published in the 19505
(Schimanuki, 1993). Varroa jacobsoni was discovered for the first time in the US. in 1987
(Dietz and Hermann, 1988). Currently, the mite is present in more than 85 countries and is
considered one of the most serious pests of A. mellifera colonies in most of the world

(Matheson, 1993, 1994, 1995).

The female mite is a dark red-brown and is almost 1.5 mm in diameter. It is the only
common parasite of honey bees that can be seen with the naked eye and identified with a

hand lens.

The mite reproduces mainly on drone pupae of A. cerana (De Jong, 1988), and on
worker pupae of A. mellifera, although it prefers the drones. Live mites usually occur inside
sealed brood cells or are partly hidden between the abdominal segments of adult bees so, in
spite of their size, they are not easily noticed by beekeepers. Dead mites, which fall from

adult bees, can be seen fairly readily in debris from the hive floor.

LIFE CYCLE

The previously inseminated reproductive female mites enter a drone or worker brood
cell before the honey bee larva (5-days old) in the cell is capped (T ewarson, 1983). They
may feed on the larva initially, but they quickly crawl underneath the larva and immerse
themselves in the brood food, at least some of which they ingest (De J ong et al., 1982), with
only their movable finger-like periterrnes (breathing tubes) exposed ( Morse and Hooper,
1985). They remain in the brood food, oriented with their ventral side towards the opening
of the cell, until the larva eats the food and thereby cleans off and frees the mites. As many
as 21 adult female mites may be seen, apparently immobilized, immersed in the brood food
in one cell (De Jong et al., 1982). Bradbear (1988) and Tewarson (1983) stated that the
greatest number of mites (up to 20) can hatch from a drone cell. When the larva does not eat
all the food, the mite is caught and dies (Beetsma, 1983). After the mites are freed, they
begin feeding on the haemolymph of the larva and later of the pupal bee (De J ong et al.,
1982). Once the Varroa jacobsoni female has fed on the bee haemolymph, maturation of the

first sperrnatocyte and fertilization of the first oocyte occurs (Ramirez and Otis, 1986).

The female mite lays the first male egg about 60-64 h after the cell is capped (Martin,
1994; Rehm and Ritter, 1989); the next eggs, usually female, are laid at 30 h intervals. The
eggs are laid singly on the cocoon, on the covering of the pupa, or rarely, on the larva or
prepupa, or on the wall of the cell (Tewarson, 1983), or between the developing larva and
the wall of the cell (Bradbear, 1988; Martin, 1994). One to seven eggs are laid (Martin,
1995). The mite probably feeds for the last time, and oviposition ends, when the worker or

drone pupal eyes become dark (for workers, day 18, or hour 216-220 of pupal development;

6

for drones, day 19 or hour 240 after leaving the first larval instar) (Ifantidis, 1983). If several

infective females enter the same brood cell, not all of them necessarily reproduce (Ramirez

and Otis, 1986).

A six-legged protonymph hatches from the egg and feeds actively for 48 h after which
it molts into an eight-legged deutonymph which also actively feeds on the haemolymph of
the pupa. It is similar to the adult mite except for its white color. After about 3 days, the
deutonymph molts into an adult which still has a lighter color than the mother mite. The
complete development inside the bee cell of the female mite takes 7.5 days and 5.5 days for
the male (Ifantidis, 1983), (8-10 for the female and 6-7 for the male; Grobov, 1977;

Tewarson, 1983).

The female V. jacobsom' mites reach maturity inside the capped brood cell, 24 h after
becoming adults (Ifantidis, 1983). By that time the males that moult to adults, approximately
220 h after larval engorging, are probably ready to mate with the females, and they
presumably search for the females and mate with them. The male and immature female

offspring die shortly after the brood cell is uncapped and the adult bee emerges.

The infective adult female mite, whether fertilized or not, leaves the honey bee cell
by phoresy on an emerging bee, or by herself, after which the mite’s chelicerae (mouthparts),
pierces the intersegmental membranes (or other soft parts) of the bee and feeds on its
haemolymph for several days. The mites are phoretic between 1-3 weeks before they enter

a cell with a larva (Boot et al, 1994), where they feed on the prepupa and begin laying eggs.

7
IMPACT OF MITE ON BEE POPULATIONS

The mite can impact the bee population in a variety of ways. First, it has a direct
impact because it parasitizes the bee. Secondly, it can change the behavior of the adult bee.
In addition, it often brings secondary diseases into the bee colony.

Direct impact. Mite infestations shorten the life span of the bee. A summer bee
lives for approximately 28 days, whereas a bee infested with two mites during its
development may live for only seven days, and if infested with 6 mites dies shortly after
hatching (De Jong et al., 1982). Goncalves et al. (1985) stated that the mean life span of
bees is reduced by half. The life span of the Varroa diseased bee depends on the bee’s age
when the mite fixes itself on it (Dietz and Hermann, 1988).

When a larva is infested with 1,2,3,4, or 6-8 mites, the loss of weight of the resulting
adult is 6.5%, 10%, 11%, 14,% 17% and 25% respectively (Beetsma, 1983). The life
expectancy of adult bees is 28, 18, and 9 days when the pupae have been infested with 0, 1,
and 2 or more mites (Beetsma, 1983).

Schneider (1986) stated that the number of spermatozoa decreased by 50% when the
drone was infested with more than 3 mites. In varroatosis colonies, worker and drone

emergence is delayed by 2-4 days (Smimov, 1978).

Indirect impact. Along with directly shortening the bee’s life span, the mite is
responsible for carrying pathogens to the bee. It has been determined that the Varroa mites
can carry the bacterical pathogens of haffniosis, virus paralysis and nosema diseases
(Kutsenko, 1975; Mikitiuk et al., 1975; Sidorov and Stolvov, 1975). Platukhina et al. (1975)

separated 22 species of micro—organisms from Varroa, of which two are pathogens of bees.

8

In general, varroa infested bees have 2.2 times greater bacteria in the haemolymph
than healthy bees (Smimov, 1978). Koch and Ritter (1987) could find significantly more
bacteria in the haemolymph of adult bees which had been infested with three mites than the

uninfected control group or in bees which were infested with only one or two mites.

Smimov and Kudriavtsev (1977) obtained experimental evidence of the Varroa mites

being able to carry the pathogen agent of AFB from diseased to healthy colonies.

Batuev (1979) did laboratory tests where Varroa mites were placed on bees that had
been injected with acute paralysis virus (APV). The following day these mites were
transferred to healthy test bees. After 20 days only about 9% of the bees were alive,
compared with 60% of the control. This virus is relatively common, but it only passes into
the haemolymph and attacks vital organs if it can propagate on a massive scale as in the
presence of varroatosis (Bailey, 1981). Ball (1985) demonstrated its presence in heavily
infested colonies. Bailey and Ball (1991) reported that mites activate the acute paralysis
virus when they attack the bees and then transmit it to kill further individuals. It is estimated
that varroa can transmit APV from one honey bee pupa to another with an efficiency of about
70% (W iegers, 1986). Varroa also act as a vector for sacbrood and black queen cell virus.
Bailey and Ball (1991) stated that the deformed wings of the adult that are produced from the
parasitized pupa is invariably caused by deforrned wing virus which is transmitted by mites.
Strick and Made] (1988) stated that Hafnia was successfully transmitted from septic infected
test pupa to healthy pupa by the mite. The effect of the parasitization depended upon the

number of mites that attacked. Seventeen percent of the pupae suffered from septicemia if

9

one mite attacked, 30.7% for two mites, 81.8% if attacked by four mites; and 95% if attacked

by seven mites.

The effect of the pathogen is varied, depending both on the type of pathogen and the
amount present. Secondary infections lead to decreases in bee life expectancy (Ritter, 1988),
and also to a reduction in both the drive for self-cleaning and the drive to care (Smimov,

1978).

Impact on the bee colony. Salchenko (1971) observed that infestation levels of
mites in bee colonies ranged between 3% and 85% when bees prepared for winter. German
research has suggested that a critical limit is 10 mites/day/colony on the bottom sticky sheet.
This roughly corresponds to a mite population ranging from 1000 to 1500 mites per colony.

In 1989 Jaycox lowered that threshold to three to five mites per day.

It is estimated that if two percent of the worker brood is infested that the population
of the hive will be reduced by 1%. Colonies in which 20 mites have been found for every
100 bees in the autumn become very weak. In those in which greater than 50 mites per 100
bees were found, all the bees died. In colonies fed sugar (9-10 kgs) in autumn, all the bees
died when the ratio was 10-14 mites per 100 bees (Smimov, 1975; Nikolski and
Evdokimova, 1975). It has been estimated that the critical threshold of mites which cause
damage to a normal colony is approximately 5,000 and that bee colonies with mite

populations that reach an excess of 10,000 will not survive (Mautz, 1987).

Impact on honey production. Since the mean life expectancy of bees infested with
varroa during development is reduced by one half, then it is reasonable to estimate that honey

production is also reduced by the same half. If two percent of the worker brood is infested,

10

the population of the hive will be reduced by one percent, and it may be estimated that this
would cause a one percent reduction in honey production (Issa et al., 1985). Woo (1993)
also found higher infection rates of 30.0 and 52% corresponded to a 30.65 to 46.0%

reduction in honey yields.

LITERATURE CITED

Bailey, L; Ball, BV (1991): Honey Bee Pathology. 2nd ed. Academic Press, New York.

Bailey, L (1981):Honey bee pathology. Acacemic Press Inc.,London. 123 pages.

Ball, BV (1985): Acute paralysis virus isolates from honeybee colonies infested with Varroa
jacobsoni. J. Apic. Res. 24(2), 115-119.

Barbina, MT; De-Paoli, M; Valentino, A (1990): Determination of tau-fluvalinate residues
in honey. Pestic. Sci. 28(2), 197-202.

Batuev, Yu M (1979): New Information about viral paralysis. Pchelovod. 7, 10-11.

Beetsma, J (1983): Some results of studies by Dr. D. Jong (Brazil). In: Proceedings of a
Meeting of the EC Experts’Group/Wageningen 7- 9 February 1983. (Ed: Cavalloro,
R) A. A.

Boot, WJ; Calis, JNM; Beetsma, J (1993): Invasion behavior of Varroa mites into honey bee
brood cells. In: Asian Apiculture. (Eds: Connor, LJ; Rinderer, TE; Sylvester, HA;
Wongseri, S) Wicwas Press, Cheshire, Connecticut, 491-498.

Boot, WJ; Sisselaar, DJA; Calis, JMN; Beetsma,J (1994): Factors affecting invasion of
Varroa jacobsoni (Acari: Varroidae) into honeybee, Apis mellifera (Hymenoptera:
Apidae), brood cells. Bull. Entomol. Res. 84, 3-10. Bowman,C) Ellis Horwood,
Chichester, 349-359.

Bradbear, N (1988): Varroa Jacobsoni. International Bee Research Association Leaflet 3,
4.

De J ong, D (1988): Varroa jacobsoni does reproduce in worker cells of Apis cerana in South
Korea. Apidol. 19, 241-244.

De Jong, D; Morse, RA; Eickwort, GC (1982): Mite pests of honey bees. In: Annual Review
of Entomology. Vol. 27. (Eds: Mittler, TE; Radovsky,FJ; Resh, VH) Annual
Reviews, Inc., Palo Alto, CA, 229-252.

11

Delfinado-Baker, M; Houck, MA (1989): Geographic variation in Varroa jacobsoni (Acari,
Varroidae): Application of multivariate morphometric techniques. Apidol. 20(4),
345-358.

Dietz, A; Hermann, H (1988): Biology, Detection and Control of Varroa jacobsoni: A
parasitic Mite on Honey Bees. Lei-Act Publishers, Commerce. 75 pages.

Fries, I; Aarhus, A; Hansen, H; Korpela, S (1991): Development of early infestations by the
mite Varroa jacobsoni in honey-bee (Apis mellifera) colonies in cold climates.
Exptl. & Appl. Acar. 11(2-3), 205-214.

Goncalves, L; DE Jong, D; Morse, R (1985): The truth about Varroa in Brazil. In: The
XXXth Int. Apic. Cong., Japan. Apimondia,.

Griffiths, D; Bowman, C (1981): world distribution of the mite Varroa jacobsoni, a parasite
of honeybees. Bee World 62, 154-163.

Grobov, O (1977): Varroasis In Bees. In: Varoasis, a honeybee diseases. Apimondia
Publishing House, Bucharest,, 46-69.

Hansen, H; Petersen, JH (1988): Residues in honey and wax after treatment of bee colonies
with bromopropylate.Tidsskr. Planteavl. 92(1), 1-6.

Ifantidis, MD (1983): Ontogenesis of the mite Varroa jacobsoni in worker and drone
honeybee brood cells. J. Apic. Res. 22(3), 200-206.

Issa, M; Goncalves, L; De Jong, D (1985): Study of the preference of the mite Varroa
Jacobsoni for Apis Mellifera Drones. In: The XXXth International congress, Japan
from October 10-16, 1985. Apimondia, 157-158.

J aycox, E (1989): Integrated pest management programs. Gleanings in bee culture October,
556-558.

Kraus, B; Page, RE, Jr. (1995): Population growth of Varroa jacobsoni Oud in
Mediterranean climates of California. Apidol. 26, 149-157.

Kutsenko, I (1975): (in Russian) cited in Smimov 1978.
Langhe, A (1976): Biology of Varroa mites. Veterinariya 7, 74-77.

Lodesani, M; Colombo, M; Spreafico, M (1995): Ineffectiveness of Apistan registered
treatment against the mite Varroa jacobsoni Oud in several districts of Lombardy
(Italy). Apidol. 26(1), 67-72.

Loglio, G (1994): (Varroa jacobsoni: is it becoming resistant to fluvalinate?). Apiculture
abstracts 45. No. 3, 273.

Martin, S (1994): Ontogenesis of the mite Varroa jacobsoni Oud. in worker brood of the
honeybee Apis Mellrfera L. under natural conditions. Exptl. & Appl. Acar. 18, 87-
100.

12

Mautz, D; Hirschmann, W; Kemnitzer, F (1986): The embryonic development of Varroa
jacobsoni Oudemans 1904 (V arroinae, Mesostigmata). Acarol. 27(3), 203-210.

Mautz, D (1987): Uber die Folgen der Varroatose und eines Verzichts auf Bekamp
fungssnahmen. Irnkerfreund 42, 330-333.

Mikityuk,V; Korzjhova, L; Sedin, I (1976): [Experiments on the biology of the Varroa mite].
Pchelovod. 12, 19-20 (in Russian).

Milani, N (1994): Possible presence of fluvalinate-resistant strains of Varroa jacobsoni in
northern Italy. In: New perspectives on Varroa. (United Kingdom) (Ed: Matheson,A)
International Bee Research Association, Cardiff, 87.

Moretto, G; GonHalves, LS; De Jong, D; Bichuette, MZ (1991): The effects of climate and
bee race on Varroa jacobsoni Oud. infestations in Brazil. Apidol. 22, 197-203.

Morse, R; Hooper, T (1985): The Illustrated Encyclopedia of Beekeeping. Blandford press,
Link House, pool, U K, 432.

Nikolki, O; Evdokimova, A (1975): Cited by Smimov 1978.

Oudemans, AC (1904): On a new genus and species of parasitic acari. Notes Leyden Mus.
24, 216-222.

Platukhink, N; Egorova, N; Stolbov, N (1975): (in Russian) cited in Smimov 197 8.

Ramirez, B,W; Otis, GW (1986): Developmental phases in the life cycle of Varroa
jacobsoni, an ectoparasitic mite on honeybees. Bee World 67(3), 92-97.

Rehm, S; Ritter, W (1989): Sequence of the sexes in the offspring of Varroa jacobsoni and
the resulting consequences for the calculation of the developmental period. Apidol.
20, 339-343.

Ritter,W ( 1988): Varroa jacobsoni in Europe, the tropics, and subtropics. In: African Honey

Bees and Bee Mites. (Eds: Needharn, GR; Page, RE, Jr; Delfinado-Baker, M;
Bowman, C) Ellis Horwood, Chichester, 349-359.

Ruttner, F (1983): Varroatosis in honeybees: extent of infestation and effects. In: Varroa
jacobsoni Oud. Affecting Honey Bees: Present Status and Needs. (Ed: Cavalloro, R)
A. A. Balkema, Rotterdam, 7-13.

Salchenko,V (1971): Cited in Smimov 1978.

Schneider, P (1986): In: Varroa workshop. Feldafing, 20.

Shimanuki, H (1993): Current status and future prospects for the control of bee mite. In:
Asian Apiculture. (Eds: Connor, L; Rinderer,T; Sylvester, H; Wongsiri, S) Wicwas
Press, Cheshire, Connecticut, USA, .

Sidorov, N; Stolbov, N (1975): (in Russian) cited in Smimov 1978.

13
Smimov, A; Kudriavtsev, E (1977): cited in Smimov 1978.

Smimov, AM (1978): Research results obtained in USSR concerning etiology, pathogenesis,
epizootiology, diagnosis and control of Varroa diseases in bees. Apiacta 13 (4), 149-
162.

Smimov, B (1975): cited in Smimov 1978.

Strick, H; Made1,G (1988): Transmission of the pathogenic bacterium Hafnia alvei to honey
bees by the ectoparasitic mite Varroa jacobsoni. In: African Honey Bees and Bee
Mites. (Eds: Needham, GR; Page, RE, Jr; Delfinado—Baker, M; Bowman, C) Ellis
Horwood, Chichester, 462-466.

Tewarson, N (1983): Nutrition and reproduction in the ect0parsitic honeybee (Apis sp.) mite,
Varroajacobsoni. Doktors Thesis, Biologi, Eberhard-Karla—Univ. Tubingen,

Tubingen, German Federal Republic. 71 p.

Wiegers, PP (1986): Transmission of honeybee viruses by Varroa jacobsoni Oud. In:
European Research on Varroatosis Control: proceedings of a meeting of the EC
Experts’ Group, Bad Homburg, 15—17 October 1986. (Ed: Cavalloro,R) Balkema,
Brookfield, VT, 99-104. (ill.)

Wienands, A (1988): Varroa research makes headway in West Germany. Amer. Bee J.
128(4), 265.

Woo, K (1993): The bionomics, distribution and control of bee mites (Varroa jacobsom) in
South Korea. In: Asian Apiculture. (Eds: Connor,L; Rinderer,T; tylvester,A)

DEVELOPMENT OF THE IVIITE VARROA JA CUBS 0N1 IN THE HONEY BEE
COLONIES APIS AIELLIFERA IN MICHIGAN AND COIVIPARISON
OF DIAGNOSTIC NIETHOD FOR DETECTION

OF LOW INFESTATION LEVEL IN THE COLONIES

ABSTRACT

Twenty honey bee colonies were divided into four groups one control and the other
three received 5, 10, and 25 mites respectively with five replications for each group. The
groups were separated from each other to reduce drifting. The development of the mite
infestation was monitored every other week from May until October 1994. Estimation of the
difl'erent mite populations were based on daily mite downfall, or a sample of 100adult bees,
and 100 worker and 100 drone broods and estimate of number of adult bees and brood cells.

Over the period of one summer, the mite population increased 81, 188, and l93-fold
for the group that were infected with 25, 10 and 5 mites respectively. Based on the daily mite
downfall, the population estimates were 2,032 ,1,880 and 968 mites for the groups that
started with 25, 10, and 5 mites respectively. The mite estimate fiom the live bee population
was larger than mite estimate obtained from sticky board. However, variation in mite
populations between the colonies was large. The sticky board method was better than adult

bees and brood samples for the initial detection of mite population at low infestation level.

14

15

INTRODUCTION

The mite Vm'roa jacobsoni Oudemans (Acari: Mesostigrnata) is a parasite of the
honey bees Apis cerana and Apis mellifera. It is present in more than 85 countries and is
currently considered one of the most serious pests of A. mellifera colonies in most of the
world (Matheson, 1993, 1994, 1995). There are a number of factors that efi‘ect the
reproductive rate of the mite. Reproduction occurs in sealed honey bee brood cells. Hence,
mite population growth occurs only in the presence of brood cells. Other factors include
host-specific effects of the bee (Moritz and Hanel, 1984; Buchler, 1990; Rosenkranz et al.
1990; Fuchs, 1991; Moretto et al., 1991b; Otten, 1991; Kulincevic etal., 1992), geographic
and climatic factors (De Jong et al., 1984; Moretto et al., 1991) and possibly Varroa
genotypes (Delfinado-Baker and Houck, 1989; Guzman et al., 1996).

In the Mediterranean climate of California the initial population is capable of
increasing 300-fold during one year (Kraus and Page, 1995). In colder, temperate climates
the increase averages about 10-fold per year (Ritter, 1984; Fries et al., 1991b; Korpela et al.,
1992); but can increase up to lOO-fold within one summer (Fries eta1., 1991). In tropical
climates the parasite seems to be less virulent (Ritter and De Jong, 1984). In sub-tropical
climates the infestation rate is lower than in temperate climates (Moretto et al., 1991b). If
the population is not controlled, colonies infested with V. jacobsoni die in three to four years
(Ritter, 1984).

An important part of control is the initial detection of the mite. Early detection with
low infestation rates is important in bee colony management. Ritter (1981) and De Jong
(1984) suggest the following diagnostic method: count mites that fall from the bee due to

natural causes, examine bee samples for phoretic mites, check capped brood samples for mites

16

while they are reproducing, and treat the colony with acaricides. For low infestation levels
(below ten mites), the use of acaricides may be the only efi‘ective method for detecting the
mite with acceptable levels of precision in broodless colonies (Ritter, 1984). If the mite
population is between ten and 100, then the examination of hive debris should allow for
detection (Ritter, 1984). In fact, Ritter reported that live bee or brood examinations are
insufficient for population levels below 100 mites per colony (1985). Liebig et al. (1984)
reported a close correlation between mites collected in hive debris and the size of the Varroa
mite population. Fries et al. (1991a) also compared difl‘erent diagnostic methods for
detection of low Varroa mite infestation levels. They concluded that debris was more
efi'ective than examining the brood itself for low infestation rates. They said it was preferable
to other methods because of its simplicity and eficacy.

The objective of this research project was to investigate the population growth of
Varroa jacobsoni under deest conditions and investigate methods of detecting mite
infestations at low rates and correlate it to mite population levels in Michigan honey bee
colonies. Initial observation suggested that the impact might be greater on bee colonies in
Michigan than in Europe. Delfinado-Baker and Hauch (1989) suggested that the European

mite has lower virulence based on the hypothesis that the mite originated in South America.

MATERIALS AND METHODS

Twenty packages of bees (0.9k of bees per package) were separated into four groups,
each containing five colonies. They were installed into single chamber Langstroth hives
containing honey and comb foundation on the first of May in East Lansing, Michigan. Each

colony was treated with two Apistan strips (10% fluvalinate). The four groups were placed

17

in separate locations on the Michigan State University campus in an effort to reduce drifting
between groups. Robbing screens also were used (Hoopingamer, 1982). All groups were
managed optimally as for honey production and were treated with Fumidil B (Mid-Con) in
May and treated with terramycin antibiotic (Pfizer) in May and July. One group was the
control (0 mites), and the other three were inoculated with 5, 10 and 25 mites, respectively.
The mites were removed from their host bees using C02, collected in small tubes and
introduced directly upon the bees on May 15. The experiment lasted about 30 weeks.

Total mite population was estimated using both live bees and hive debris. Live bee
estimates were taken every two weeks, beginning May 29th. Between 100 and 200 live adult
bees were taken from brood combs and stored in a deep freezer. The mites were separated
fiom the bees by vigorously shaking the bees in 70% ethanol for 3 to 4 minutes. The mites
were washed from the bees using a hand-shower over a double wire screen. Number of bees
and mites were counted to determine the level of infestation on adult bees, data were adjusted
to present number of mites per 100 bees. In addition, samples of 100 sealed worker and 100
drone brood cells were examined when brood was present. The cells were opened and the
number of adult females in the cells and on the bees were counted. The amount of brood in
the colonies was estimated in each colony using the double-sarnpling technique described by
Roger et al. (1983). The bee population was estimated as described by Burgett and Burikam
(1985)

The hive debris was collected weekly on a paper sheet placed on the bottom of the
hive to monitor the natural mortality. A wire screen prevented bees from gaining access to

debris (Ritter, 1981), and a plastic sheet was smeared with cooking oil to catch the mites.

18

Sheets were left in the hive for three days, and adult females were counted directly on the
paper. This was done over a 30 weeks period.

Total number of mites in each colony with brood was estimated by Fries et al. (1991b)
using the following two methods: (1) daily downfall: average daily mite downfall x 120
(Liebig et al., 1984); and (2) live bees: the infestation rate of live bees x number of bees +

infestation rate of brood x number of brood cells (Fuchs and Koeniger, 1984, 1996).

RESULTS

Although the experiment started with four treatments, the control colonies were lost
in the first week because of contamination by an infested swarm. One week after the start of
the experiment, one sticky board had 60 mites, the nearest one to that had 16 mites. The
other three sticky boards were without mites. The colony with 60 mites had 80 mites on the
sticky board and 59% adult infestation in November. The other colonies in this treatment
group gradually became infested also. Two of the five colonies survived the winter, and both
of them lud peak mite downfall of 20 and 30 on the sticky board in September, respectively,
and they entered the winter with an 18 and 22% adult bee infestation rate. Therefore, only the
three other treatments were analyzed. They included those inoculated with 5, 10, or 25 mites.

The entire five colonies that were inoculated with 25 mites were lost during the
winter. Their deaths could not be attributed completely to Varroa, but may be also due to
the cold winter. The maximum number of mites recovered from the sticky board was
recorded on the September 17th sampling date. The number of dead mites ranged from 40

to 63 and the adult infestation was between 24 to 34%.

19

Four of the five colonies that were inoculated with 10 mites survived the winter. The
colony that died did not have any more mites than the colonies that survived. The range of
mites recovered from sticky boards on September 24 ranged fi'om 25 to 71. The number
found in the dead colony was 65 and had 28% infestation of the adults in November. Three
of the five colonies that were inoculated with 5 mites died. Once again, there were no
difi‘erences in the number of mites recovered on the sticky board among the colonies. The
two that survived had 26 and 29 mites on the sticky board in September, while the three that
died had 18, 25, and 62 mites on their boards and entered the winter with 14, 17 and 21%
adult bee infestation rate. Two of the colonies that died had more foulbrood infection than
the rest, although they received the same medication treatments in May and July as the other
colonies.

Using the sticky board method of calculating mite density (Figure 1), choosing the
same dates when live bee samples were taken, the population ranged fiom 1640 to 2520
mites, with an average of 2,032 in the 25 mite inoculated treatment. The 10 mite inoculated
treatment had a smaller peak, ranging fiom 1,200 to 2,200, with an average of 1,880 mites.
The last treatment (5 mites) had an average of 968 mites, with a range from 520 to 1,480.

The mite estimate fiom the live bee population (Figure 2) was larger than the mite
estimate obtained fi‘om sticky board counts for all treatments. It averaged 1760 (range 578-
2,375), 2,247 (range 1,347-2,775), and 3,119 (range 2,130-3,834), for the 5, 10, and 25 mite
treatments, respectively.

Over the period of one summer, the mite population increased 193, 188, and 81-fold
for the groups that were infected with 5, 10 and 25 mites respectively, when using the sticky

board method of estimating mite population. This was an average of 154-fold increase. When

20

 

 

 

 

 

 

 

 

 

 

 

 

2500
2000 -
g 1500 -
E
'6
2 1000 ~
500 -
o l A
15-May 15-Jun 15-Jul 15-Aug 1569p 15-Od 123-Nov
Sawing data
Figure 1. Average number of varroa mites per treatment group (5, 10, 25 mites in

initial inoculation), calculation based on average daily mite downfall (x 120,
Liebig et al., 1984). The last sampling date estimate has no brood in the

 

 

 

   

 

 

 

 

 

 

colonies.
4
3 r—
E 1;; l .25 wires
3 g 2 - +10 «to:
Q "E at- 5 «its
1 h
o l j I
154mm 15-Jul 1.5-Aug 1560p 15-Oct
Sarrphg We
Figure 2. Average number of varroa mites per treatment group (5, 10, 25 mites in

initial inoculation), calculation based on number of mites per live bee *
number of bees + number of mites per brood cell * number of brood cells.
Samples on 15th of October from adult bees only.

21

estimating the population fiom the live bee method, there was a 352, 225, and 125-fold
increase, respectively. This method had an average increase of 23 4-fold.

The colonies that did survive the winter were weak the following spring and only
covered one or two combs and they had an average of 50% brood infestation rate on May 15.
Attempts were made to strengthen these colonies so they could be utilized for other
experiments, but they were unsuccessful. All of the colonies collapsed in September.

The number of mites found on the sticky boards (Figure, 3) increased slowly until the
middle of August for all three treatments. In the middle of August there was a sharp increase
in the number of mites found on all the sticky boards. This increase peaked on September
17th for the 10 and 25 mite inoculated colonies. This peak plateaued for two weeks and then
gradually declined until the population started to increase in March and April. For the third
treatment (colonies inoculated with 5 mites), the downfall began to sharply increase at the end
of August similarly to the other two treatments. It sharply increased until the 17th of
September (where the other two treatments peaked), but continued to increase until the 7th
of October. This group did not experience a plateau, but declined immediately, until the
downfall increased in March and April when the bee colony began its brood rearing.

The infestation rate on the adult and brood population (Figures 4, 5, 6), increased
drastically when the brood and adult population declined at the end of the season. In general,
the trend in these graphs are in agreement with the results fi'om natural death rate. The level
of infestation on adult bees (Figure 4) ranged from 0.3 to 3% the majority of the summer and
was the highest on wintering bees at the end of November when there was no brood and the
bee population was low. The adult infestation rate varied widely between colonies within the

same group. For example, in November there were between 25 to 34%, 13-28% and 7-17%

22

Ewen 303m 2: so 380:8 99% ooh: .8; E23302: 355 E
825 mm .2 .3 macaw 2258: con 8:8 85> How me .6983: omega

.m 2&5

 

 

839.5

22.10 33.00 non—+0 0.0...» 9.0.2. 806m «5.: sin 5.30 05w use .3 .io Earn, :3
33.: nonfi- ::p 0.0.0. .62.: 80.; 80.3 5.0.: .3..." can. 3N 3. 5:3 cgw scab"

d udi 11‘1,d4 ulfiJ‘d - u d u q 1. I J ‘1ng
{/ \Ifl‘x\l\£

 

  
  

HICOw I?

 

Sign“ ul-

 

 

 

 

0'

“PW

 

 

23

 

 

I of mites

7O

 

 

 

50 '

40 -

 

 

: ------ 25mites
3° ' a - - - 1O mites
20 I 1 ’ —5mites

 

 

 

 

 

10

 

 

 

 

 

27- 15- 29- 15- 2- 19- 6- 12- 4-
May Jun Jun Jul Aug Aug Sep Sep Oct

Months

 

Figure 4. Average number of mites per 100 worker cells per treatment per group

(5, 10, 25 mites in initial inoculation)

 

 

I of mites

 

 

18

16

14

12

1O - - - 25mites
------ 1O mites
—-—5mites

 

 

 

 

8
6
4
2
0

27-May 15-Jun 29-Jun 15-Jui 2-Aug 19-Aug 6-Sep
Days

 

Figure 5. Average number of mites per 100 drone cells per treatment group

(5, 10, 25 mites in initial inoculation)

 

24

Aces—:85 3:5 5 3:8 mm .2 .3 963 “segue: con 8:8 not? .3 Bamomfi moon :33 Lo 033023 2F

 

 

 

_ was?! 8.22.9. 8.2.3...

 

2% 8.95m

 

 

 

 

 

 

 

 

.e same

25

for groups one, two and three, respectively. Also wide variation occurred in the number of
mites in worker and drone cells, e. g., the number of mites ranged between 30 -200 mites per
100 worker cells in group one on October 4th.

One of the objectives of this study was to compare the two methods for estimating
the mite population The sticky board method is statistically better for the initial detection of
mite populations at low infestation levels (p<0.001, Colton, 1974). For example, on the first
sampling date (15 June), the sticky board detected mites in 12 of the 15 colonies while the
adult and brood population detected mites only 2 out of 15 times. When populations grew,
by 15 July, the sticky board detected mites in all 15 colonies, whereas the live adult and brood
population detected mites in 9 of the 15 colonies. Once the population grew even larger by
15 August, both methods detected mites in all the colonies. This continued on 15 September.
Then, on 15 October, the live bees and brood population detected mites in 13 of the 15
colonies, while the sticky board tested positive in all 15 colonies. Although the sticky board
had a higher rate of detection (100%), the two methods were not statistically different fiom
each other on October sampling date (p<0.171). When there was no brood present, there was
no significant difference between sticky board and sampling of live bees.

The next question concerned the difi‘erences between examining the bee brood vs. the
adult bee population Neither one was good at detecting mite populations at low levels. On
all five dates, there was not a significant difi‘erence between the two groups (p=0. 1710,
0.4032, 0.7941, 1.0, 0.1710, respectively).

Correlating the relationship between the live bee samples and sticky board samples,

it was found that all the correlations were positive, except for some correlations with the

26

drone brood. The best correlation was between the sticky board and worker brood (0.79 in
all treatments, Table 1).

Along with the correlations, the data were fit to a series of linear regression models.
It was found that the number of mites recovered on the sticky board could be explained by
the following linear regression (r2=0.656, p<0.001):

TMS = 199.8 + 0.242*AB + 0.682*WB - 9.467*DB
where:

TMS = total mites on sticky board

AB = mites found on adult bees

WB = mites found in worker brood

DB = drone brood
Since the p-value associated with the constant and adult bees and worker brood were almost
significant (p<= 0.053) while the p-value associated with the drone brood was not (p<0.341),
the model was revised to exclude the drone brood from the estimate. When the drone brood
is removed, the r2 is still significant (r2=0.518, p<0.001), and the equation is as follows:

TMS = 282.449 + 0.348*AB + 0.594*WB
where:

TMS = total mites on sticky board

AB = mites found on adult bees

WB = mites found in worker brood
The p-value associated with each of these coefficients was highly significant (p<0.005). The
number of mites found in the worker brood alone was the best single predictor of number of
mites on the sticky board. This factor explained 46% of the variability in sticky board
numbers (p-value < 0.001). Figures 7 and figure 8 show the brood and bee population,

respectively.

27

.333 ES woos 53 2a 82. 35.

.3085 v8.5 07:
.2885 25.6 855 2:0N
.8085 389

 

 

 

 

o8... 3o .82. :2; .> 28.. gag
23 cm: 82 82. 08.? 8.? 82.. 2.2.. .> 83 .33.
~86 8:. $3 89¢ 83 one- 88.. use .> “8.. oz
23 83 so... 83 $8 3.0 .82. is .> ”8.. 2,:
e 8 82s 83 82- 8 _ .o- vac- 25% 5 2:: .> 383 a a:

83 22. 88 23 z .o- 88- 23- 2 .o- 82.. 25:. .> 28.. 3.5
82 mod 83 83 945 as 82. =23 5983 a as.

83 .83 .85 8; . a; an; 23 as .885 .333 .> 28.. .325

83 N8... 83 :56 who one: 38 3o 88 .32. .> .88 oz

82. Sec 83 83 Rd Sec 23 m3 82. 02. .> 28.. ream

82. e: .o :3 82. 93 can... 88 who .83 =2... .> 28.. 32am

:5 i n 2 m :5 a. 8 2 m

 

 

 

 

 

 

 

 

 

 

cog—8:. mafia—am

 

 

A3 3585.»? he .26. 8889.8 05 v5 3 286508 sous—oboe :05 memes—2:

.m_o>o_ nauseous avenge “a maaeeeexgk me .8823 8.. aficion— mEEEam Schema e838: agree—coo

._ use.

28

 

0|
0

 

‘6’ 8

N
O

No.0fm'tes

Thoumnde

 

   

d
O

 

 

 

o l l J l l l

15-May 15-Jun 15¢." 15-Aug 15-Sep 15-Od
Sanpingthte

 

 

 

Figure 7. The average bee population per treatment group (5, 10, 25 mites in initial
inoculation).

 

22

 

20
18
16

14

Noofm'tas

Thousands

12

 

 

1O

 

 

 

6 r J r r
15-Jun 154a! 15-Aug 1.5-Sep
Sanping dda

 

 

 

 

Figure 8. The average total brood population per treatment group (5, 10, 25 mites in
initial inoculation).

29

DISCUSSION

When mite population estimates were based on the natural mite downfall in the hive
debries the data show that the mite population can increase an average of 154-fold during one
summer. Fries et al. (1991b) reported that mite populations could increase more than 100-
fold within one summer in a cold climate very similar to Michigan with beginning inoculation
levels of five to eight mites. When examining the data from the group inoculated with five
mites in the beginning of the experiment it noted that the mite population increased at least
50 times faster than in the Fries et al. study. This difference maybe due to immigration of
mites fiom outside the experimental unit despite precautions taken to avoid such a situation.
This also may explain the large variation between the colonies within the group. Greatti et
al. (1992) found daily reinfestation rates in high density honey bee areas of up to 100
mites/colony. Irndore and Kilchenmann (1991) estimated between 3000 - 4000 mites/colony
were transported by robber bees when they attacked dying colonies in a neighboring apiary
in June.

The estimate from the live bees, adults and broods, yielded a higher and more variable
population level than those obtained fi'om the natural mite fall in the hive debris, and this was
in agreement with Liebig (1996) and Fries et al. (1991a). That probably was due to the
noticeably wide variation in infestation between the colonies in the same group, between the
combs within a colony and between cells within one comb where a few cells would have no
mites and the next would have several mites. This observation was in agreement with others.
Rosenkranz et al. (1984) and Fuchs (1985) reported that infestation within a single colony
varies fi'om one brood comb to another and even from one area of the comb to another. Also

the infestation of the adult bees vary from one comb to another (Pappas and Thrasyvoulou,

30

1986). Liebig (1996) reported live bee estimates are more likely to be affected by the part of
the hive from which the sample has been taken. Ellis and Baxendale (1994) stated that the
distribution of mites among adult bees and brood are probably afl‘ecting the results of the
sampling method. Both Fries er al. (1991) and Liebig (1996) suggested that the natural mite
downfall is more reliable for estimating incidence of Varroa.

Colony-mite population was highest in September and that was possible because
brood production was high in August and relatively high in the first two weeks of September
at the same time the mite population became high. The brood production was highest in June
and July but the mite population was small. The sharp increase from mid August to mid
September also could be a result of immigration, if that had an efl‘ect, although it has been
reported that in isolated apiary after removing the Apistan strips in August, which were in the
hives for 65 days, the mites increased to about 1,700 within 122 days (Delaplane and Hood,
in press).

The sticky board is the most reliable method for detecting mite infestations when the
population is low. In fact, at no time in the study did the live bee and brood estimates surpass
the sticky board as a superior method of detecting mite populations. Detection rates from the
sticky board were either greater or equal to those occurring from the examination of live bee
populations. As the mite population grew, the live bee results became closer to the mite
downfall results. This concurs with Fries et al. (1991a). Therefore, it is recommended that
beekeepers use the sticky board to detect initial populations.

Although the examination of the worker and drone brood was not as good a method
of detecting mites as examining mite downfall from the sticky board, especially at low

population levels, the number of mites found in the worker brood was highly correlated to the

31

mite downfall (r = 0.79). It also was significant in both the multiple regression and the simple
linear regression model (r2 = 0.656 and r2 = 0.460, respectively). This is in agreement with
the findings of Boot et aI.(1995) when he reported that an18% of the mites introduced into
a colony were found in the sticky board when bees emerged from brood cells. The correlation
with drone brood was negative (-0. 14). This was probably due to the small number of drone

cells available for the mites.

LITERATURE CITED

Barbina, MT; De-Paoli,M; Valentino, A (1990): Determination of tau-fluvalinate residues in
honey. Pestic. Sci. 28(2), 197-202.

Boot, WJ; Sisselaar, DJA; Calis, W; Beetsma, J (1994): Factors afi‘ecting invasion of
Varroajacobsoni (Acari: Varroidae) into honeybee, Apis mellifera (Hymenoptera:
Apidae), brood cells. Bull. Entomol. Res. 84, 3-10.

Burgett, DM; Burkikarn,I (1985): Number of adult honey bees Hymenoptera: Apidae)
occupying a comb: A standard for estimating colony populations. J. Econ. Entomol.
78,1154-1156.

De Jong, D (1984): Current knowledge and open questions concerning reproduction in the
honeybee mite Varroa jacobsoni. In: Advances in Invertebrate Reproduction 3:
Proceedings, 3rd International Symposium, International Society of Invertebrate
Reproduction, Tubingen, Federal Republic of Germany, 22-27 August 1983. (Ed:
Engels,W) Elsevier Science, Amsterdam, 547-552.

De Jong, D (1988): Varroajacobsoni does reproduce in worker cells of Apis cerana in South
Korea. Apidol. 19, 241-244.

Delfinado-Baker, M; Houck,MA (1989): Geographic variation in Varroajacobsoni (Acari,
Varroidae): Application of multivariate morphometric techniques. Apidol. 20(4), 345-
358.

Ellis, MD; Baxendale, FP (1994): Comparison of formic acid sampling with other methods
to detect Varroa mites (Vmoajacobsoni Oud.) and mite distribution within colonies
in Nebraska. Bee Sci. 3 (3), 139-144.

32

Engels, W; Rosenkranz, P; Hertl, F; Staemmler, G (1984): [Effect of drone brood removal
on Varroa infested honeybee colonies]. Biologische Varroa Kontrolle durch
Drohnenbrutentnahme. Apidol. 15(3), 246-248.

Fries, I; Aarhus, A; Hansen, H; Korpela, S (1991a): Comparison of diagnostic methods for
detection of low infestation levels of Varroajacobsoni in honey-bee (Apis mellifera)
colonies. Exp. & Appl. Acar. 10(3-4), 279-287.

Fries,I; Aarhus, A; Hansen, H; Korpela, S (1991b): Development of early infestations by the
mite Varroajacobsoni in honey-bee (Apis mellifera) colonies in cold climates. Exp.
& Appl. Acar. 11(2-3), 205-214.

Fuchs, S (1985): [Quantitative diagnosis of the infestation of bees hives by Varroajacobsoni
Oudemans and distribution of the parasitic mite within the hives] Untersuchungen zur
quantitativen Abschaetzung des Befalls von Bienenvoelkem mit Varroa jacobsoni
Oudemans und zur Verteilung des Parasiten in Bienenvolk. Apidol. 16(4), 343-368.

Fuchs, S; Koeniger, N (1984): Rechnen oder Raten-das Dilemmabei der Abschatzung des
Varroabefalles. Allg. Dtsch. Irnkerztg. 18, 294-296.

Greatti, M; Milani, N; Nazzi, F (1992): Reinfestation of an acaricide-treated apiary by Varroa
jacobsoni Oud. Exp. & Appl. Acar. 16(4), 279-286.

Guzman, L; Rinderer, T; Delatte, G; Macchiavelli, R (1996): Vmaajacobsoni Oudemans
tolerance in selected stock of Apis mellifera L. Apidol. 27, 193-210.

Hoopingamer, R (1982): The individual hive robbing screen. Glean. Bee Cult. (110:92,109)

Korpela, S; Aarhus, A; Fries, I; Hansen, H (1992): Varroajacobsoni Oud. in cold climates:
Population growth, winter mortality and influence on the survival of honey bee
colonies. J. Apic. Res 31(3-4), 157-164.

Kraus, B; Page, RE, Jr (1995): Population growth of Varoajacobsoni Oud in Mediterranean
climates of California. Apidol. 26, 149-157.

Kulincevic, JM; Rinderer, TE; Mladjan,VJ; Buco, SM (1992): Five years of bi-directional
genetic selection for honey bees resistant and susceptible to Varroa jacobsoni.
Apidol. 23(5), 443-452.

Liebig, G (1996): (The natural mite fall (in hive debris)- not suitable for rearing work).
Apiculture abstracts 47 (4), 382.

Liebig, G; Schlipfi U; Fremuth, W; Ludwig, W (1984): Ergenbnisse der Unterscuhungen uber
die Befallsentwicldung der Varroa-Mible in Stuttgart-Hohenheirn 1983. Allg. Dtsch.
Imkerztg. 2, 6-11.

33

Lodesani, M; Colombo, M; Spreafico, M (1995): effectiveness of Apistan registered
treatment against the mite Varroa jacobsoni Oud in several districts of Lombardy
(Italy). Apidol. 26(1), 67-72.

Martin-SJ (1994): Ontogenesis of the mite Vmoajacobsom‘ Oud. in worker brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 18(2), 87-
100. ~

Matheson,-A (1993): World bee health report. Bee World 74(4), 176-212.
Matheson,-A ( 1994): World bee health: where to now?. Bee World 75(1), 3-4.
Matheson,-A (1995): World bee health update. Bee World 76(1), 31-39.

Milani, N (1994): Possible presence of fluvalinate-resistant strains of Vmoajacobsoni in
northern Italy. In: New perspectives on Varroa. (United Kingdom) (Ed: Matheson,
A) International Bee Research Association, Cardim 87 .

Moretto, G; Concalves, S; De Jong, D (1991a): Afiicanized bees are more efficient at
removing Vmoajacobsom’ - Preliminary data. Amer. Bee J. 131, 434.

Moretto, G; Goncalves, LS; De Jong, D; Bichuette, M2 (1991): The efl‘ects of climate and
bee race on Varroajacobsoni Oud. infestations in Brazil. Apidol. 22, 197-203.

Otten, C (1991): Reproduction and population dynamics of Varroa jacobsoni Oud. in
colonies of Apis mellifera L. of difi‘erent origin. In: Proceedings of the International
Symposium on Recent Research on Bee Pathology - Gent, Belgium, Sept. 1990.

Pappas, N; Thrasyvoulou, A (1986): Searching for an accurate method to evaluate the degree
of Varroa infestation in honeybee colonies. In: European Research on Varroatosis
Control: proceedings of a meeting of the EC Experts‘ Group, Bad Homburg, 15-17
October 1986. (Ed: Cavalloro,R) Balkema, Broolcfield, VT, 85-92.

Ritter, W (1981): Varroa disease of the honeybee Apis mellifera. Bee World 62, 141-153.

Ritter, W; Leclercq, E; Koch, W (1984): [Observations on bee and Varroa mite populations
in infested honey bee colonies] Observations des populations d' abeilles et de Varroa
dans les colonies a difi‘erents niveaux d' infestation. Apidol. 15(4), 389-399.

Ritter, W (1981): Varroa disease of the honeybee Apis mellifera. Bee World 62, 141-153.

Ritter, W (1984): Neuester Stand der diagnostischen und therapeutischen Massnahmen zur
Bekampfung der Varroatose. Tierarztliche Umschau 39, 122-127.

34

Rogers, L; Gilbert, R; Burgett, M (1983): Sampling mathematical model of brood production
in honeybee colonies. J. Apic. Res 21, 157-160.

Rosenkranz P; Rachinsky, A; Strarnbi, A; Strarnbi, C; Roepstorf, P (1990): Juvenile hormone
titer in capped worker brood of Apis mellifera and reproduction in the bee mite
Varroajacobsoni. Gen. Comp. Endocrinol. 78(2), 189-193.

Rosenkranz, P; Tewarson, NC; Engels, W (1984): Optimal host selection by reproductive
female Vmoajacobsoni. In: Advances in Invertebrate Reproduction 3. Vol. 3. (Ed:
Ritter, W; Van Laere, 0; Jacobs, F; De Wael,L) Apimondia - Intemat. Fed.
Beekeepers Assoc, Rome, 67-69.

Ruttner, F (1983): Varroatosis in honeybees: extent of infestation and efl'ects. In: Varroa
jacobsoni Oud. Afi‘ecting Honey Bees: Present Status and Needs. (Ed: Cavalloro, R)
A. A Balkema, Rotterdam, 7-13.

REPRODUCTIVE BIOLOGY OF VARROA JA COBSONI IN
WORKER AND DRONE BROOD OF THE HONEY BEE APIS

MELLIFERA UNDER MIDWEST CONDITIONS

ABSTRACT

The reproductive biology of the mite Varroajacobsoni Oudemans was studied fi'om
June 30 to October 15, 1995 under Midwestem conditions in Apis mellifera colonies that
were highly infested fi'om the previous year. A total of 353 worker cells containing 697
mother mites and 192 drone cells containing 498 mother mites were found in 959 worker cells
and 344 drone cells that were examined. Number of ofi‘spring were calculated two difi'erent
ways, one included infestations that did not produce ofl‘spring, or produced male only
ofi'spring, and included dead ofi‘spring. The second method that is presented in some of the
literature excluded these infestations. In an effort to compare this study with studies available
in the literature, both methods are presented. It was found that the mean number of female
ofi‘spring reaching maturity before the bee emerged in worker and drone cells containing a
single mother mite are 1.41 and 2.47 ofi‘spring respectively, when non-reproduction and male
only reproduction were included in the average. When these components were excluded the
number increased to 1.82 for workers and 2.69 for drones. In multiple infested cells, the
average number of ofi‘spring was 1.09 for workers and 1.87 for drones, when non-

3S

36

reproduction and male only reproduction were included. These increased to 1.26 and 2.03,
respectively, when non-reproduction and male only reproduction were excluded. This study
found that 86.75 and 93% of the mites were fertile in worker and drone cells, respectively,
when including those mother mites that produced male only offspring. When excluding these
ofi‘spring, the fertility rate decreased to 82 and 90%, respectively, in worker and drone cells.
The percentage of female mites that did not produce eggs are 11 and 7% in worker and drone
cells, respectively. In addition, mortality of mother mites accounted for 2.29 and 2.7% of the

total infested mites in worker and drone cells, respectively.

INTRODUCTION

The mite Varroajacobsoni (Acari: Mesostigrnata) is a parasite of the honey bee (Apis
mellifera) and is currently considered one of the most serious pests of A. mellifera colonies
in most of the world. On its natural host, A. cerana Fabr., the Varroa mite population is
generally under the damage threshold because the mite parasitizes primarily drone brood
(Konger et al., 1981; De Jong, 1988; Tewarson et al., 1992; Rosenkranz er al., 1993) and A.
cerana has developed methods to protect itself such as a more eficient grooming behavior
(Peng et al., 1987) and its ability to remove parasitized worker brood (Rather and Drescher,
1980; Boecking, 1992).

The mite behaves differently in A. mellifera than it does in A. cerana colonies. It
regularly enters both worker and drone brood, but has a higher fertility and fecundity rate in
the drone brood. The number of viable female 063ng produced by invading mother mites

depends, in part, upon the type of cell the mite enters, whether it is drone or worker; and the

37

number of invading mother mites per cell. The number of mites entering a brood cell is
inversely proportional to the number of ofi‘spring produced per female mite.

There is a wide variation in the fertility rate of mites reported among different
countries (Ritter and De Jong, 1984; Ruttner et al., 1984; Ifantidis, 1984; Thrybom and Fries,
1991; Camazine, 1986; Sulimanovic er al., 1981; and Rezenkran, 1994). In addition, there
is also variation in the fecundity rate (Shulz, 1984; Fuchs and Langenbach, 1989; Ifantidis,
1984, 1990; Engels et al., 1986; and Martin, 1994, 1995).

Without control, the mite population can breakdown a colony in Michigan in one to
one and a half years. In Europe the breakdown is much slower, occurring within three to four
years (Rosenkranz and Engles, 1985).

The objective of this study was to record the reproduction, fecundity and fertility of
Varroa under Michigan conditions and compare these findings with results from other

countries.

MATERIALS AND METHODS

Ten colonies of Apis mellrfera were used in this research. The site was located in East
Lansing, Michigan The colonies were highly infested with Varroa fi'om the previous year and
were kept in three story Langstroth hives. Two weeks before the beginning of the study,
attempts were made to strengthen and equalize the colonies used. None of the colonies
examined had previously been exposed to chemical acaricides.

A total of 959 worker and 344 drone cells were examined on four difi‘erent occasions
from June 30 to October 15, 1995. Two frames from each colony were removed on each

sampling day and the contents of at least 30 infested cells were recorded. This procedure was

38

repeated monthly for four consecutive months. Some of the colonies died during the study
and were replaced. In order to estimate the average fecundity of the mites, both worker and
drone brood cells were examined. When possible, only those cells containing adult bees that
were near emerging were used. Ifthis was not feasrble, pupae with dark eyes and light brown
thorax were counted (<230 and 322 hours post-capping for workers and drones,
respectively). Slightly younger pupae were used for the fertility estimation.

The procedures used for examining and recording the data were the same procedures
used by Ifantidis (1990) and are summarized as follows: worker and drone brood cells were
examined in cells that contained emerging adults while they were in the process of emerging,
the number of Varroa female adults, as well as the number of skins of the last molt of young
females, were counted. In addition, the number of adult male(s) and developing mites were
also recorded (to use in case there was a question concerning the number of skins present).
The number of original mother mites found in the cell was determined to be the number of
female adults present minus the number of skins from female deutochrysalis. By dividing the
number of skins by the number of original mother mites one finds the average number of
offspring produced per mite in the cell.

The examined brood combs were not put in the refiigerator before the examination
of their infested cells. Some of the characteristics of reproduction per cycle that were
recorded and analyzed inchrded: (1) total number of ofi‘spring per producing mother mite; (2)
the fiaction of mother mites without 035ng (3) the fiction of mother mites with only male
ofi'spring; and (4) the fiaction of mother mites that were dead in the cells.

Reproduction was calculated using single mite infested brood cells. Some cells

containing old pupae are included in the data. The presence of V. jacobsoni eggs, proto- or

39

deutonymphs, confirmed that the female mites were reproducing. The total number of
ofi‘spring per mother was calculated using two methods. One included mother mites that did
not produce offspring and the other included only mother mites that produced viable
ofi'spring. Both methods are published in the literature and were utilized here because one
of the objectives of this research was to compare mite behavior in Michigan colonies with

data already published.

RESULTS

The total number of worker cells examined was 959, from 20 brood combs. Of these
cells, 353 worker cells were infested with 697 mites. From the four sampling dates, the
infestation rates were 40% in June, 50% in July, 52% in August and 67% in October.
Reproduction rates averaged 86.8% in worker cells when including male only reproduction
and averaged 82.0% when excluding male only (Table 1). The percent of cells that contained
only male ofi‘spring averaged 5.1% in worker cells. The number of offspring per mother mite
averaged 1.41 when including mites that produced only males and dead female mother mites
using single infested cells. The fecundity rate rose to 1.82 when excluding these worker
brood cells from single infested cells (Table 2). The average number of ofi‘spring decreased
to 1.09 when the whole mother mite population was included, regardless of the number of
mother mites infesting the worker brood cell when including male only and dead mother
mites. The average number of ofi‘spring rose to 1.26 when excluding these mites.

There were 192 drone cells fi'om 11 brood combs. These contained 498 mother mites.

Infestation rates for the drone brood were 45% in June, 71% in August and 90% in

September. Reproduction rates averaged 93.0% in drone cells when including male only

40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 2 9: can a: 0%
ed N c.~ _ Q: o v.3 me new me E 3% S:
Om N c.» m S: a odw ow 93 we we 3% (a
ed N 0m v «5 5 ed» me «.3 S we mQNNR
ed c «a m «.2 e 93 3. new we mm 38m?
.x. a e\_. c .x. a {a e .x. a
Ame—«E memes—0x8 988c 8.86
econ 3:0 032 330385 oz 8:338; eeuosvoaox Z .955
88238 came—£2 8pc: m=oo cue—co? 98.332 3&4 E comes—voice aeeaccoexcetek ._ 035.

41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 2. Number of daughter mites produced per Varroajacobsoni mother mites in
worker brood under Michigan conditions.
Date All mother mim (including those Mother mites that produce female
producingonlymalcs anddeadmitm, oflspfingaxcludirgg those producing only
excluding those producing only males, dead mites and matures)
immatures)
Single mated Total mite Single infected Total mite population
cells population cells
n offsprifl n ofl‘spring n ofi'sgrigg n oflgirinL
6/30/95 46 1.46 105 1.04 37 1.81 93 1.11
7/22/95 68 1.47 130 1.16 55 1.82 112 1.31
9/15/95 69 1.41 128 1.22 53 1.83 106 1.38
10/15/95 31 1.29 68 0.93 22 1.81 53 1.25
‘ Aviage 1.41 1.09 1.82 1.26

 

 

42

reproduction and declined to 89.9% when excluding male only reproduction (Table 3). The
percent of cells that contained only male ofi‘spring in drone brood averaged 2.3% in worker
cells. The number of offspring per mother mite averaged 2.47 in single infested cells that
including male only and dead mites and increased to 2.79 ofi'spring when excluding male only
and dead from the average (Table 4). The average number of offspring decreased to 1.87 and
2.03, respectively, when the whole mother mite population was included, regardless of the
number of mother mites infesting the drone brood cell.

Comparing worker and drone fertility and fecundity data, it was found that drone
fertility was higher than worker (93 to 86.8%, respectively). There were more brood cells
containing only males in worker cells as compared to drone cells (5.1 to 2.3%, respectively).
The mite fecundity rate was also higher in drone cells than in worker cells (2.03 to 1.26
average number of ofl‘spn'ng when the whole mite mother mite population was included,
respectively).

The rate of decrease in mite fecundity is illustrated in Figure l for both worker and
drone cells. The graph was generated fi'om the ofi‘spring data that excluded male only, dead
mites, and immatures). The logarithmic equations are as follows:

Y = -l .423 *log(x) + 2.523 for mites infesting worker cells

Y = -1.196*log(x) + 1.486 for mites infesting drone cells
where:

Y = expected ofl‘spring

x = number of mother mites in brood cell

The worker and drone models had r2 values of 0.751 and 0.836, respectively. The

data generating the model demonstrated the negative correlation between number of infesting

43

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.3 an 3 a3 :8 $52
ad ad o he g fine 3 03 S n _ 3% {m
EN ed _ m.m N mam em .Na mm mm maRNR
ed 3. m a.» v :3 3 03 me 3. onmB
.\. .3 a .\. a .x a .\e .—
Ame—«E mafia—0x5
2:0 032 eouosvoiox oz 852.38% 9:80.:— 8.36
8533.5 2 BE
@5238 Semi—£2 cows. «:8 26.6 Embacfi 35x 5 330389.. accuccoexuctc; .m 038.

44

 

 

Table 4. Number of daughter mites produced per Varroajacobsoni mother mites in drone
brood Imder Michigan conditions.
Date All mother mites (including those Mother mites that produce female
producing only males anddeadmites, oflspring (excluding those producing
excluding those producing only only males, dead mites and
immatures) immatures)

Single infested Total mite Single infested Total mite

cells population cells population

 

n oflspfing n offspring n offspring n ofl‘spring

 

 

 

6/30/95 41 2.39 48 2.29 35 2.80 41 2.66
7/22/95 76 2.45 117 1.66 27 2.81 109 1.74
9/15/95 14 2.57 44 1.66 13 2.77 42 1.70

 

 

 

 

 

 

 

 

 

 

 

, Av e 2.47 1.87 2.79 2.03

 

45

N
l

i:
1

  
   

Drone Cells

No. of Offspring
T

Worker Cells

0
b.
l

 

 

Ililr‘l Ill
O—NMVU‘O‘OI‘QO‘

No. of mother mites (excluding male only,
dead mites. and immatures)

Figure 1. Average number of adult females produced by mother mites.

46

mothers and average ofl‘spring per mother. On average, worker cells with one mother mite
produce 1.41 ofi‘spring and cells with four mothers produce 0.62 ofi‘spring in worker cells,
when including male only and dead mothers (Table 5). Ifexcluding these mother mites, the
fecundity rate in worker cells would be higher: 1.82 for one mite infestations, 1.4 for two
mites, 1.1 for three mites, and 0.68 for four mites. In drone cells, the increased fecundity
rates are: 2.8 for one mite infestations, 2.63 for two mites, 2.23 for three mites, and 1.7 for

four mites.

DISCUSSION

There is a wide variation in both fertility and fecundity rates recorded in the literature.
This is at least partly due to the methodology used in the study (especially in fertility rates)
and the definition of fecundity. For example, Fries et al. (1994) demonstrated that the
variation in results from numerous studies was because of the difl‘erent methods of reporting
the data.

Comparing reproduction rates obtained in this study (Tables 1 and 3) with other
studies, it is found that the mite reproduction rate of 86.8% in worker and 93% for drone
broods was very representative of what other researchers found. Blum (1989) reported mite
reproduction rates at 88.7%, Fuchs and Langenbauch (1989) recorded 92.7% and Buchler
(1990) found 86.6% in worker brood cells. In drone cells, Fuchs and Langenbach (1989) also
recorded 92.2% fertility rates in drone cells.

The percent of mites that did not reproduce (11.0% in workers and 7.1% in drones)
also was similar to figures presented by other researchers. In worker cells, Sulimanovic et al.

(1982), Schulz (1984), Moosbeckhofer et al. (1988), Fuchs and Langenbach (1989), Ifantidis

47

 

 

 

 

 

 

 

 

 

 

 

Table 5. Average number of ofl‘spring per mother mite taking number of mother mites per cell
into consideration.
Type of Mites per cell
cell Author
1 2 3 4

Workers 1.50 1.40 1.20 0.96 Moosbeckhofer et al., 1988
1.50 1.58 1.13 0.83 Blum, 1989
1.40 1.09 1.16 0.91 Fuchs and Langenbach, 1989
1.60 1.50 1.70 Martin, 1994
1.41 1.30 1.00 0.62 This study

Drones 2.21 1.90 1.52 1.51 Fuchs and Langenbach, 1989
2.10 1.90 1.60 1.40 Martin, 1995
2.47 2.30 1.99 1.67 This study

A cerana 2.30 1.70 1.20 Rath, 1991

drones

 

 

 

 

 

 

 

 

48

(1990) and Boot et al. (1995) reported infertility rates in worker cells of 13, 16, 7, 7, 14.1,
and 8-12%, respectively. Fuchs and Langenbach (1989), Ifantidis (1984), and Schulz (1984)
found lower levels of infertility in drone cells also. They reported 8, 4, and 5%, respectively.

Some of the mother mites produced only male ofl‘spring. This is probably because they
had not mated, since haploid eggs of Varroa mites develop into males (de Ruijter and Pappas,
1983). Martin (1995) attributes this partly to the death of the male before he is able to
fertilize his sisters. He found in earlier studies, that 20% of the males died before they mated
in worker brood (1994) and 10% in drone cells (1995).

In this study, it was found that 5.1% of the ofi‘spring were male only in worker cells
(Table 1). This was similar to other studies. Boot et al. (1995), Schultz (1994),
Moosbeckhofer et al. (1988) and Fuchs and Langebach (1989) reported a rate of 8-10, 6, 3
and 3%, respectively. For drone cells, a rate of 3% was observed (Table 3). Fuchs and
Langebach (1989) reported a rate of 1%.

The other parameter that was measured was the percent of mother mites that died in
brood cells. It was that 2.3% died in worker cells (Table 1) and 2.7% in drone cells (Table
3). Only one other researcher reported this statistic, Martin (1994). He found a higher
percentage in drones (7.7%), and a similar rate in workers (2%). He found that 32% of the
deaths in drone cells were caused by failure of the mite to emerge fi'om the brood food and
found they were trapped in the cell wall. The percentage rose to 50% in worker cells.

Fecunditywas measured in avariety ofways inthe literature both because ofdifi'erent
experimental techniques used and because of differing definitions of fecundity. For example,
there were difi‘erences in the way that researchers extrapolated estimates where cells had to

be opened before development was completed (Boots et al., 1995). In addition, some studies

49

included all mothers that infested brood cells, which included those that died, did not
reproduce, and reproduced males only. These methods, in general, had less offspring per
mother than those that excluded these groups from the study. This study compiled fecundity
rates utilizing both methods in single infested cells and multiple infested cells calculated
separately (Tables 2 and 4). Varroa females are biologically capable of laying up to six eggs
in worker cells and seven in drone cells (Ifantidis, 1984; Martin, 1994, 1995). The fact that
fecundity rates are much lower than this is attributed to offspring mortality, which occurs
primarily in the deutonymph stage (Martin, 1994, 1995).

When excluding problematic mothers from the analysis, it was found that mites
averaged 1.82 053ng in single cell infestations of worker brood (Table 2). In drone cells,
this figure increased to 2.79 ofi‘spring (Table 4). This is fairly consistent with other
researchers. Fuchs and Schulz (1984), Ifantidis (1984), Fuchs and Langebach (1989), and
Martin (1994, 1995) reported 1.82 and 2.69, 1.33 and 2.77, 1.69 and 2.76, and 1.45 and 2.2
in worker and drone cells, respectively.

When including all the female mites found in single infested cells these figures
decreased. In Michigan, average ofi‘spring for worker brood decreased to 1.41 (Table 2) and
to 2.47 for drones (Table 4). These statistics differed widely among other researchers.
Schulz (1984) found 1.3 in worker cells and 2.6 in drones. Ifantidis (1984) reported 0.71 in
worker and 1.7 in drone cells. Fuchs and Langebach (1989) found 1.4 in worker and 2.21
in drone.

When the whole mother mite population is taken into consideration (single and
multiple infested cells), the average number of ofi‘spring decreases for both calculation

methods (Tables 2 and 4). The negative correlation of average ofi‘spring per mite and

50
number of mother mites infesting the cell is well documented in the literature and is supported
by the study reported here (Figure 1, Table 5). The Michigan data includes both dead
mothers and only male reproduction.

Fuchs and Langenbach (1989) attribute this lower fecundity rate to suppressed
reproduction while Martin (1995) shows that it is due largely to increased mortality of the
offspring. Donze and Guerin (1994) showed that mites normally feed at a single feeding site.
Martin (1995) explained that with increased number of eggs, there is more competition for

the feeding site, and stronger deutonymphs and adults will out compete younger ofl‘spring.

LITERATURE CITED

Blum, R (1989): Einfluss einer Unterschiedlichen proteinernahrung von Honigbienen aufdie
Reproduktion der hamophagen Milbe Varroa jacobsoni. Diplornarbeit Thesis, Fakultat
Biologie, Universitat Tubingen, Germany. 51 pp.

Boecking, O (1992): Removal behavior of Apis mellifera colonies towards sealed brood
cells infested with Varroajacobsoni: techniques, extent and eficacy. Apidol. 23,
371-3 73.

Boecking, O; Drescher, W (1992): The removal response of Apis mellifera L. colonies to
brood in wax and plastic cells after artificial and natural infestation with Varroa
jacobsoni Oud. and to fi'eeze-killed brood. Exp. & Appl. Acar. 16(4), 321-329.

Boot, WJ; Schoenmaker, J; Calis, JNM; Beetsma, J (1995): Invasion of Varroajacobsoni
into drone brood cells of the honey bee, Apis mellifera. Apidol. 26(2), 109-118.

Buchler, K Drescher, W (1990): Variance and hertiability of the capped developmental stage
in European Apis mellifera L. and its correlation with increased Varroa jacobsoni
Oud. infestation. J. Apic. Res 29(3), 172-176.

Camazine, S (1986): Differential reproduction of the mite, Varroajacobsoni
Mesostigmata: Varroidae), on Africanized and European honey bees
(Hymenoptera: Apidae). Ann. Entomol. Soc. Amer. 79(5), 801-803.

Donze, G; Guerin, PM (1994): Behavioral attributes and parental care of Varroa mites
parasitizing honeybee brood. Behav. Ecol. Sociobiol. 34(5), 305-319.

51

Engels, W; Goncalves, LS; Steiner, J; Buriolla, AH; CavichioIssa, MR (1986): Varroa
infestation in Carnolian honey bee colonies under tr0pical climate. Varroa Befall
von Carnica Voelkem in Tropenklima. Apidol. 17(3), 203-216.

Fries, I; Aarhaus, A; Hansen, H; Korpella, S (1991): Comparison of diagnostic methods for
detection of low infestation levels of Varroajacobsoni in honey-bee (Apis mellifera)
colonies. Exptl. & Appl. Acar. 10, 279—287.

Fuchs, S; Langenbach, K (1989): Multiple infestation of Apis mellifera L. brood cells and
reproduction in Varroajacobsoni Oud. Apidol. 20, 257-266.

Ifantidis, MD (1984): Parameters of the population dynamics of the Varroa mite on
honeybees. J. Apic. Res. 23(4), 227-233.

Koeniger, N; Koeniger, G; Wijayagunasekaran, H (1981): Beobachtungen uber die
Anpassung von Varroa jacobsoni an ihren ursprunglichen Wirt Apis cerana in Sri
Lanka. Apidol. 12 (1),37-40.

Martin, SJ (1994): Ontogenesis of the mite Vmoajacobsoni Qud. in worker brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 18(2), 87-
100.

Martin, SJ (1995): Ontogenesis of the mite Vmoajacobsoni Oud. in drone brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 19(4), 199-
210.

Martin, SJ (1995): Reproduction of Varroajacobsom‘ in cells of Apis mellifera containing
one or more mother mites and the distribution of these cells. J. Apic. Res 34, 187-
196.

Moosbeckhofer, R; Fabsicz, M; Kohlich, A (1988): Untersuchungen gber die Abhngigkeit der

Ritter, W; De Jong, D (1984): Reproduction of Varroajacobsoni O. in Europe, the Middle
East and tropical South America. Z. Angewandte Entomol. (J. Appl. Entomol.) 98(1),
55-57.

Rosenkranz, P; Engels, W (1994): Infertility of Vmoajacobsom’ females afier invasion
into Apis mellifera worker brood as a tolerance factor against varroatosis. Apidol.
25(4), 402-411.

Ruttner, FH; Marx G (1984): [Observation about a possible adaptation of Varroa jacobsoni
to Apis mellifera L. in Uruguay] Beobachtung neber eine moegliche Anpassung von
Vmoajacobsoni an Apis mellifera L. in Uruguay. Apidol. 15(1), 43-62.

Schulz, AE (1984): Reproduction and population dynamics of the parasitic mite Varroa
Jacobsom‘ Oud. and its dependence on the brood cycle of its host Apis Mellifera.
Apidol. 15 (4), 401-420.

Sulimanovic, D; Ruttner, F; Pechhacker, H (1982): Studies on the biology of reproduction
in Vmoajacobsoni. Honeybee Sci. 3, 109-112.

52

Thrybom, B; Fries,1 (1991): Development of infestations by Varroajacobsoni in hybrid
colonies of Apis mellifera monticola and Apis mellifera ligustica. J. Apic. Res
30(3-4), 151-155.

MODELING OF HONEY BEE AND MITE

POPULATION DYNAMICS

ABSTRACT

The life history of the honey bee Apis mellifera and the bee mite Varroa jacobsoni
and their interactions were modeled using a commercial software package known as Stella
IL Stella II provides a non-mathematically intensive modeling technique that allows the user
to generate a series of differential equations that can track a population through time with a
minimum set of parameters. The model generates population statistics at regular intervals
throughout a designated time period and outputs diagrams, a series of equations, tables and
graphs.

Mite parameters included in the model are: initial mite population, number of female
offspring per mother mite in both worker and drone brood, number of reproductive cycles,
fertility, mite preference for drone versus worker brood, phoretic period, mortality, and
removal of infested brood cells.

Interaction between the bee and mite population were also modeled. For example, the
impact of various mite infestation levels on honey bee colony dynamics was modeled using
three different mite initial infestations rates (5, 10 and 20 mites). The first year infestation
rate resulted in different maximum mite populations levels, and the actual estimate was 690,
1,339 and 2,521 for the three rates, respectively. In the second year, the mite population

53

54

maximums grew to 11,000, 17,000 and 25,000, respectively. However, the colony collapse
was inversely proportional to the initial infestation rate. For example, the colony with an
initial rate of 20 mites collapsed first.

It is also possible to change other parameters of the model such as post-capping
period, fertility and fecundity rate, or to introduce chemical and biological control
mechanisms. These simulations may help beekeepers alter their control strategies for the
mites.

The objective of this modeling project is to develop a tool that will both predict bee
and mite populations under specific conditions and allow a scientist to vary and check the
influence of single factors on the mite population development. In addition, it will help in
the development of research hypotheses to test under field conditions. It also was the hope
of this modeling project to integrate past research results and identify research gaps that need
further investigation in an effort to find better ways to control mite populations in bee

colonies.

INTRODUCTION

Modeling is a valuable tool that helps in the understanding of the complex societies,
such as honey bees, by incorporating the many factors that compose or influence the colony
population dynamics.

There have been several models developed that describe honey bee colony population
dynamics. Degrandi-Hoffman et al. (1989) developed a model that simulates honey bee
(Apis mellifera L.) colony population dynamics. Her model was based on the egg laying

potential of the queen, foraging characteristics of the colony, degree days, and amount of

55

sunlight. Harris (1985) developed a model that predicts adult honey bee population
dynamics based upon field estimates of scaled brood or daily egg laying rates, survival rates
of the immature and adult bees, developmental rates of eggs, larvae, and pupae, and the
initial size of the adult population. McLellan et al. (1980) and Rowland and Mclellan (1982)
developed models to predict brood production throughout the year based upon algorithms
describing the increased and decreased egg laying potential.

In addition, there have been models developed for the Varroa mite. Camazine (1988)
modeled some factors affecting severity of Varroa on European and Africanized bees.
Omholt and Crailsheim (1991) developed a model to predict the degree of infestation of
honey bee colonies estimated by a mean of their natural death rates. Fuchs (1992)
demonstrated in model simulations the importance of choice between worker and drone
brood cells at varying ratios of the brood types for optimal reproductive success of the mite.
Fries et al., (1994) developed a comprehensive model based on both bee and mite population
dynamics. However, their bee p0pulation dynamics were static and did not change with mite
interactions. Marcangeli et al. (1995) used Camazine’s model to analyze the population
growth of V. jacobsoni in the temperate climate in Argentina

The population dynamics of a honey bee colony is dependent on many parameters
that interact with each other. These include oviposition, colony population, weather and
brood availability. The egg laying has the most dramatic effect on colony population

dynamics (DeGrandi-Hoffman et al., 1989; Mclellan et al., 1980).

Egg laying potential depends on sunlight, degree days, and the foraging population.

Under Midwestern conditions temperature ranges for egg laying is between 0-31C°, and the

56

photoperiod between 9.1 to 15.25 h. of light per day (DeGrandi-Hoffman et al., 1989). Nolan
( 1925, 1928) found that egg-laying rates of less than 2,000 per day over a 12-day period were

to be expected.

According to Fukuda & Sakagami’s (1968) life table, 95.8% of the eggs hatch and
become unsealed brood. After five days, 85.7% of the unsealed brood become sealed brood.
Twelve days later, 98.8% of the sealed brood emerged as adults. During their twenty one

days as house bees in the hive, 5.54% die.

It takes approximately 21 days for eggs to become adult workers and 24 days for eggs
to become adult drones (Jay, 1963). It is estimated that the egg and larvae mortality rate is
approximately 19% (Fukuda & Sakagrni, 1968). The natural adult bee life span depends on
the season (Ribbands, 1953; Fukuda & Sakagami, 1968). The average longevity of June bees
is 28.3 days, July bees 32.4 days, wintering bees 154.1 days and postwintering 23.4 (Fukuda
and Sakagami, 1968). Under optimal colony conditions, the adult drone life span was
estimated to be 59 days (Howell and Usinger, 1933). It is normal to expect a 50% decline in

honey bee population during the winter (Avitabile, 1978).

Under Midwest conditions, colony p0pulations peak at most at 50,000 adults in the
middle of July while drone populations tend to peak in the middle of June. Bees do not
produce brood from late autumn to midwinter (late January) and are confined to the hive (I.

e., do not forage) from late October to early April (DeGrandi-Hoffman et al., 1989).

All factors that increase the brood activities of the bees have a strong effect in the
development of the mite populations because the growth of the mites population is closely

correlated with the availability and type of brood.

57

In addition, mite population is effected by factors such as fertility and fecundity of
the mite, reproductive cycle, phoretic period and mortality either natural or caused by the

bees as result of hygienic or grooming behavior of the bees.

Koeniger et al., (1981) and Anderson (1994) reported that V. jacobsoni does not
normally reproduce when it infests worker brood of the Eastern honey bee, A. cerana.
Extensive mite reproduction occurs only on drone brood. In Apis mellifera, a portion of the
infesting Varroa mites do not reproduce. Blum (1989) reported mite reproduction rates at
88.7%, Fuchs and Langenbauch (1989) recorded 92.7% and Buchler (1990) found 86.6% in
worker brood cells. In worker cells, Sulimanovic et al. (1982), Schulz (1984),
Moosbeckhofer et al. (1988), Ifantidis (1990) and Boot et al. (1995) reported infertility rates
in worker cells of 13, 16, 7, 4.1, and 8-12%, respectively. Fuchs and Langenbach (1989) ,
Ifantidis (1984), and Schulz also (1984) found lower levels of infertility in drone cells. They
reported 8, 4, and 5%, respectively. Ritter and De Jong (1984) observed only 43% of the
mites in A. m. ligustica in the worker cells in South America to be fertile. Marcangeli et al.,
(1992) estimated that, depending on the season, between 56% and 72% of the mites in A. m.
ligustica colonies were fertile. Rosenkranz and Engels (1994) compared Africanized and
Eur0pean colonies of A. mellifera, and found less than 40% of female mites were fertile in
Africanized bees, whereas, in European bees between 80-90% were fertile.

Ruttner (1984) reported colonies in Uruguay which could resist Varroa infestation
without any treatment; this was attributed to the very low fertility (IO-30%) of the mite in
worker brood cells. In Tunisia, Ritter (1990) reported the comparatively low number of
fertile mites produced (SO-80%). This demonstrated the increased tolerance of Am.

intennissa to Varroa.

58

The number of offspring produced by Varroa females is also an important factor
effecting the mite population. The number of offspring depends on the type of brood, that
they are produced on, whether it is worker or drone cells. Schulz (1984) measured offspring
reproduction by fertile female mite in worker brood as 1.8 (including infertile mites 1.6). For
the drone the reproductive factor was 2.7. Fuchs and Schultz (1984), Ifantidis (1984), Fuchs
and Langebach (1989), and Martin (1994, 1995) reported 1.82 and 2.69, 1.33 and 2.77 , 1.69
and 2.76, and 1.45 and 2.2 in worker and drone cells, respectively.

The number of times a mother mite enters a brood cell to reproduce is a key factor
in determining the population growth of V. jacobsoni (Fries et al., 1994). Ruijter (1987)
artificially transferred mites from cell to cell and found that the mother mite is able to
reproduce as many as seven times. Schulz (1984) reported that 78% of the mites reproduce
only once and 22% reproduce twice. Mikityuk et al. ( 1976) observed an additional
reproduction cycle. He stated that 78% of the Varroa produce only once, 18% produce twice
and 4% produce three times for an overall mean number of 1.26 reproductive cycles per
female mite. Mikityuk (1979) stated that 1.9% of the mites reproduce four times. Fries and
Rosenkranz (1993) reported that 13% of the mites reproduce three times, their sample size
was 475 mother mites. Wended and Rosenkranz (1993) found that 4.1% of the mites produce
three times with an overall mean of 0.88 reproductive cycles per mother mite. Recently, Fries
and Rosenkranz (1996) using full-size colonies reported that under optimal conditions the
mean number of reproductive cycles by Varroa is greater than 1.5 but less than 2.

After emergence from the brood cells, the female mites reside a certain period on
adult bees in the colony before they invade new brood cells ( Boot et al., 1993; Boot et al.,

1994). Boot et al., (1995) stated that the length of this period strongly affects the population

59

dynamics of the mite, because mites cannot reproduce while they reside on adult bees and
therefore reproduction is delayed. In addition, the period on adult bees may affect the
population dynamics of the mites since some of the mites will die during their stay on adult
bees and it might reduce the number of offspring per mother mites (Beetsma and Zonneveld,
1992). Thus, mite fitness increases by minimizing their stay on adult bees (Boot er al. 1993).
Schulz (1984) reported after a phase of 1-20 days (44% within 6 days) on adult honey bees,
the female mite enters the brood cell for reproduction. He found that the phoretic period was
4.5 days for older mites, 10.7 days for younger mites and averaged 7.4 days for a mixed
population of mites. Boot et al. (1993, 1994) stated that during the brood rearing the mean
residence time of mites on adult bees is maximally 1-3 weeks, depending on the number of
brood cells available for mite invasion. Woyke (1987) reported that mites have an average
phoretic period of 4.7 days in a mixed population and 5.9 days for younger mites. Grobov

(1977) reported a range of 4-13 days.

Mites begin to invade brood cells during a limited period preceding cell capping with
a fairly constant rate until cells are capped, about 50 and 20 hours for drone and worker,

respectively ( Ifantidis 1988 and Boot et al. , 1992).

Muller (1987) stated that the Varroa mite does not change hosts if the host drops from
the winter cluster. A 50% reduction in the number of bees is normal during wintering in cold
climates (Avitabile, 1978), and a similar effect on the population of Varroa could be
expected (Fries et al., 1991). Muller (1987) reported losses of between 3 and 38 mites per
day throughout the European winter. However, other authors suggested different percentages,

3-10% (Weiss, 1984; Rademacher and Geiseler, 1986); Moosbeckhofer (1991) reported

60

between 3% and 40%; Korpela et al. (1992) estimated a total mortality of 40% over a
broodless period of 125 days during the winter, which corresponds to a mean mortality of
0.4% per day. Boot et al. (1995) estimated 0.6% per day during the brood rearing periods and
stated that there is no reason to think that mortality of mites on adult bees should depend

much on the time of the year.

Fries et al. (1994) stated that one important source of summer mortality of phoretic
mites is the loss of mites on foragers that fail to return to the hives. A colony whose
population is in steady state will have approximately 1,500 adult bees eclosing and dying
each day, or about 5% of the population. Mite mortality in the sealed brood cell is

approximately 1.5% of the mother mites (Kustermann, 1990).

Hygienic and grooming behavior also has an effect on mite populations. In the Asiatic
honeybees A. cerana F., the original host of V. jacobsoni, the infestation remains at low
levels and the parasite does not severely harm the colony. That may be due to the number of
defense mechanisms that this bee has. One of these mechanisms, as described by Peng et al.
(1987), is the active removal of adult mites from the bodies of worker bees. This process
involves self—cleaning behavior. After showing signs of irritation, the bee performs a
grooming dance, and then nestrnate cleaning and group cleaning behavior. This resulted in
removing (within two hours) more than 99% of mites added to the colony. Only 0.3% of the
mites were removed by grooming in colonies of A. mellifera. Buchler et al. (1992) also
compared grooming in A. cerana and A. mellifera and found successful mite were removal
in 75% of the cases in A. cerana. In A. mellifera, 48% of the mites removed by grooming.

Fries et al. (1996) reported lower numbers in full-sized colonies of A. cerana, 56% of 220

61

mites were removed by the bees in 6h. and, of those, 30% were damaged; results for A.

mellifera colonies were 21% of 280 mites were removed and 12% were damaged.

Ruttner and Hanel (1992) examined the natural mortality of five A. m. carnica
colonies for about one year and found on average 26% of the mites collected from inserts
showed injuries to the legs but rarely to the cuticle of the idiosoma. Moretto et al. (1991)
reported that 5.75% of the mites were removed by A. m. ligustica bees within 30 min after
infestation, and an average of 38.5 (range 10-70%) were removed by Africanized hybrids of
A. mellifera bees.

Hygienic behavior of bees was described by Rothenbuhler (1964) in relation to
resistance against American foulbrood (Bacillus larvae). He showed that hygienic behavior
consists of two independent behavioral events: the uncapping of cells containing larvae or
pupae, and the removal of the dead brood. The hygienic behavior of A. mellifera against
Varroa mites was observed in both A. cerana and A. mellifera. Rath and Drescher (1990)
showed that the A. Cerana workers were success at detecting, uncapping and removing 98%
of artificially infested worker brood cells within 5 days. Boecking and Drescher (1990)
reported that artificially infested worker brood cells were detected, uncapped and removed
to various degrees and they show that brood cells infested with one Varroa mite were rejected
from 14.3 up to 95.8%, those with two Varroa from 25 up to 100% after ten days. Boecking
and Drescher (1991) reported that the removal of brood cells infested with one mite in Am.
carnica was 5.5% (minimum) up to 95.8 (maximum). Within the same colonies; brood cells
infested with 2 Varroa mites showed a removal from 4.8% (minimum) to 100.0%

(maximum). In another study Boecking and Drescher (1994) stated that mites more

62

effectively removed infested brood from their cells. When one mite was in a cell, the removal
rate are 10.9% for their own brood and 15.4 for another brood, when number of mites
increased to two, removal rates rose to 32.2% and 41.9%, respectively. Boecking and Ritter
(1993) reported workers in 15 test A. m intennissa colonies detected and removed up to 75%
of artificially infested brood and removed up to 97-99% of freeze-killed brood in each of two

trials.

MATERIALS AND METHODS

The bee and mite population dynamics were modeled using a modeling package
known as Stella 11 (High Performance Systems). Stella provides a non—mathematically
intensive modeling tool that automatically computes mathematical equations. The program
presents the model in a variety of ways, such as through diagrams, a series of equations,
tables and graphs that represent model inputs and outputs that are simulated through a
specific time. The model was constructed using literature values (T able 1) to simulate honey
bees, Apis mellifera and Varroa jacobsoni population dynamics.
Model description

This biological model has two major components, one focusing on the honey bee and
the other on the mite (Figures 1 and 2). It simulates five years of colony development with
1800 time intervals (one per day). The initial model assumes that the bee population begins
with 15,000 adults and the mite is introduced with an initial population of 10 mites. The
model is designed so the user can change these initial parameters.

W. The honey bee component of the model is largely driven by the

number of eggs laid per day (Figure 3). The eggs laid per day equation is first based on the

 

 

 

 

 

 

 

 

Worker Brood
W Egg Flow
Drone Brood
D Egg Flow
= Mite population
0 Flow I

Figure 1. Honey bee component of the modelling project.

65

.865 95.25.: on. Co E8858 2:2 .N 2:9...

0? i

3:888 .8332

32m 0 3:25 8:5.

8%

noem >> 3:25 852 =

 

£23.: 3323 8m

32m 0 so: «was.

01

 

 

 

 

 

 

 

QAflPL

8.3 Seat

 

ll"

8.2m 2, ea. 8.3

 

 

63

Table 1. Parameters used for model simulation of Varroa jacobsoni and honey

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

bees Apis mellifera.
Systems Parameters Honey Bee Mite
Initial Population 10,000 10
Length of simulation 5 years Sycars
Length of time interval 1 day 1 day
Starting date January 1 January 1
Maximum eggs laid per day 1500 us:
Brood Mortality 0.15 (Sakagami and Fukuda, 1968) ***
Hours of sunlight Michigg average accumulated sunlight) ***
DD 50 year average of acc. temperature with *“
base of 0°C (Published database on
max/min temperatures on CD Rom)
Foraging Percent foragers under Midwestern ***
conditions (DeGrandi-l-loffman, 1989)
Adult factor Coefficient of adult factor taken from ***
Midwes research data (DeGrandi-
l-loffman, 1989)
Worker postcapping period 12 days ***
Drone postcapping period 14 days ***
Total worker brood Conveyer stock, with 12 day interval ***
Total drone brood Conveyer stock, with 14 day interval *"
Worker proportion Graph function, Table l (Nolan, ***
1925,1928)
Natural mortality Graph function (Sakagami and Fukuda Graph function (.004 for
.1968) winter, .006 for summer)
(Fries er al., 1994; Boot er al.,
1996)
Bee Mortality due to mite Dependent on no. of mites per pupa
invasion (Beetsma, 1983)
Mite Mortality due to bee Mites on infested bees will die
mortality with bee, additional mortality
is equal to the number of mites
on bees that die (Muller,
1987), see equation 8.
Phoretic Period *** 5.9 days (Woyke, 1987)
Mite preference for drones *** Graph function (Fuchs.
(fraction to drones) 1990)
Number of offspring per mite Graph function, Table 4
Mite fertility on worker brood *** 0.85 (Schulz, 1984)

 

Mite fertility on drone brood

tit

0.95 (Schulz, 1984)

 

Number of reproductive cycle

tit

1.4

 

 

Number of female offspring
produced in a worker cell
Number of female offspring

produced in a drone cell

 

 

1.3 (Ifantidis, 1984)

2.7 (Schulz, 1984; Ifantidis,
1984)

 

 

66

e m...
Co tant o

.. dult factor -
Sunlight factor . O ..

r
all da
99 y Queen' ‘28

J

Sunlight . .

Initial Popul ion factor

.. dd factor ".
Q ' foraging factor .

. Initial Pop

Foraging 96 Bees

dd

Figure 3. Honey bee egg laying components of the modelling project.

66

ll . Adult pop
Co . ant o
.. ‘dult factor

Sunlight factor . .

I
l i .~- d
, ggs ' r ay Queen' ‘ee

Sunlight . Q

.. Initial Popul ion factor
dd factor .

Q ‘ foraging factor .
dd
0 Initial Pop

Foraging 96 Bees

 

Figure 3. Honey bee egg laying components of the modelling project.

67

assumption that the adult population must have at least 1,000 members to support the
queen ' 5 egg laying (If Adult population > 1000 then use the regression line to estimate
number of eggs laid, otherwise enter 0). If the population is greater than 1,000 the regression
equation is the primary factor determining egg laying potential unless it returns a negative
value. If it returns a negative value then, once again 0 is entered as the number of eggs laid
that day. The regression equation was based on four factors and a constant. These factors
included: sunlight, degree days, foraging and number of adults in the colony. These factors
explained 95.5% of the variability in egg laying (p<0.001).

E = -1303 - 0.0154A + 188.7038 +0.859D + 5.525F Eq. 1
where:

E = Maximum number of eggs laid per day

A = Number of adults in bee colony

S = Hours of sunlight

D = DD accumulation incremented per day

F = Percent of foragers in bee colony

If the regression equation is positive, that number of eggs is multiplied by two factors,
one corresponding to the age of queen and the other to the size of the initial adult population.
The queen ' 5 age coefficient ranges from 0.9 to 1.0, depending on the age of the queen.
Older queens do not lay as many eggs. In addition to queen’s age, it is believed that number
of worker bees have a positive correlation to egg laying. If the initial population is low, the
number of eggs laid does not meet the queen’s egg laying potential. However, if the
population is 15,000 or more, then the queen lays eggs close to her potential. The initial

population factor ranged from 0.75 to 1.25, depending on the size of the colony (ranging

from 1000 to 30,000).

68

Eggs develope to worker or drone brood. The pr0portion that became worker brood
was multiplied by the number of available eggs. The rest (one minus the proportion
becoming worker brood * available eggs) develop to drone brood. The proportion that
develops to worker and drone brood was dependent on time of year and taken from Michigan
specific research papers.

A delay function was used to determine the number of eggs going from the egg stage
to the brood stage for both drone and workers. This function allowed a three- day delay,
which is the biological development time for the egg. More specifically, if the queen laid the
egg on day 1, it would pass through the egg phase on day 4. The number of eggs available
at any given time was accumulated in separate stocks for worker and drone eggs. The
equations are as follows:

dED(t) = ED(t-dt) + (ED(t)-FD,(t))dt Eq. 2a

dEw(t) = Ew(t-dt) + (Ew(t) - Fm(t))dt Eq. 2b
where:

dED = Change in drone eggs

FDI = Flow from drone eggs

dEW = Change in worker eggs

FWI = Flow from worker eggs
This number was then multiplied by an egg mortality factor before it entered the brood stock,
which once again was accumulated. The mortality rate was 0.15 for workers and 0.35 for
drones. The brood accumulated as follows:

dBD(t) = BD(t-dt) + (0.65*FD,(t) - FDn(t))dt Eq. 3a

dBw(t) = Bw(t-dt) + (0.85*Fw,(t) - Fwn(t))dt Eq. 3b

69

where:

dBD = Change in drone brood

BD = Drone brood

FDI = Eggs becoming drone brood

FD“ = Flow from drone brood

dBW = Change in worker brood

BW = Worker brood

FWI = Eggs becoming worker brood

PM = Flow from worker brood

The developmental period for larvae was 21 days for drones and 18 days for workers.
Therefore, the total number of drone brood is equal to the number of brood already present,
plus the number entering that specific day from the eggs laid three days previously, minus
the number of brood becoming adult bees, either those entering the brood 21 days before
(drones) or 18 days for the workers.

Total honey bee population was determined by the number of honey bees already
present, plus the honey bees coming from brood cells, minus honey bees lost due to mortality
factors. The mathematical equation for bee population growth is:
where:

BP = Bee Population

Fwn = Flow from worker brood

Fm, = Flow from drone brood

M3,. = Bee Population Mortality

Honey bee mortality came from either death due to natural causes or death due to
varroaosis. Honey bees typically survive 32 days in the summer, and 154 days in the winter.

If the bee had one mite enter its pupa cell, the bee’s life span is reduced by one-third; and if

two or more mites entered its brood stage, its life span is reduced further (by about two-

7O

thirds). These mortality rates were also taken from the literature. Thus, the equation for bee
mortality is as follows:

M3,, = (BP - MPDC - MPWC)*MB + (MPDC + Mch)*2*MB Eq. 5
where:

MBP = Mortality of bee population

BP = Bee Population

MPDC = Mite pOpulation entering drone cells

MPWC = Mite population entering worker cells

ME = Mortality rate for bee

Mitesuhmgdel. Once a bee colony was established, the model assumed that mites
were introduced into the system. In many ways, the mite population dynamics is much more
complicated than the honey bee. It depends on both the bee population and the density of the
mites in relation to the bee population (Figure 4).

The mite population dynamics begins with the number of phoretic mites and their
density compared to the number of available worker and drone brood cells. The number of
phoretic mites is a function of the number of mites already present, plus the mothers and
offspring exiting worker and drone brood, minus mite mortality, minus mites entering into
worker and drone brood that are available for reproduction. The number of phoretic mites
in a colony at any given time is represented by the following equation:

dMP(t) = MP(t-dt) + (MPED + MPEW ' MMP ' MPDC ‘ Mch)dt Eq. 6
where:

dMP = Change in mite population

MP = Mite population

MPED = Mite population exiting drone cells

MP1;w = Mite population exiting worker cells

MMP = Mortality of phoretic mites

MPDC = Mite population entering drone cells
MPWC = Mite pOpulation entering worker cells

71

-I

New 0‘ onngWMax

Pc ime D

Pc influence I

fertility o . . Pc influence w

Upcapping factor

 

fertility w

Figure 4. Mite population components of the modelling project.

72

When mites are ready to reproduce offspring, they enter brood cells, with a preference for
drone brood cells. The number that enter each brood cell is very important because it
determines the number of offspring that each female will produce. The actual number of
offspring is determined by the fertility rate of the mite, the number of female mites available
to produce eggs, the number of available worker and drone brood cells, and the length of the
post-capping period.

The mites have a higher fertility rate in drone cells (0.95) than in worker cells (0.85)

and can also produce more offspring in drone cells (up to 2.7) than in worker cells (up to

1.3).
The equations for offspring produced in worker and drone brood is as follows:
Own = Fw*DDw*Iw*UF Eq. 7a
ODB = FD*DDD*ID*UF Eq. 7b

where:

Own = Offspring produced in worker brood
FW = Fertility rate in worker cells

DDW = Density dependence in worker cells
IW = Post-capping influence in worker brood
UF = Uncapping factor

ODB = Offspring produced in drone brood
FD = Fertility rate in drone cells

DDD = Density dependence in drone cells

ID = Post-capping influence in drone brood

Density dependence for mite offspring in worker and drone brood depends on the
number of mites entering each cell. This is dependent on both the number of mites and the
number of cells available. The number of available cells is determined by the total number
of cells that are of the correct age to attract the mite. This part of age corresponds to the

eighth and ninth day after egg laying plus an additional two hours on the tenth day for drone

73

brood, and to 20 hours of the eighth day for worker brood. Delay functions were used to
simulate this period of attraction. Total number of available drone brood cells was equal to
the number of eggs laid eight days previously plus the number of eggs laid nine days
previously plus one-twelfth times the number of eggs laid ten days ago, while, total number
of available worker brood cells was equal to the number of eggs laid eight day ago,
multiplied by 20/24 as the attractive period is equal to 20 hours.

Then the number of mites available to enter the cells was divided by the number of
brood cells available to determine the average number of mites per cell. This amount
determined the number of offspring each mite would have. This density dependent offspring
function ranged from 0.0 to 3.0 for drone brood cells, with number of offspring falling from
3.0 to 0.0 when number of mites per cell was greater than 6.0. The range was less for worker
cells, extending from 0.0 to 1.3, with the number of offspring falling to 0.0 when number of
mites per cell was greater than 3.0.

Post-capping period was dependent on type of brood cell. Worker brood had a period
of 12 days and drone brood 14 days. It is assumed that there is a small amount of variation
in the actual Post-capping time, and if the time is somewhat shorter less mite offspring will
be produced. Post-capping influence is a function that ranges from 0.5 to 1.5 which reflects
the influence of the variability of the Post-capping period on reproduction. The model
assumes that the Post-capping period is a random function, and is normally distributed with
a mean of 12 and a standard deviation of 0.1 for workers; and a mean of 14 and a standard

deviation of 0.1 for drones.

74

Total number of mite offspring is multiplied by 0.95 due to an uncapping factor.
Adult worker bees will destroy brood cells if they know that the mites are present and this
accounts for a mortality rate of approximately 0.05.

The number of offspring is multiplied by 1.4 which is the average number of
reproductive cycles per female mite to obtain the number of mites leaving drone and worker
cells to become phoretic mites.

Therefore, the total number of mites exiting worker and drone cells is a combination
of the number of mothers entering the cells and the number of live offspring that each
produces.

Mite mortality is a function of natural mortality. Since the mite life expectancy is
approximately twice as long as the honey bees, the honey bee mortality rate was used for the
mite but it was multiplied by 0.5. An additional mortality factor is due to honey bee
mortality because if a phoretic mite is on a honey bee it will die if the honey bee dies. The
mite will not leave the dead bee. In addition, approximately 1.5% of the mother mites died
in the sealed brood. The equation for mite mortality is as follows:

MMP = (MP*MB)*.5 + 0.015*(MPDC+MPWC) Eq. 8
where

MMP = Mite population mortality

MP = Mite population

MB = Mortality rate for bee

MPDC = Mite p0pulation entering drone cells
Mch = Mite population entering worker cells

75

RESULT AND DISCUSSION

The model simulates bees and mites populations and illustrates the effect of the bees
on the mite population and the effect of the mites on the bee population.

The model incorporates information from previous models such as models cited in
Degrandi-Hoffman et al. (1989) and Fries et al., (1994) and from experimental data. The
model has the flexibility to alter some of the parameters that effect both bees and mite
populations. By altering those parameters, the model will be a valuable tool that indicates
which values have the most effect on populations. This should lead to more effective
application for control or more effective bee breeding programs to select for tolerant or
resistance bees.

When starting with 10,000 adult bees in January, the colony p0pulation peaked at
50,000 adults in the end of July (Figure 5). Worker brood population peaked at about 19,000
in the end of June while drone brood started in the end of March and peaked at about 1,900
the first week of June (Figure 6) with a ratio of about 11/89 to workers. The egg laying was
maximum the first week of June at about 1,500 eggs/day.

When starting with higher initial population in January the population will remain
higher and the peak is higher. Figure 5 shows that there are almost 7,000 more adults if the
model starts with 5000 more initial population. The colony will enter the winter with about
2,000 more adult bees which will result in more bees the following spring. These results are
in agreement with Nolan (1925) and Degrandi-Hoffman et al. (1989). Under Midwest
conditions, the overwinterd colony must be over 10,000 in January to be able to survive and

the number of bees will strongly effect the initiation and amount of egg laying because more

 

76

 

 

 

 

 

 

 

 

 

 

 

 

 

60000 !’_\ 20,000—9/1‘
ll: ‘- }\ 15,000 —3-; '.‘\\
50000 ,: - ‘.i -‘-' ‘ "n
’l/ 2‘ 10,000 \\
40000 E
i ’7 1
m l'
a 30000 It;
* "
1
0
'33§E§§§§§§§§§§
Days

 

Figure 5. Adult bee population development when starting with different initial

 

 

 

 

 

 

 

 

 

populations in January
2000 , 20000
1800 " I ‘ 18000
di— ‘ \ .1:-
16001» I ‘\ \\ -- 16000
140° “ \, \VVorker brood I ‘\ ‘“ 14000
§ 12001. I I \ «L 12000 g
2 I t r" \ g
e 1000 -- lI \ I \ ~- 10000 3
g 800 -- , \\ Drone brood \ V“ -- 8000 “5
l ~ 4*
m.“- I, \—J’\ I ~ “ T601”
400-. , l 1' 1 --4000
200 -- 1’ ‘1 ,I K -- 2000
O 4”, ‘A-” ‘V 0
"925%:“323‘88833392
— —' N 3 N m M V V tn '0 In \D ‘O 1‘
nova

 

 

Figure 6. Worker and drone brood population development.

 

 

77

bees are needed to increase the nest temperature to the degree where the brood will be able
to survive.

When using the values in Table 1, and starting first of April with 10 mite’s, about
10,000 bees and about 4,000 worker brood cells. The infestation will stay relatively low until
the drone and worker populations reach their maximum. At this time the number of mite will
increase sharply (Figure 7). The relatively high amount of brood in September and early
October and the increased mite population resulted in mite population peak at about 1,375
in the end of October (Figure 7). The second year, the bee population decreases up to the end
of August and dies early September (Figure 9). These results concur with the study by Ritter
(1984).

The times at which a certain mite number is reached, as well as whether these
numbers lie above the damage threshold, depends mainly on the initial degree of infestation
in Spring (Ritter, 1988), the amount of the brood reared and the degree of immigration.
Figures 8 and 10 shows the mite population with different initial infestation. Starting with
5, 10 or 20 mites, result will be peak populations of 690, 1,339 and 2,521 in the first year,
and as the initial infestation increases, the population peaks early in the season in the second
year. The same trend will continue the next year until the damage threshold is reached. It
seems that the threshold depends on the ratio of mites to bees. If the number of mites is high
in proportion to bees, the bee population can not support the colony and it will die.

Mite population has been studied in USA recently (Dclplane and Hood, 1997) in
Georgia Started with new packages of bees which contained a small incipient population of
V. Jacobsom'. The mite population development was in June 4271110, in August 31721324

and 666212127 in October. In their experiment, the development of mite population in

78

 

 

 

20000 18000
18000 16000
16000 14000
14000 12000
12000
8 10000 a
2 10000 g
m 8000
8000
6000 6000
4000 4000
2000 2000
0 0
'-‘ m V) I\ °\ "" m V) I‘ C\ '-‘ «1 V3 1‘ ex —‘ «1 V3
V co N \D —‘ W O m 1‘ N \O O V 00 M 1‘ '—
—t '— N N N ('0 M V V V3 V3 V3 \0 \D 1‘
Days

 

 

Figure 7. Mite population in correlation with worker and drone brood development.

 

 

3000

 

2500 r‘x

 

2000 , \

 

1500 20 mites / \

 

 

I of mltes

1000

 

500

 

 

 

 

 

 

 

 

Figure 8. Mite population development in the first year when starting in April with
different initial infestation.

79

 

 

 

 

 

 

 

 

 

 

 

30000
25000 Ix
\
I I
20000 I t
a I l
o l
3'.’ o“,
'E- 15000 20 mites J1 '1 l.‘
O o 0
* N s. ‘o
l ' 1
10000 10 mites \éa'; I 1
jmites I ' l i
5000 - i '1 i .‘
’- ' ‘ .0
’r.’...oo.\.\ --__—:v."{" ‘ I
0 Wu“ \
-- m tn r~ at m m In 1‘ Ch -— m tn b ca — tn
s1- co N to —- tn cm to h N c G <1- co en r~
— -— N N N en en er v on in In to to

 

 

 

Figure 9. Comparison between mite population development when starting with

different initial infestations in April until the collapse of the colony the
second year.

80

Georgia were compare with mite population in North Carolina in North Carolina. They
attributed the higher number in NC colonies to higher number of bees. Kraus and Page
(1995) started with 50 mites at the end April and the mite population was up to 2,367 in
October. And when started with 50 mites at the end of October, the mite population was up

to 1,620 in April. They stated that the mite population increased up to 300 fold during one

summer.

LITERATURE CITED

Anderson, DL (1994): Non-reproduction of Varroa jacobsoni in Apis mellifera colonies in
Papua New Guinea and Indonesia. Apidol. 25, 412-421.

Avitabile, A (197 8): Brood rearing in honeybee colonies from late autumn to early spring.
J. Apic. Res. 17 (2), 69-73.

Beetsma, J; Zonneveled, K ( 1992): Observations on the initiation and stimulation of
oviposition of the Varroa mite. Exp. & Appl. Acar. 16(4), 303-312.

Blum, R (1989): Einfluss einer Unterschiedlichen proteinernahrung von Honigbienen aufdie
Reproduktion der hamophagen Milbe Varroa jacobsoni. Diplomarbeit Thesis,
Fakultat Biologic, Universitat Tubingen, Germany. 51pp p.

Boecking, O; Drescher, W (1994): [Rating of signals that trigger Apis mellifera L. bees to
remove mite-infested brood] Bewertung von Ausloesefaktoren des
Verhaltensmerkmals Ausraeumen milbeninfizierter Brut bei Apis mellifera L. Apidol.
25(5), 459-461.

Boecking, O; Drescher, W (1991): Response of Apis mellifera L. colonies infested with
Varroa jacobsoni Oud. Apidol. 22, 237-241.

Boecking, O; Drescher, W (1990): The reaction of different Apis mellifera colonies to
Varroa infested brood cells. In: Proceedings of the international symposium on recent
research on bee pathology, Septemper5-7 1990, Gent, Belgium. (Ed: Ritter),, 41-42.

Boecking, O; Ritter, W (1993): Grooming and removal behavior of Apis mellifera intermissa
in Tunisia against Varroa jacobsoni. J. Apic. Res 32 (3/4), 127-134.

Boot, WJ; Calis, JNM; Beetsma, J (1992): Differential periods of Varroa mite invasion into
worker and drone cells of honey bees. Exptl. & Appl. Acar. 16(4), 293-301.

81

Boot, WJ; Calis, JNM; Beetsma, J (1993): Invasion behavior of varroa mites into honey bee
brood cells. In: Asian Apiculture. (Eds: Connor,LJ; Rinderer, TE; Sylvester,HA;
Wongseri,S) Wicwas Press, Cheshire, Connecticut, 491-498.

Boot, WJ; Calis, JNM; Beetsma, J (1995): Does time spent on adult bees affect reproductive
success of Varroa mites? Entomol. Exp. Appl. 75(1), 1-7.

Boot, WJ; Sisselaar, DJA; Calis, JMN; Beetsma, J (1994): Factors afecting invasion of
Varroa jacobsoni (Acari: Varroidae) into honeybee, Apis mellifera (Hymenoptera:
Apidae), brood cells. Bull. Entomol. Res. 84, 3—10.

Buchler, R; Drescher, W (1990): Variance and hertiability of the capped developmental stage
in European Apis mellifera L. and its correlation with increased Varroa jacobsoni
Oud. infestation. J. Apic. Res 29(3), 172-176.

Camazine, S (1988): Factors affecting the severity of Varroa jacobsoni infestations on
European and Africanized honey bees. In: African Honey Bees and Bee Mites. (Eds:
Needham,GR; Page, RE, Jr; Delfinado-Baker, M; Bowman,C) Ellis Horwood,
Chichester, 444-451.

DeGrandi-Hoffman, GD; Roth, SA; Loper, GM; Erickson, EH, Jr (1987): BEEPOP; a
computer simulation model of honey bee colony population dynamics. Amer. Bee J.
127, 846-847.

Fries, I; Aarhaus, A; Hansen, H; Korpella, S (1991a): Comparison of diagnostic methods for
detection of low infestation levels of Varroa jacobsoni in honey-bee (Apis mellifera)
colonies. Exptl. & Appl. Acar. 10, 279-287.

Fries, I; Aarhus,A; Hansen, H; Korpela, S (1991b): Development of early infestations by the
mite Varroa jacobsoni in honey-bee (Apis mellifera) colonies in cold climates.
Exptl. & Appl. Acar. 11(2-3), 205-214.

Fries, I; Rosenkranz, P (1993): Number of reproductive cycles of the Varroa mite. Apidol.
24 (5), 485-486.

Fries, I; Rosenkranz, P (1996): Number of reproductive cycles of Varroa jacobsoni in honey-
bee (Apis mellifera) colonies. Exp. & Appl. Acar. 20(2), 103-112.

Fries, F; Camazine, S; Sneys, J (1994): Population dynamics of Varroa jacobsoni: A model
and a review. Bee World 75 (1), 5-28.

Fuchs, S (1992): Choice in Varroa jacobsoni Oudemans between honey bee drone or worker
brood cells for reproduction. Behav. Ecol. Sociobiol. 31(6), 429-435.

Fuchs, S; Langenbach, K (1989): Multiple infestation of Apis mellifera L. brood cells and
reproduction in Varroa jacobsoni Oud. Apidol. 20, 257-266.

82

Fukuda, H; Sakagami, SF (1968): Worker Brood survival in honeybees. Res. Popul. Ecol 10,
31-39.

Grobov, O. 1977 Varroasis In Bees. In: Varoasis, a honeybee diseases. Apimondia
Publishing House, Bucharest,, 46-69.

Harris, JL 1985: A model of honeybee colony population dynamics. J. Apic. Res. 24(4), 228-
236.

Howell, D; Usinger, R 1933 Observation on the flight and length of life of drone bees.
Annals of the Entomological Socity of America 26, 239-246.

Ifantidis, M (1988): Some aspects of thr process of Varroa jacobsoni entrance into honeybee
Apis mellifera brood cells. Apidol. 23 (4), 227-233.

Ifantidis, MD (1984): Parameters of the population dynamics of the Varroa mite on
honeybees. J. Apic. Res. 23(4), 227-233.

Ifantidis, MD (1990): Reexamination of reproduction parameters of the mite Varroa
jacobsoni Oudemans, .

Jay, SC (1963): The development of honeybees in their cells. J. Apic. Res. 2, 117-134.

Koeniger, N; Koeniger, G; Wijayagunasekaran, H (1981): Beobachtungen uber die
Anpassung von Varroa jacobsoni an ihren ursprunglichen Wirt Apis cerana in Sri
Lanka. Apidol. 12 (1),37-40.

Korpela, S; Aarhus, A; Fries, I; Hansen, H (1992): Varroa jacobsoni Oud. in cold climates:
population growth, winter mortality and influence on the survival of honey bee
colonies. J. Apic. Res 31(3-4), 157-164.

Kraus, B; Page, RE, Jr. (1995): Population growth of Varroa jacobsoni Oud in
mediterranean climates of Califomia Apidol. 26, 149-157.

Kusterrnann, T (1990): Populationsstruktur der varro-milbe in Arbeiterinnenebrut. Deutsches
Imker- Journal 1 (11), 436-437.

Marcangeli, JA; Eguarae, MJ; Fernandez, NA (1992): Reproduction of Varroa jacobsoni
(Acari: Mesostigmata: Varroidae) in temperate climates of Argentina. Apidol. 23, 57-
60.

Marcangeli, J; Eguaras, M; Fernandez, N (1995): Population growth of Varroa jacobsoni
(Garnasida: Varroidae) in colonies of Apis mellifera (HymenopterazApidae) in
temperate climates using camazine’s model. Apiacta 30 (1), 13-19.

Martin, SJ (1994): Ontogenesis of the mite Varroa jacobsoni Qud. in worker brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 18(2), 87-
100.

83

Martin, SJ (1995): Ontogenesis of the mite Varroa jacobsoni Oud. in drone brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 19(4), 199-
210.

McLellan, AR; Rowland, CM; Fawcett, and RH (1980): A monogynous eusocial insect
worker population model with particular reference to honeybees. Insects Sociaux 27,
305-311.

Mikityuk, V (1979): [Reproductive capacity of female varroa mites]. Pchelovod. 9, 21.

Mikityuk, V; Korzjhova, L; Sedin, I (1976): [Experiments on the biology of the varroa mite].
Pchelovod. 12, 19-20 (in Russian).

Moosbeckhofer, R (1991): Varroaverluste wahrend der uberwinterung. Bienenvater 112 (9),
300-

Moosbeckhofer, R; Fabsicz, M; Kohlich, A (1988): Untersuchungen Uber die Abh'a’ngigkeit
der

Moretto, G; Concalves, S; De Jong, D ( 1991a): Africanized bees are more efficient at
removing Varroa jacobsoni - Preliminary data. Amer. Bee J. 131, 434.

Moretto, G; Concalves,S; De Jong, D (1991b): Africanized bees are more efficient at
removing Varroa jacobsoni - Preliminary data. Amer. Bee J. 131, 434.

Muller, M (1987): Befallsentwicklung der Milbe Varroa jacobsoni den Wintermonaten.
(summary). Allg. Dtsch. Imkerztg. 2, 6-11.

Nachkommensrate von Varroa jacobsoni Ouf. vom Befallsgrad der Bienenvolker. Apidol.
19, 181-208.

Nolan, W (1928): Seasonal brood-rearing activity of the Cyprian Honeybee. J. Econ.
Entomol. 21, 392-401.

Nolan, WJ ( 1925): The brood-rearing cycle of the honeybee. U.S.D.A. Dept. Bull. No. 1349.
(55 pp)

Omholt, S; Cralisheim, K (1991): The possible prediction of the degree of infestation of
honeybee colonies Apis mellifera by Varroa jacobsoni Oud. by means of its natural
death-rate: a dynamic model approach. Norwegian Journal of Agriculture Sciences

5, 393-400.

Peng, Y-S; Fang, Y-Z; Xu, S-Y; Ge, L-H (1987): The resistance mechanism of the Asian
honey bee, Apis cerana Fabr., to an ectoparasitic mite, Varroa jacobsoni Oudemans.
J. Invertebr. Pathol. 49(1), 54-60.

Rademacher, E; Geiseler, E (1986): "Die Varroatose der BienenzGeschichte, Diagnose,
Therapie. In: (Eds: Verlag, S; Jeep),, .

84

Rath, W; Drescher, W (1990): Response of Apis cerana Fabr. towards brood infested with
Varroa jacobsoni Oud. and infestations rates of colonies in Thailand. Apidol. 21,
311-321.

Ribbands, CR (1953): The Behavior and Social Life of Honeybees. International Bee
Research Association, London. 352 pages.

Ritter, W; De Jong, D (1984): Reproduction of Varroa jacobsoni O. in Europe, the Middle
East and tropical South America Z. Angewandte Entomol. (J. Appl. Entomol.) 98(1),
55-57.

Ritter, W; Leclercq, E; Koch, W (1984): [Observations on bee and Varroa mite populations
in infested honey bee colonies] Observations des populations (1’ abeilles et de Varroa
dans les colonies a differents niveaux d’ infestation. Apidol. 15(4), 389-399.

Ritter, W (1984): Neuester Stand der diagnostischen und therapeutischen Massnahmen zur
Bekampfung der Varroatose. Tierarztliche Umschau 39, 122-127.

Ritter, W (1988): Varroa jacobsoni in Europe, the tropics, and subtropics. In: African Honey
Bees and Bee Mites. (Eds: Needham,GR; Page,RE,Jr; Delfinado-Baker,M;
Bowman,C) Ellis Horwood, Chichester, 349-359.

Ritter, W (1990): Development of the Varroa mite population in treated and untreated
colonies in Tunisia. Apidol. 21(4), 368-370.

Rosenkranz, P; Engels, W (1994): Infertility of Varroa jacobsoni females after invasion into
Apis mellifera worker brood as a tolerance factor against varroatosis. Apidol. 25(4),
402-411.

Rothenbuhler, WC (1964): Behavior genetics of nest cleaning in honey bees. IV. Responses
of F1 and backcross generations to disease-killed brood. Amer. Zoo]. 4, 11-123.

Ruijter, A, De (1987): Reproduction of Varroa jacobsoni during successive brood cycles of
the honey bee. Apidol. 18, 321-326.

Ruttner, F; Hanel,H (1992): Active defense against Varroa mites in a Camiolan strain of
honeybee (Apis mellifera carnica Pollmann). Apidol. 23, 173-187.

Ruttner, FH; Marx, G (1984): [Observation about a possible adaptation of Varroa jacobsoni
to Apis mellifera L. in Uruguay] Beobachtung neber eine moegliche Anpassung von
Varroa jacobsoni an Apis mellifera L. in Uruguay. Apidol. 15(1), 43-62.

Schulz, AE (1984): Reproduction and population dynamics of the parasitic mite Varroa
Jacobsoni Oud. and its dependence on the brood cycle of its host Apis Mellifera.
Apidol. 15 (4), 401-420.

85

Sulimanovic, D; Ruttner, F; Pechhacker, H (1982): Studies on the biology of reproduction
in Varroa jacobsoni. Honeybee Sci. 3, 109-112.

Weiss,K (1984): Bienen-patholgie (Summary). Franz Ehrenwirth Verlage,
Woyke, J (1987): Infestation of honeybee (Apis mellifera) colonies by the parasitic mites

Varroa jacobsoni and Tropilaelaps clareae in South Vietnam and results of chemical
treatment. J. Apic. Res. 26(1), 64—67.

USING MODEL SIMULATIONS TO PREDICT POPULATION RESPONSES IN
HONEY BEES AND MITES BY INTRODUCING BIOLOGICAL CONTROL,
CHEMICAL CONTROL, REINVASION, AND GENETICALLY MANIPULATED

CHARACTER TRAITS INTO THE SYSTEM

ABSTRACT

Treatment efficiency and time of treatment were simulated in the honey bee and
Varroa mite model and the results show that one chemical treatment is sufficient to control
the mites for two years if there is no reinvasion from outside and if the beekeepers follow the
label concerning length of treatment. However, reinvasion is inevitable in the field and that
explains the failure of control measures and colony death even when treated once a year. This
result shows how important it is for beekeepers to try to reduce reinvasion by cooperating
with each other in the same area, by treating in the same season, and by including in their
management decisions, practices that help in reducing drifting, robbing, swarming etc.

Controlling the mites by means of non-chemical measure is labor intensive, time
consuming and is not practical for large operation beekeepers. However, some of the hobby
and sideline beekeepers who have a few colonies and are willing to put in the effort, might
find these techniques useful. Although, intensive drone removal reduces a high portion of
the mite population, it still is not successful in controlling the mite population during the
second year, when the population becomes high.

86

87

Trapping comb and hyperthermia techniques alone will not work for more than two
years in high and moderately high population colonies. These techniques reduce the mite
population significantly but not enough to control it.

The honey bee and Varroa mite model is used to evaluate some of the important mite
and bee traits involved in resistance or tolerance of honey bees against Varroa colonies such
as, grooming, brood removal, brood attractiveness, infertility of mites, post-capping period,
and reproductive cycle.

The results obtained indicate that the model is a valuable tool for predicating the
changes in mite and bee populations resulting from varying a single factor concerning the
above mentioned traits. The values for these simulations were obtained from the research
literature when available, or expert opinion when data was not available. The results from
the model simulations illustrate the effect of these traits on the bee and mite populations and
it shows that any thing that effects mite reproduction such as number of offspring, fertility,

reproductive cycle and phoretic period are the most important traits for which to select.

INTRODUCTION

Modeling is an important tool that is used to predict the effect of some factors on mite
populations and to show which factors have major effects. Studying of such factors in the
field might take a long time and be costly. In addition, the model can consider the affect of
these factors separately or simultaneously.

In the honey bee and Varroa mite model we attempt to show the effect of different
control measures and the effect of reinfestation on mite population dynamics. In addition, we

use the model to evaluate the effect of separate resistant traits on the population growth of

88

the mite and hopefully the model will clarify which resistance traits should be chosen for

selection.

BIOLOGICAL AND CHEMICAL CONTROL AND THE EFFECT
OF THE REINVASION ON MITE POPULATION DEVELOPMENT
Introduction

Controlling the mite population is necessary to keep the mite density under the
economic threshold and prevent the death of bee colonies of European races. More than 145
pesticides have been tested and used for control of Varroa (W ienands, 1988). Pyrethroid
Fluvalinate (Apistan), the most common acaricide that has been used to control Varroa all
over the world, is the only one approved for controlling these mites in the United States.
Reports about Varroa mites developing resistance to Apistan (Lodesani er al., 1995) and the
pesticide residues in honey and wax from Apistan and other registered products (Maria and
De Paoli, 1994; Milani, 1994; Bomeck and Merle, 1990; Hansen and Petersen, 1988;
Barbina et al., 1989) make it important to research alternative methods of reducing the mite
population and limiting the use of acaricides inside the colonies.

Attempts have been made to control the mite population without using chemicals or
with minimal use with various degrees of success. Although these methods are time
consuming, labor intensive and may not be suitable for commercial beekeepers, those who
do not want to use chemicals inside the hives, such as hobby and side line beekeepers may

find these alternative methods useful.

89

Drone removal

Literature review. Varroa prefer drone brood to worker brood, proportionally drone
cells are infested by approximately 12 times as many mites as worker cells (Calis er al.,
1996; Boot et al., 1995). This preference can be utilized by beekeepers to use drone combs
to trap Varroa.

The essence of this technique is the insertion of a drone comb or frame with drone
foundation into the colonies and regularly remove the brood once it has been fully sealed.
Schulz et al., (1983) reported that drone comb removal was effective in controlling Varroa
in lightly infested colonies, in heavily infested colonies it was only 15% better than the
control and thus failed in slowing down the build up of the Varroa population. For this
procedure to be effective, all drone brood present on any other comb in the colony must be
cut out and removed (Weiss, 1984) otherwise mites may invade drone cells that are not
removed from the colonies. Another way to do this is to cut and destroy all the drone sealed
brood. Ritter et al. ( 1981) reported a 10% reduction of mite populations in both high and low
infested colonies. Shilor (1980) reported a 54% reduction.

Lavagnino and Marletto (1996) periodically removed drone brood in two apiaries,
cutting the drone brood 12.2 and 10.8 times. On average, the total number of mites removed
averaged 1448 and 10303, receptively. In the second year the numbers were lower. Although
this method was effective, it could be used for only a few months of the year, and by itself
could not ensure colony survival.

Calis et al. (1996) reported that trapping mites in drone brood will be much more
effective when it is applied during periods when no other brood is present in the colony, this

will not allow the mites to escape by invading brood cells on other combs.

90

Boot et al. (1995) stated that trapping mites in drone brood can be an effective non-
chemical method to control Varroa when it is applied during periods when no brood other
than that introduced for trapping is present in the colony. They found that 462 drone cells are
estimated to be sufficient to trap 95% of the mites in a colony of 1 Kg of bees.

Calis et al. (1996) designed a method to trap mites in drone cells which resulted in
an overall effectiveness of 93%. In this study they used five pairs of colonies. One colony of
each pair was made broodless. The brood was given to the other colony. This colony was
split into a brood less part with the queen and a part in which a new queen was reared and
which became broodless when all the old brood had emerged. This provided three times the
optimal opportunity for trapping mites in drone brood, i.e., a broodless (part of a) colony.

Model Simulation. From the literature cited above, it is clear that there is a wide
range of drone removal techniques attempted by researchers. Methods ranged from
removing only a part of the drone brood to removing all the brood from the colony. The goal
of this model simulation is to demonstrate the effect of drone brood removal on the mite
population. The natural development of the mite and bee populations are shown in Figure
1. There was no attempt to remove the drone brood from this simulation. If the mite
population is not controlled, then the bee colony will collapse after two years.

Then the simulations were run removing a range of drone brood. If one removed
99% of the drone brood after sealing, the simulation showed that the mite population will be
reduced by more than 50% the first year (Figure 2). A 20%, 50% and 70% drone brood
removal corresponded to a 9%, 30% and 42% mite reduction in the first year (Figure 2).
However, the bee colony will still collapse during the winter of the second year or early in

the spring of the third year (Figure 3) because of the buildup in the mite population during

91

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50000 18000
45000 16000
40000 14000
35000
12000
10000
0 Q
§ 25000 g
a
20000 8000
15000 6°00
10000 400°
5000 2000
o 0
‘— ¢D '- 1- (D 1- CD 1— CD 1- CD 1- ‘0 1'- CD '- CD 1- Q
o~§ztaaaaaaeeeaaaaa
Days
Figure 1. Mite and bee population development when starting with 10 mites in
April and 10,000 adult bees in January until the colony collapses in the
second year.
1200 A
/ \\
20% I
1000
O
800
a \Z \
T: 600
'6 .'
* 990/ '0 ' \‘\‘\
400 I a' . t- ‘
I ' '. ~‘, \
I o) I \‘s‘ \
200 ,1“, a...
j .l
0

22
43
295

 

Figure 2. Mite population deve10pment in the first year when drone removal

technique is used to control the mite population.

 

 

92

 

16000

 

14000

 

1 2000

 

10000

 

8000

 

 
    

it of mites

6000

 

/'

4000

 

2000

 

 

 

 

 

 

m3———-“
715 - - 1 - '

757

799

841

 

 

 

Figure 3. Mite population development when using drone removal technique to

control the mite population: different percentages from the whole brood
is removed during the whole season.

93

the second year. Hence, drone brood removal alone is not sufficient to control the mite
population and it is necessary to use other methods of biological control along with drone
brood removal.

Trapping comb technique

Literature review. In this method of biological control, the queen is confined to one
comb, within a cage made of queen excluder. After a seven to nine day period, the queen is
removed from the brood comb and placed on a new comb. After a total of 28 days, or 4
transfers, the queen is released and allowed to move freely (Maul, 1983). The combs are
removed from the colonies after capping. Maul et al. (1988) recommended using trapping
comb technique for 2 or 3 years, because mite levels continued to increase after that amount
of time. Trapping comb technique gives an estimated reduction of 90% of the mites from
infested colonies (Fries and Hansen, 1989; Maul and Klepsch, 1986; Rademacher and
Geiseler, 1986).

There is some concern about colony survival without new bees, especially if the
technique is applied for four successive cycles (28 days + 21 days = 49 days without new
bees). Fries and Hansen (1993) reported if the sealed worker brood combs are destroyed, the
bee colony will still be able to survive the winter. Their results demonstrated that it is
possible to control Varroa using management methods only. However, they reported that the
relatively high mite mortality during the third and fourth experimental year indicated that the
use of the trapping comb technique by itself might not be sufficient for control of the mite
in all colonies over a period of several years.

Camilla and Hansen (1994) showed that mite control using the trapping comb

technique followed by heat treatment of the sealed brood for four hours at 44°C, combined

94

with drone brood removal, was sufficient to control Varroa infestations in most cases.
However, they suggested that even for skillful beekeepers this control procedure can cause
difficulties and as a solution, they suggested treating the colonies with a soft chemical after
nectar flow, when trapping comb technique has not been successful carried out. In this
procedure, the sealed worker brood was not destroyed but removed and treated with heat for
four hours at 44° C and then returned to the colonies. Calis et al. (1993) suggested killing the
mites trapped in capped worker brood outside the colonies with formic acid. They reported
that 73.5 - 80% of the dead mites recovered from the colonies were from the treated brood
cells.

Model Simulation. This model simulates the effect of the trapping comb technique
on mite and bee populations. The simulation prevents the worker brood from emerging for
49 days and also examines the effect of applying the treatment at different dates to illustrate
the impact that date of application will have on the mite and bee populations.

Two different times were chosen to apply this technique. The first time was early in
the season, at the end of April (Figures 4, 6), and the second time was at the end of July or
just after the peak of the adult population (Figures 5, 7). If one chooses to use this technique
early in the season, then the number of mites remaining after the treatment will be
significantly reduced, which will result in less mites after the peak of the drone and worker
broods. The drawback of this time period is that the honey bee population will also be
reduced by about 20% (going from 50,000 to 40,000), which will result in less honey
production.

Using the technique later in the season has one major disadvantage in that the mite

population will not be affected until it has already become high. After the treatment the mite

 

95

 

 

50000 55000
45000 50000
40000 45000
35000 40000
3 30000 35000
= 30000
5 25000
'3 25000
" 20000 20000
10000 10000
5000 5000
0 0
v- 0) In B 0) v- 0) IO N O) P 0) IO N O) '07) ('3 IO N O) P 00 ID
sweeaaaasssaaaoaxasaas
days

 

 

 

Figure 4. Mite and bee population development when using trapping comb
technique at the end of April the first year to control Varroa population.

 

60000 9000

 

 

 

 

Boos
p‘
\
h
0".
I
b.
C
.-
q
0'...

20000

. , . 3000
a.-." .' ..\ ..-" / “ 1‘. 2000
10000 . ‘ ". '

 

”i

 

 

 

 

4 '0
MNOVNFVQngNmO’NCGO
FPFNNNQC’) VVVIDIDIDCDCDN

 

Figure 5. Mite and bee p0pulation development when using trapping comb
technique at the end of July the first year to control mite populations.

 

96

 

45000

35000
30000
25000
20000
1 5000
1 0000

5000

 

In
C
P

 

126

300
250
200
150 a":
a
100

50

N 00 O) o v— N ('3 V I!) CD IN

V (D as '- (‘J ID 1‘ O) 1- 8 ID

'- v- v- N N N N N ('3 G)
Days

 

Figure 6. Mite and bee population development in the first year when using
trapping comb technique at the end of April to control mite populations.

 

 

 

 

 

 

 

 

 

 

 

60000
Mites A
50000 .*" . ‘ ..350
Bees /\ Y \
40000 ‘ .‘ . 30°
\' ‘. +250
3 30000 ' '- 3
s / ‘- V "2°“;
20000 7 “ 15°
...... f ~....,..\
10000 . . .
«50
0_ o
0 v- N ID ‘0 N a) a, O 1- N ('3 V IO N
«eggeszeeasssaa‘fia
Days

 

Figure 7. Mite and bee population development in the first year when using
trapping comb technique at the end of July to control mite populations.

 

 

97

numbers will increase again because of the high number of mites entering the broods to
multiply. The advantages of this time frame are that more honey will be produced because
the honey bee population will also be greater. There will be more space available to store
the honey. Also this technique might increase the number of foragers by shifting the workers
duties from nursing to foraging.

This model estimates a mite population maximum of 1339 in the first year without
using trapping comb technique (Figure 1). This number is reduced to 279 mites when using
the technique at the end of April (Figure 6), and to approximately 418 mites when using it
at the end of July (Figure 7).

The model can also be used to compare colony survival from the effect of treatment
timing. The overall effect on colony survival shows that the colony actually survived longer
when the mites were reduced early in the season because the overall mite population was less
going into winter (Figure 6). The colony survived through the third year, although it was not
healthy that third year. It collapsed immediately after the mite population drastically
increased after the peak of the worker brood. When the trapping occurred at the end of July,
the colony still collapsed in October of the second year (Figure 5). It collapsed early because
the population was higher coming out of the first year of treatment.

The next step was to see the effect on the mites and the bee colony if the trap combs
were used for two years. Both times of treatment allowed the colony to survive a third year
when the trapping technique was used in two consecutive years (Figures 8, 9). The peak bee
population was lower the first year with the earlier treatment, but both treatments resulted
in a bee population ca 50,000 during the second year. The colonies survived until June and

July, respectively, the third year. They collapsed because of the high mite population

9E3

 

50000
45000
40000
35000
30000
a
8 25000
at
20000
15000
10000
5000

 

@ml‘mv-MIONOQv-COID
vmmtov-mmmhcucoo
FPNNNGGVVU)
Days

547

589
631

673
71 5

757

799

841

883
925

 

40000
35000
30000
25000
20000 g
1 5000
1 0000

5000

 

Figure 8. Mites and bee population development until the collapse of the colony
when trapping comb technique is used in the end of April the first year

and end of July the second year.

 

I at Mltas

 

 

50000
45000

35000
30000
25000
20000
1 5000
1 0000

 

Figure 9. Mite and bee population deve10pment until the collapse of the colony
when using trapping comb technique in the end of July the first year and

end of July the second year.

 

 

99

following the brood peak during the third year. However, the technique was successful at
removing about 60% of the mite population in the second year (Figure 10 & 11). The
relatively high numbers of brood that are present in August and September allow the mite
population to increase after the treatment.

The last model was used to simulate the effect of using the trapping comb technique
for three years (Figure 12). It illustrates that the trapping comb is good at controlling the mite
population for two years in highly infested colonies. With low infested colonies, it might
work a bit longer. This result is in agreement with Maul (1988) who reported that the
technique is good for 2 or 3 years, but reported that mite population levels continue to
increase after that time frame. Fries et al. (1993) stated that the trapping comb technique
might be insufficient for controlling mites in all colonies over a period of several years due
to the high mortality during the third and fourth experimental year.

The next simulation used the two years of trapping comb technique combined with
a third year of 99% drone removal. The colony still collapsed at the spring the third year
(Figure 13).

Hyperthermia

Literature review. Varroa females are more sensitive to temperature above that of
the normal honeybee brood nest level than are larval and pupal bees. Based on this slight
difference in tolerance to heat, a heating method is used to eliminate Varroa (Rosenkranz,
1988). Rosenkranz (1987) exposed frames of brood without adhering bees to high
temperatures in an effort to kill Varroa mites within the cells. Between 80—100% of adult
Varroa females and 100% of the nymphal stage of the mite were killed in the broods that

were exposed to 40C° (104°F) for 12h, to 44C for 5h, or to 45C (1 P3 F) for 4h. when

100

 

 

ii of mites

71
106

141

176

211

 

246
281

Treatment in July

Treatment in April

F0
38
DOV!

31 6
421
456
491
526
561
596
631
701

 

Figure 10. Mite population development showing the effect of trapping comb
technique in first and second year.

 

I 0! mites

 

3500

2500

8

1500
1000

500

 

 

/"\

 

 

End of June 8 end of July I \r \
\ V

 

End of April & end of .10wa ]A\

 

/7\7\/

 

Z/

 

 

 

 

701

 

Figure 11. Mite populations for the first two years when using trapping comb
technique in the end of April or the end of July the first year.

 

 

101

 

 

 

25000

 

20000

 

15000

Mites

 

10000

5000
O t M

 

 

 

 

 

 

 

43
85
127
169

21 1
253
295
337
379
421
463

P 505
.3 547
589
831
673
715
757
799
841
883
925
967

 

 

Figure 12.A comparison of mite populations when using the trapping comb
technique for three years 1- end of April the first year & end of
June the second year & end of April the third year 2- end of June the
first year & end of June the second year & end of April the third year.

 

 

30100

25100

20100

15100

#01 Bees

10100

5100

 

100

E§§§§§E§§§E§§EEE§§§§§

 

 

Figure 13. Mite and bee population development when using trapping comb
technique in the end of July first year and end July second year and 99%
drone removal in the third year

102

considering all stages of the brood comb, the maximum rate of damaged brood was about
5%, although the early stages were slightly more sensitive to the high heat. Appel and
Buchler (1991) found similar success when the brood was heated in an incubator shortly
before the emergence of the bees, 100% of the mites were killed after at least 4h treatment
with 44C°. They recommend heat-treatment for trapping combs when there is precision in
controlling temperature and heat regulation.

Engels (1994) stated that hyperthermia was 100% effective, about one day after the
treatment all adult female mites together with the nymphs were found dead in the capped
brood cells. According to the amount of capped brood present in the colony, with one
hyperthermia application between 50% and 80% of the total mite population in a hive can
be eliminated. Hoppe and Ritter (1986) applied thermal therapy in the field using 6 artificial
swarms of normal size (ca. 15,000 bees). The swarms were treated at 48°C for 20 minutes
and 6 others at 50°C for 15 minutes. In comparison to the controls, an average of only 23 and
38% of the mites, respectively, were killed. The low mite mortality is due to the strong
tendency of the bees to reduce the temperature by cooling and heavy ventilation (Engels and
Rosenkranz, 1992; Hoppe and Ritter, 1986). Therefore, hyperthermia of capped brood frames
outside the hive remains the only applicable method .

Model Simulation. This is different from the previous two control strategies because
the brood will not be damaged and the bees will emerge inside the colony. The model
predicts an approximate 50% reduction in the mite population (Figures 14, 15). The
simulation shows the differences in mite populations in the first year when no control was
used, and when the hyperthermia was 90% and 100% effective (Figure 14). The colony

collapsed sooner with 90% efficiency than with 100% (Figure 15). The results, once again,

103

 

 

 

1400

1200 No control ___;,/\

.00. [90}

800 / ‘ \

600 .

400 / 100% \\

200 / \\
/ \

 

 

 

 

ii of Mites
\ ‘q$
/

 

 

 

 

 

0) ID to N m a: O V- N 0') V In ‘D N CD 0) O 1-

>~ v- (V) L0 N 03 N V (D m C N V (D Q C ('0 ID

8 v- v- v- v- N N N N N (O CO
Days

 

 

Figure 14. Comparison between mite populations in the first year with and without
hyperthermic control.

 

1 8000
1 6000
1 4000
1 2000
1 0000

No control

4 0i Mites

6000

 

1
36
71

106
141

176
21 1

246
281

316
351

366
421

456
491
526
561
596
631
666
701
736
771
806

Days

 

 

Figure 15. Mite population development until the collapse of the colony when
hyperthermic technique is used to control the mite population.

 

104

indicate that this single biological control mechanism is not sufficient to control the mite
population alone.
Chemical control

Literature review. More than 145 pesticides have been tested and used for control
of Varroa (W ienands, 1988). Pyrethroid Fluvalinate (Apistan), the most common acaricide
that has been used to control Varroa all over the world, is the only one approved for
controlling these mites in the United States. Efficacy of Apistan generally ranges from 98-
100% (Hillesheim et al., 1996). Reports about Varroa mite developing resistance to Apistan
(Lodesani et al., 1995) and the pesticide residues in honey and wax from Apistan and other
registered products (Maria and De Paoli, 1994; Milani, 1994; Bomeck and Merle, 1990;
Hansen and Petersen, 1988; Barbina et al., 1989). By using the model simulation, we attempt
to illustrate when the most effective time to treat and to show the effect of different doses and
length of treatments.

Model simulation. The model can be used to simulate the effect of chemical control
on mite p0pulation dynamics. This model assumes there is no reinvasion from outside. Two
different efficacy rates were chosen, 70 and 90%. If a chemical control is used for 65 days
and is only 70% effective it will control the mite population for three years and the colony
will collapse after the brood peak in the fourth year, if not treated again (Figure 16). Similar
effects were found when the control was 90% effective (Figure 17); although, the mite
population was considerably lower at its last peak.

If chemotherapy is used, with an efficacy of 99%, for seven days the mite population
is reduced by about 50% to a peak of ca. 625 the first year (Figure 18); but the colony will

still collapse at the end of the second year (Figure 19). The mites that were in sealed cells

105

 

# of Bees

50000
45000
40000
35000
30000
25000
20000
1 5000
1 0000

5000

 

57

1 13
1 69
225
281
337
393
449
505
561
61 7
673
729
785
841
897
953
1 009

Days

1 065

1121

1177
1233

1289

35003

30003

25003

20003

I of Mites

1 5003

E

'5003

 

 

Figure 16. Mite and bee populations when using chemical control with 70%

efficacy at the end of August for 65 days.

 

 

11 of Mites

50000
45000
40000
35000
30000
25000
20000
1 5000
1 0000

5000

 

CID
If)"
v-

400
457
51 4
571
628
685
742
799
856
91 3
970
1 027

01030
BNO
PNN

Days

1084

1141

1198
1255
1312

25003

20003

15003

10003

 

Figure 17. Mite and bee populations when using chemical control with 90%

efficacy at the end of August for 65 days.

 

 

106

 

 

 

 

 

 

 

 

 

 

 

60000 703

50000 Mites\_\_/\ "603

Bees\)/ / \ 503

40000

. / f l .. ~
5’ / / V\A\

~303

20000 // / 203

10000 i J ‘ 103

0 3

 

 

 

Figure 18.Mite and bee populations in one year when using chemical control with
99% efficacy in the end of July for 7 days only.

 

 

 

60000

50000

30000

20000

10000

12003

 

mites

\\/\

10003

 

\

8003

 

 

/\
Z \A

 

 

 

0

M9

SE6

8

8

36

71
106
141
176
211
246

281

5
Days

\'J

j/\
//\\
/

 

 

MQJM

 

Figures 19. Mite and bee populations until the collapse of the colony when using
chemical control with 99% efficacy at the end of August for 7 days only.

 

107

during the treatment were not effected by the treatment and were left to multiply after leaving
the cells. The colony collapsed in October the second year because seven days was an
insufficient length of time for control. When treatment lasted for 14 days (Figure 20) the
effect on mite population will be greater and the colony will survive until the brood peaked
in the spring the third year.

Reinfestation

Literature review. If a large number of colonies are highly infested within the flight
radius, the mite p0pulation can increase in five months from zero to an average of 6000 mites
per colony according to an investigation by Ritter (1986). In comparison, the increase in
infested colonies with little invasion pressure remains small during the same time period. A
maximum of 2000 mites was found in the colonies in the fall. Greatti et al. (1992); Milani
et al. (1993) reported that the daily reinfestation rate in highly density areas with honey bees
was low during spring, varied between 1.6-13.7 mites/day/colony during June, July and first
week of August, and rose during September and October up to 100 (on average, 31.6)
mites/day/colony; it was relatively high when nectar was scarce. Drones were prevented from
entering the colonies by means of queen excludes.

Irndore and Kilchenmann (1994) found that, numbers of mites started to increase in
the monitored colonies (they were free of mites) at one apiary in June when infested colonies
in a neighboring apiary started to die. Robbing of the weakened (infested) colonies was
observed. It is concluded that mites were carried by robber bees back to the healthy colonies;
number of mites transported is estimated as 3000-4000 mites/colony.

The zone within a radius 7 km a round the center of an infested apiary is considered

contaminated. The zone within a radius of 100 km (double of the flight range of swarm) is

 

108

 

 

 

 

 

 

 

 

 

 

60000
50000 f‘
1'
40000 :
'5 Treatment for 14 days ,'
E 30000 \ 5
'6 .'
.. \_.
20000 ':
Treatment for 7 days .i'
0
'- C a: Q N ‘0 ID V to N F C a: m N ‘D IO N '- O 03 Q
v5299955559958585582558

Days

 

 

Figure 20.Comparison of mite populations when using chemical control in the end
of August with 95 % efficiency for 7 or 14 days.

109

considered exposed (Sulimanovc, 1984). As one colony of bees reaches the point of total
collapse, the few bees that leave the hive carry 10 or more mites each, and fly to neighboring
colonies or apiaries (Mussen, 1993). Foraging bees also carry the mites. In foraging bees the
infestation increased from 0.2% in April to 21-32% in October, and in some colonies the
level of infestation of foraging bees increased 20-25 times (Artemenko et al., 1978).
Huttinger er al. (1981) report that 5.2% of the flying drones were infested and 6.1% of the
drones caught at the congregation area, while only 3.9% of the flying foragers and 0.7% of
water-carrying bees. Huttinger and Pechhacker (1988) observed a mean of 4.5 foragers
drifting in a day from each of three colonies 10-80 m apart, and even fewer bees drifted from
more distant colonies. Sulimanovic (1984) found that 6% of the bees drifted to another bee
hive in a stationary apiary, and in a migratory apiary the number of drifting bees increased
to 18%.

Marked drones have been found 8 miles from their home apiary, and some
researchers indicate that 23% of the Varroa mites outside cells may be on flying drones,
while 17-20% of the young bees of nursing age carried Varroa mites (Mobus and Connor,
1988). Currie and Jay (1992) found that 47% of the drones drifted from their parent colonies,
and 21% of those drones drifted more than once. The proportion of drones that drifted
decrease with increased spacing between colonies, but only at the longer distance. Although
30—60% of drones drifted to colonies up to 50m away, only 20% of the drones drifted to
colonies that were 100m away. However, the workers drifting was 44.3%, 26.3%, and 12.6%
when the distance between hives was 3, 6, and 9 m respectively (Jay, 1966).

Filipov (1978) found that infestation spread in 32 days within a 100 m radius, in 73

days within a 500 m radius, and in three months within a 6-11 km radius depending on the

110

density. Gandinger (1985) estimated that the spread from apiary to apiary was about 3.5
km/yr. Natural progression occurs at the rate of 30-40 km annually (Robaux, 1988). Ritter
and Leclercq (1987) reported that the risk of infestation can be reduced only by treating all
colonies in an area in the same season.

Model Simulation. The model can also be used to simulate the impact of reinvasion
on mite population dynamics (Figure 21). Geratti et al. (1992) estimated reinvasion patterns
for one year. They estimated that 2.4 mites per day reinvaded in June, 4.7 in July, 5.9 in
August, 31.6 in September and October, and 2.4 in November. Thus, the mite population can
increase by about 4,000 mites as a result of reinvasion pattern. According to Ritter (1996),
if there is a large number of colonies infested within flight radius, the mite population can
increase from undetectable levels to an average of 6000 mites in a five month period.

Figure (22) illustrates the difference in mite populations when there is no reinvasion
compared to one where 300 mites were introduced by robber bees on one day at the end of
August. Hence, the results presented here show the important role that reinvasion plays in
treatment protocol and on the outcome of research results when studying mite population

dynamics in the field.

SELECTION TOWARD VARROATOSIS RESISTANCE
Introduction

Selection toward varroatosis resistant or tolerant strains has become a major issue in
practical honey bee breeding (Koeniger and Fuchs, 1988; Kulincevic and Rinderer, 1988;
Otten, 1990; Mortiz and Jordan, 1992; Buchler, 1994 ). In a comprehensive review, Buchler

(1994) outlined the current scientific findings in colony defense mechanisms . Boot et al.

 

 

# of Mites

6000

5000

4000

3000

2000

1 000

 

Reinfestatioq

 

\/A\

 

/ \

 

No reinfestation/ \

 

AA\

 

 

 

/ \

 

(OFCDPCDFCDFCDFCDPCDFQPQFCDFCDPCDP
v-(‘OVCONODCNCOIDCDQOJFNVIONQOv-mgw
v-v-v-v-v-v-v-NNNNNNNC‘OO) 0)

Days

 

Figure 21. Mite population development in the first year with and without

 

 

 

 

 

 

 

 

 

 

 

 

 

 

reinfestation.
1800
1600 ,5 ',
1400 v". “-
1200 :" A ‘.
3 Reinvasiql I / \.
a .
'6 800 1.
" / \‘
.. / r
200 /
0
CO v- (D v- ‘D v- v- 0 v- CD v- v- (D v- (D '- CD '0- 0 v- v-
rero~e§99292558555858558
Days

 

Figure 22. Mite population development without re invasion or when 300 mites are

introduced in one day by robber bees at the end of August.

 

 

112

(1994) and Buchler (1994) listed the important traits involved in resistance of honey bees
against Varroa colonies, namely, grooming, brood removal, brood attractiveness, infertility

of mites, and post-capping period.

In these model simulations we attempted to evaluate some of the resistance traits
mentioned above and to compare the results between some of the published parameters to

see their effects on the mite population dynamics.
Grooming and hygienic behavior

Literature review. In the Asiatic honeybee Apis cerana F, the original host of V.
jacobsoni, the infestation rate remains at low levels and the parasite does not severely harm
the colony, and that may be due to the number of defense mechanisms that these bees have
developed. One of these mechanisms, as described by Peng et al. (1987), is the active
removal of adult mites from the bodies of worker bees. This process involves self-cleaning
behavior. After showing signs of irritation, the bee performs a grooming dance, and then
nestrnate cleaning and group cleaning behavior occur. This resulted in removing (within two
hours) more than 99% of mites added to colonies. Only 0.3% of the mites were removed by
grooming in colonies of A. mellifera. Buchler et al. (1992) also compared grooming in A.
cerana and A. mellifera and found successful mite removal in 75% of the cases in A. cerana.
In A. mellifera, 48% of the mites were removed by grooming. Fries et al. (1996) reported
lower numbers in full-sized colonies of A. cerana, 56% of 220 mites were removed by the
bees in 6h and, of those, 30% were damaged; results for A. mellifera colonies were 21% of

280 mites and12%, respectively.

Ruttner and Hanel (1992) examined the natural mortality of five A. m. camica

113

colonies for about one year and found on average 26% of the mites collected from inserts
showed injuries to the legs but rarely to the cuticle of the idiosoma. Moretto er al. (1991)
reported that 5.75% of the mites were removed by A. m. ligustica bees within 30 min after
infestation, and an average of 38% (ranging from 10-70%) were removed by Africanized

hybrids of A. mellifera bees.

Hygienic behavior of bees was described by Rothenbuhler (1964) in relation to
resistance against American foulbrood (Bacillus larvae). The hygienic behavior of bees
against Varroa mites was observed in both A. cerana and A. mellifera. Rath and Drescher
( 1990) showed that the detection, uncapping, and removal of Varroa-infested worker brood
cells by the A. cerana worker, artificially-infested worker brood cells were removed to 98%
within 5 days. Boecking and Drescher (1990) reported that artificially infested worker brood
cells were detected, uncapped and removed to various degrees and they showed that brood
cells infested with one Varroa mite were rejected from 14.3 to 95.8% of the time. Those with
two Varroa from 25.0 to 100% were found after ten days. Boecking and Drescher (1991)
reported that the removal of brood cells infested with one mite in Am. camica was 5.5%
(minimum) up to 95.8 (maximum). Within the same colonies, brood cells infested with two
Varroa mites showed a removal rate from 4.8% (minimum) to 100.0% (maximum). In
another study, Boecking and Drescher (1992) reported the removal rate was 24-41% of cells
containing 1-2 mother mites after 10 days. Removal rates 15.4% versus 10.9%; 2mite/cell:
removal rates 41.9% verses it was found that the rate was higher from other colonies, 32.2%
(Boecking and Drescher (1994). Boecking and Ritter (1993) found workers in 15 test A. m.

intermissa colonies detected and removed up to 75% of artificially infested brood and

114

removed up to 97-99% of freeze-killed brood in each of two trials.

Moosbeckhfer (1992) reported a significant negative correlation between the number
of damaged mites found in the period between August/September, and the infestation of

brood and bee samples and the total infestation found in a field test with 111 colonies.

Model Simulation. The model was used to simulate the effect of hygienic behavior
on the overall mite population. When workers opened the infested cells, some of the mothers
will be killed (Boecking and Drescher, 1991); but, more importantly, the most serious impact
will be on the reduction in offspring. Figure 23 illustrates the impact of a 5, 3, and 1%
reduction of number of offspring in worker and drone cells, as a result of cell uncapping by

workers, on the overall mite population.
Infertility and fecundity

Literature review. Koeniger et al. (1981 ; Anderson,l994) reported that Varroa
jacobsoni does not normally reproduce when it infests worker brood of the Eastern honey
bee, Apis cerana, extensive mite reproduction occurs only on drone brood. In A. mellifera
only a portion of the Varroa mites do not reproduce. Blum (1989) reported mite reproduction
rates in worker cells at 88.7 %, Fuchs and Langenbauch (1989) recorded 92.7% and Buchler
(1990) found 86.6% reproduction in worker brood cells. Sulimanovic et al. (1982), Schultz
(1984), Mossbeckhofer et al. (1988), Ifantidis (1990) and Boot et al. (1995) reported
infertility rates in worker cells of 13, 16, 7, and 8-12% respectively. In drone cells, Fuchs
and Langenbach (1989), Ifantidis (1984), and Schultz (1984) also found lower levels of
infertility in drone cells. They reported 8, 4, and 5%, respectively. Ritter and De Jong (1984)

observed only 43% of the mites in A.m. ligustica in worker cells in South America to be

 

 

2000
1800
1600
1400
1200
1000

800

t of Mites

200

 

lav-[~00
#60me
FP N

PF

1
1 7
33
49
65
81
97
1 13
1 29
225
241
257
273
289
305
321
337
353

Days

 

 

Figure 23. Mite p0pulations in the first year when starting with 10 mites in April
and when having different percentages reduction of offspring per worker
and drone cell as the result of hygienic behavior.

116

fertile. Marcangeli et al. (1992) estimated that, depending on season, between 56% and 72%
of the mites in A. m. ligustica colonies were fertile. Rosenkranz and Engels (1994) compared
Africanized and European colonies of A. mellifera, and found less than 40% of females mites
were fertile in Africanized bees, whereas, in European bees between 80-90% were fertile. In
another study by Aumeier et al. (1996) the percentage of infertile mites was 49% in

Africanized bees comparing to only 17% in camica colonies.

Ritter and De J ong (1984) found that in tropical Brazil fewer (ZS-60%) female mites
reproduce when entering brood cells than in Germany and Turkey where the rate was about
80%. Camazine (1986) reported that in European bee colonies in Brazil, 75% of infested
brood cells have immature mites compared with only 49% of infested brood cells in colonies

of Africanized bees.

Ruttner (1984) reported that colonies in Uruguay could resist Varroa infestation
without any treatment; this was attributed to the very low fertility rate (IO-30%) of the mites
in worker cells. In Tunisia, Ritter (1990) reported that comparatively low numbers of fertile
mites reproduced (SO-80%). This demonstrates the increased tolerance of A. m. intermissa

to Varroa.

The number of offspring produced by each mother Varroa mite entering the brood
cell is an important factor effecting the mite population, those numbers depend on the type
of brood they are, whether it is worker or drone cells. Schulz (1984) measured offspring
production by fertile female mites in worker brood as 1.8 (including infertile mites 1.6) and
drone cells the reproductive factor was 2.7. Fuchs and Schultz (1984), Ifantidis (1984), Fuchs

and Langebach (1989), and Martin (1994, 1995) reported 1.82 and 2.69, 1.33 and 2.77, 1.69

117

and 2.76, and 1.45 and 2.2 in worker and drone cells, respectively.

Model Simulation for fertility. The model can be used to simulate the impact of
different fertility rates on the total mite population obtained from worker and drone cells.
This particular simulation is done utilizing worker cells. The fertility rates chosen for this
simulation were taken from the literature. They include 70, 80, 85 and 95% fertility rates.
In the first year of infestation, mite populations reached 277, 803, 2154 and 3359 for the
above fertility rates, respectively (Figure 24). This clearly illustrates the important role that
fertility rate plays in the overall population dynamics of the mite.

These various fertility rates corresponded to differences in the rate of colony collapse
(Figure 25). Fertility rates of 80% and above caused colony collapse within two years, with
the earliest collapse associated with the highest fertility rate. A fertility rate of 70% caused
the population to peak in August of the third year, with colony collapse in the spring of the
fourth year.

Model Simulation for fecundity. The model was used to simulate the impact of
number of offspring on mite population dynamic. Figure 26 illustrates the impact of small
changes in the average number of offspring produced on the overall mite population
dynamics. It can be seen, that even small changes in average number of offspring, can have
a huge impact on the mite population, with great repercussions on the health of the bee
colony. For example, change from 1.5 to 1.6 offspring can cause the mite peak population

to increase from 6,000 to 10,000 mites in one year period.

118

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3500
/"
/ \
3000 .1 ¥
I \
2500 l’ \
95% , \\
8 2000 \L’ 1 1
g 90% ‘7 ,. ~, \
I- \ I o \
1: 1500 ’ .' ‘. ‘~
BOO/o N} x \
1000 \II" A “.\
70% [y \ . .
x’" “.
500 n,-
’I' \j
0
’99888588828‘5813‘8958
1- v- v- v- v- N N N N N ('0 0') (0
Days

 

 

 

Figure 24. Mite populations when female mites have different fertility rates in
worker cells

 

50000
45000
40000
35000

25000
20000
1 5000
1 0000

5000

l of Mites

 

Days

 

 

Figure 25.Mite population development until the colony collapse when Varroa
females have different fertility rates.

 

119

 

ii of Mites

1 0000
9000
8000
7000
6000
5000
4000
3000
2000
1 000

 

17
33

49

65

81

97

113

129

145
161

g 177

193
209
225
241

’8

257

273
289

305
321

337

353

 

Figure 26. Mite population development in one year with different average number
of offspring.

 

120

Post-capping period

Literature review. The length of the bee brood post-capping period will efi‘ect the
number of offspring produced by Varroa jacobsoni and it is one of the traits that is
considered when attempting to find a stock of honeybees tolerant or resistant to Varroa Post-
capping period has been studied in difi‘erent races of Apis meIli era.

The sealed brood stage of A. cerana lasted 10.9 days. The whole development is 18.5
days for the worker (Dung et al., 1993); and 22.8 days for the drone (Tan et al., 1993). Dung
et al. (1993) observed that A. mellifera has slightly longer developmental times of 19.4 and
24 days, respectively, in their research in the same region.

Mortis and Hanel (1984) and Mortis (1995) showed that the postcapping period of
the workers of A. m. camica ranged between 11.7 and 12.5 days. Subousbo (1986) measured
maximum difl‘erences ofup to 1.15 days forthe duration ofthe capped stage ofworker brood
in A. m. Iigusfica colonies. Buchler and Drescher (1990) observed differences in the length
of the capped stage of up to 9h between different strains (A.m. camica, A.m. mellifera and
Buckfast) and up to 19 h within individual colonies.

Africanized bees in South America have an 11 day average post-capping period for
worker and pupae. This allows only one or two female ofl‘spring of Varroa per female nrites
(Camazine, 1988); whereas, mites on European honey bees, with a mean post-capping period
of 12 days, can each produce up to three ofi‘spring (Rehm and Ritter, 1989; Schulz, 1984).
The mite caused little damage to A. m. ccpensis (cape bee) because the brood stage of this
species lasts 9 days, approximately two days shorter than A.m. camica (Hanel, 1984). In this

sub species the mite reproduction is restricted in worker brood. Only 21% of the mites can

121

produce even 1 fertile offspring, and the other 79% are unlikely to produce any viable
offspring (Moritz and Mautz, 1990).

The post-capping period of the drone is approximately the same for European and
Africanized bees, approximately 14 days (Berthof, Jay 1963,Wiese 1972). Selection for a
shorter post-capping stage may be a possible way to achieve resistance, particularly in the
light of its high heritability (Moritz, 1985).

Buchler and Drescher (1990) determined the correlation of both strain and season:
for 21 colonies, checked twice, a positive correlation between the length of the capped stage
and the infestation level was estimated (=0.48) indicating that, on average, a reduction of the
length of capped stage by one hour led to an 8.7% reduction in the final mite population
level.

Model Simulation. The next simulation focused on the effect of changing
postcapping periods. The data used in this simulation is roughly adapted from Martin (1977)
as described in the above introduction. Differences in postcapping can have a major impact
on mite population dynamics in the first year of infestation (Figure 27); as well as throughout
the life of the colony (Figure 28). As the post-capping period increases, the number of
offspring increase, and the colony collapses at a faster rate. If the post-capping period can be
reduced to ten days, then the mite impact on the bee population will be minimal (Figure 29).
Reproductive cycle

Literature review. Ruijter (1987) artificially transferred mites from cell to cell and
found that the mother mite is able to reproduce as many as seven times. Schulz (1984)
reported that 78% of the mites reproduce only once and 22% reproduce twice. Mikityuk et

al. ( 1976) observed an additional reproduction cycle, he stated that 78% of Varroa produce

122

 

 

 

 

 

 

 

 

 

 

 

 

 

12days I ‘.
1000 \N: 1‘
11.5days . /'\ ‘1
800 . ' 1 \.
= a, I ’0'. ‘
z 600 11days I I .- .~ ‘\
. 7' .r \ f 1
/' ’.v' ‘. \ .
400 10.5 days - L/x ‘.\\‘
//§1v' “\ ‘
,;4.' .\\
“ #__ ‘\\‘
200 l“ T
1" ~\
0
'998885888zg‘31‘38‘358
'- v- v- v- v- N N N N N t') 1') ('3
Days

 

Figure 27. Mite population development in the first year when worker brood

have different post-capping periods.

 

 

45000
40000
35000
30000
25000
20000
1 5000
1 0000

Mites

10.5 days

 

78
155
232
309
463
540
617
694
771
848
925

1002

1079

1156

1233

1310

1387

1464

1541

1618

1695

1772

386

Days

 

Figure 28. Mite population development until the colony collapse when worker

brood have different postcapping periods.

 

 

123

 

 

60000

50000

20000

1 0000

 

 

 

 

 

 

 

 

 

70
100‘ A w. 60
; Mite Bees

' z/ «50
l\‘-. ll ll ll /\ 4°
N 11 /\ /\ l\3°
: \J \120
...... 10
-------- ‘s.....-o-' ‘ ~---------'.."".‘ 0

“895598885558358359358

Days

Mites

 

Figure 29. Mite and bee populations development when worker brood have

10 day post-capping period.

 

124

only once, 18% produce twice and 4% produce three times for an overall mean number of
1.26 reproductive cycle per female mite. Mikityuk (1979) and Grobove (1977) stated that
1.9% and 4% of the mite reproduce four times respectively. Fries and Rosenkranz (1993)
reported that 13% of the mites reproduce three times, their sample was 475 mother mites.
Wended and Rosenkranz (1993) found that 4.1% of the mites produce three times with an
overall mean of 0.88 reproductive cycle per mother mites. Recently, Fries and Rosenkranz
(1996) using full-size colonies reported that under optimal conditions the mean number of

reproductive cycles by Varroa is greater than 1.5 but less than 2.

Model Simulation. The importance of the reproductive cycle can also be easily
simulated (Figure 30). The parameters chosen for the simulation were suggested by Fries
(1994, 1996) and Engles (1994). The mite population will almost double if the reproductive
cycle is increased by 0.2.

Phoretic periods

Literature review. After emergence from the brood cells, the female mites reside
a certain period on adult bees in the colony before they invade new brood cells (Boot et al.,
1993; Boot et al., 1994). Boot et al., (1996) stated that the length of this period strongly
affects the population dynamics of the mites, because mites cannot reproduce while they
reside on adult bees and therefore reproduction is delayed. In addition, the period on adult
bees may affect the population dynamics of the mites since some of the mites will die during
their stay on adult bees and also it might reduce the number of offspring per mother mite
(Beetsma and Zonneveld, 1992). Thus fitness of mites increases by minimizing their stay on

adult bees (Boot and Calis, 1993). Schulz (1984) reported after a phase of 1-20 days (44%

125

 

 

Mites

60000

50000

40000

30000

20000

10000

 

9

 

.p-IICI"'

 

 

 

 

 

 

 

 

 

D
:WLG 1.2
'I L18
:, V11"
-: 1.4 ’.
s. I
'v L 4
:. 7
n ' l‘
. 11
'. ’ l
é“~—-'—" \
Pmmhmwmmnmwmmhmwmmnmwomh
VQNCDv-mmt‘OBNCDOEQBhv-mmngtD
v—v-NNNO‘JODVVIO ID ‘0th 0050:
Days

 

Figure 30. Mite population development when female mites have different

reproductive cycles.

 

126

within 6 days) on adult honeybees, the female mite enters the brood cell for reproduction. He
found that the phoretic period was 4.5 days for old mites, 10.7 days for young mites and 7.4
days for a mixed population of mites. (Boot et al., 1993; Boot et al., 1994) stated that during
the brood rearing the mean residence time of mites on adult bees is maximally 1-3 weeks,
depending on the number of brood cells available for mite invasion. Woyke (1987) reported
that mites have an average phoretic period of 4.7 days in a mixed population and 5.9 days

for young mites. Grobov (1977) reported a range of 4-13 days.

Mites begin to invade brood cells during limited periods preceding cell capping with
a fairly constant rate until cells are capped, about 50 and 20 hours for drone and worker
respectively ( Ifantidis, 1988; and Boot et al. 1992).

Model Simulation. The effect of different phoretic periods can also be illustrated
with this model (Figures 31, 32). The phoretic periods chosen were 4.9, 6.9, 8.9, and 10.9
days. These correspond to a maximum population of 1883, 991, 577, and 376 in the first
year of infestation, respectively. The longer the phoretic period, the less of a population
increase (Figure 31). Running the simulation over a number of years, it was found that
colonies with an average phoretic period of 4.9 and 6.9 days collapsed in the second year,

while those with longer periods (8.9 and 10.9 days) did not collapse until the third year

(Figure 32).

DISCUSSION
The model is used to evaluate some of important traits involved in resistance or

tolerance of honey bees against Varroa colonies such as grooming, brood removal, brood

127

 

 

 

2000
1800
1600
1400
1200
1000
800
600
400
200

it of mites

 

Days
15
36
57
78
99

120
141

162
1 83
204
225
246
267
288
309
330
351

Days

 

 

Figure 31. Mite population development in the first year when female mites have
different phoretic periods.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40000
35000 IJL
30000 8.9days\
4.9 03 s I
a 25000 \\y/"1 I
a
:5: 20000 J
o 6.9 da 5 I
as 15000 J Y . 10.9 days
j -7 / , 1/
10000 '. .' 4‘
/ :0 a. / .I \
5000 a". V.o" L :\
0 ..e. ‘..t..‘. ........ ’ 0.9..
v- C m 00 N (D ID V ('3 N 1- O O) m h (D ID 3 0') N P O O) Q
mmifi§8§83$8$88§§m888§§§
Days

 

 

 

Figure 32. Mite populations development until the colony collapse when female
mites have different phoretic periods.

128

attractiveness, infertility of mites, and post-capping period, and reproductive cycle. The
results obtained indicate that the model is a valuable tool for predicting the changes in mite
and bee populations resulting from varying a single factor concerning the above mentioned
traits. The values were obtained from research literature when available.

For long term solutions to the Varroa problem, priority should be given to research
on genetic improvement of colony defense mechanisms (Buchler, 1994). Since selection
program for such mechanisms will take a very long time, it is important to concentrate on the
most effective resistant trait that effects the mite population. The results from the model
simulations illustrate the effect of these traits on the bee and mite populations and it showed
that any thing that effected mite reproduction such as number of offspring, fertility,
reproductive cycle and phoretic period are the most important traits on which to concentrate.

Treatment efficiency and time of treatment were also simulated in this model and the
results show that one treatment is sufficient to control the mites for two years at least if there
is no reinvasion from outside and if the beekeepers follow the label concerning length of
treatment. However, reinvasion is inevitable in the field and that explains the failure of the
control measure and colony death even when treated once a year. This shows how important
it is for beekeepers to try to reduce reinvasion by cooperating with each other in the same
area, by treating in the same season, and by including in their management practices help in
reducing drifting, robbing, swarming, etc.

Controlling the mites by means of non chemical measures is labor intensive, time
consuming and is not practical for large operation beekeepers. Although, intensive drone
removal reduce high portions of the mite population, still it is not successful in controlling

the mite the second year when the population becomes high.

129

Trapping comb and hyperthermia techniques alone will not work for more that two
years in high or moderately high p0pulation colonies as the simulations indicate and these
results are in agreement with Maul (1988) who reported that the technique is good for two
or three years, but reported that mite population levels continue to increase after that time
frame. Fries et al. (1993) stated that the trapping comb technique might be insufficient for
controlling mites in all colonies over a period of several years due to the high mortality
during the third and fourth experimental year.

The conclusion is that one biological control technique alone would not be sufficient
to control the mite but when selecting for bees that have some traits that reduce mite

populations, naturally biological control might work.

LITERATURE CITED

Anderson, DL (1994): Non-reproduction of Varroa jacobsoni in Apis mellifera colonies in
Papua New Guinea and Indonesia. Apidol. 25, 412-421.

Appel, H; Buchler, R (1991): Heat-treatment of brood comb for Varroa control. Apidol. 22
(4), 471.

Artemenko, L; Skyrypnik, B; Sabadin, B (1978): Epizootologiia Varroatoza pchel (cited in
Smimov 1978). Pchelovod. 6, 9.

Avitabile, A (1978): Brood rearing in honeybee colonies from late autumn to early spring.
J. Apic. Res. 17 (2), 69-73.

Barbina, MT; De-Paoli, M; Valentino, A (1990): Determination of tau-fluvalinate residues
in honey. Pestic. Sci. 28(2), 197-202.

Beetsma, J; Zonneveled, K (1992): Observations on the initiation and stimulation of
oviposition of the Varroa mite. Exp. & Appl. Acar. 16(4), 303-312.

Blum, R (1989): Einfluss einer Unterschiedlichen proteinernahrung von Honigbienen aufdie
Reproduktion der hamophagen Milbe Varroa jacobsoni. Diplomarbeit Thesis,
Fakultat Biologie, Universitat Tubingen, Germany. 51pp p.

130

Boecking, O; Drescher, W (1990): The reaction of different Apis mellifera colonies to
Varroa infested brood cells. In: Proceedings of the international symposium on recent
research on bee pathology, Septemper 5-7 1990, Gent, Belgium. (Ed: Ritter), 41—42.

Boecking, O; Drescher, W (1994): [Rating of signals that trigger Apis mellifera L. bees to
remove mite-infested brood.] Bewertung von Ausloesefaktoren des
Verhaltensmerkmals Ausraeumen milbeninfrzierter Brut bei Apis mellifera L. Apidol.
25(5), 459-461.

Boecking, O; Drescher, W (1991): Response of Apis mellifera L. colonies infested with
Varroa jacobsoni Oud. Apidol. 22, 237-241.

Boecking, O; Drescher, W (1991): The reaction of worker bees in different Apis mellifera
colonies to Varroa infested brood cells. In: Proceedings of the International
Symposium on Recent

Boecking, O; Ritter,W (1993): Grooming and removal behavior of Apis mellifera intermissa
in Tunisia against Varroa jacobsoni. J. Apic. Res 32 (3/4), 127-134.

Boot, WJ; Calis, JNM; Beetsma, J (1992): Differential periods of Varroa mite invasion into
worker and drone cells of honey bees. Exptl. & Appl. Acar. 16(4), 293-301.

Boot, WJ; Calis, JNM; Beetsma, J ( 1993): Invasion behavior of Varroa mites into honey bee
brood cells. In: Asian Apiculture. (Eds: Connor,LJ; Rinderer, TE; Sylvester,HA;
Wongseri,S) Wicwas Press, Cheshire, Connecticut, 491-498.

Boot, WJ; Calis, JNM; Beetsma, J (1995): Does time spent on adult bees affect reproductive
success of Varroa mites? Entomol. Exp. Appl. 75(1), 1-7.

Boot, WJ; Sisselaar, DJA; Calis, JMN; Beetsma, J (1994): Factors affecting invasion of
Varroa jacobsoni (Acari: Varroidae) into honeybee, Apis mellifera (Hymenoptera:
Apidae), brood cells. Bull. Entomol. Res. 84, 3-10.

Boot, WJ; Schoenmaker, J; Calis, JNM; Beetsma, J (1995): Invasion of Varroa jacobsoni
into drone brood cells of the honey bee, Apis mellifera. Apidol. 26(2), 109-118.

Buchler, R; Drescher, W (1990): Variance and hertiability of the capped developmental stage
in European Apis mellifera L. and its correlation with increased Varroa jacobsoni
Oud. infestation. J. Apic. Res 29(3), 172-176.

Buchler, R (1994): Varroa tolerance in honey bees - occurrence, characters and breeding.
Bee World 75(2), 54-70.

Buchler, R; Drescher, W; Tomier, I (1992): Grooming behavior of Apis cerana, Apis
mellifera and Apis dorsata and its effect on the parasitic mite Varroa jacobsoni and
Tropilaelaps clareae. Exp. & Appl. Acar. 16 (4), 313-319.

131

Calis, J; Beetsma, W; Van den, E; Ruijter, A (1993): Control of the Varroa mite by treatment
of sealed honeybee brood with formic acid. Proceeding Experimental & Applied
Entomology 4, 217-222.

Camazine, S (1988): Factors affecting the severity of Varroa jacobsoni infestations on
European and Africanized honey bees. In: African Honey Bees and Bee Mites. (Eds:
Needham,GR; Page,RE,Jr; Delfinado—Baker,M; Bowman,C) Ellis Horwood,
Chichester, 444-451.

Colobmbo, M; Lodesani, M; Spreafico, M (1994): (Resistance of Varroa jacobsoni to
fluvalinate. Preliminary results of investigations conducted in Lombardy).

Apiculture abstracts 45 No. 3, 275.

DeGrandi-Hoffman, GD; Roth, SA; Loper, GM; Erickson, EH, Jr (1989): BEEPOP; a
computer simulation model of honey bee colony population dynamics. Amer. Bee J.
127(12), 846-847.

Engels, P; Rosenkranz, P (1992): Hyperthermie-Erfahrungen bei der Varroatose-Volekem
in sommer. Apidol. 23, 379-381.

Fries, I; Aarhaus, A; Hansen, H; Korpella, S (1991): Comparison of diagnostic methods for
detection of low infestation levels of Varroa jacobsoni in honey-bee (Apis mellifera)
colonies. Exptl. & Appl. Acar. 10, 279-287.

Fries, I; Aarhus, A; Hansen, H; Korpela, S (1991): Development of early infestations by the
mite Varroa jacobsoni in honey-bee (Apis mellifera) colonies in cold climates. Exptl.
& Appl. Acar. 11(2-3), 205-214.

Fries, I; Rosenkranz, P (1993): Number of reproductive cycles of the Varroa mite. Apidol.
24 (5), 485-486.

Fries, I; Rosenkranz, P (1996): Number of reproductive cycles of Varroajacobsoni in honey-
bee (Apis mellifera) colonies. Exp. & Appl. Acar. 20(2), 103-112.

Fries, I; Camazine, S; Sneys, J (1994): Population dynamics of Varroa jacobsoni: A model
and a review. Bee World 75 (1), 5-28.

Fries, I; Hansen,H (1993): Biotechnical control of Varroa mites in cold climates. Amer. Bee
J. 133(6), 435-438.

Fries, I; Rosenkranz, P (1996): Number of reproductive cycles of Varroa jacobsoni in honey-
bee (Apis mellifera) colonies. Exp. & Appl. Acar. 20(2), 103-112.

132

Fuchs, S (1992): Choice in Varroa jacobsoni Oudemans between honey bee drone or worker
brood cells for reproduction. Behav. Ecol. Sociobiol. 31(6), 429-435.

Fuchs, S; Langenbach, K (1989): Multiple infestation of Apis mellifera L. brood cells and
reproduction in Varroa jacobsoni Oud. Apidol. 20, 257-266.

Fukuda, H; Sakagami, SF (1968): Worker Brood survival in honeybees. Res. P0pul. Ecol 10,
31-39.

Greatti, M; Milani, N; Nazzi, F (1992): Reinfestation of an acaricide-treated apiary by
Varroa jacobsoni Oud. Exp. & Appl. Acar. 16(4), 279-286.

Grobov, O (1977): Varroasis In Bees. In: Varoasis, a honeybee diseases. Apimondia
Publishing House, Bucharest,, 46-69.

Hansen, H; Guldborg, M (1988): Residues in honey and wax after treatment of bee colonies
with formic acid. Tidsskr. Planteavl. 92(1), 7-10.

Harris, JL (1985): A model of honeybee colony population dynamics. J. Apic. Res. 24(4),
228-236.

Hoppe, H; Ritter, W (1986): [The possibilities and limits of thermal treatment as a
biotechnical method of fighting Varroatosis] Moeglichkeiten und Grenzen der
Therrnobehandlung als biologisches Bekaemfungsverfahren gegen die Varroatose.
Apidol. 17(4), 374-376.

Howell, D; Usinger, R (1933): Observation on the flight and length of life of drone bees.
Annals of the Entomological Socity of America 26, 239-246.

Huttinger,E; Pechhacker,H; Sulimanovic,D (1981): Spread of Varroa jacobsoni from one
colony to another. Apiacta 16 (2), 71-76.

Ifantidis, M (1988): Some aspects of the process of Varroa jacobsoni entrance into honeybee
Apis mellifera brood cells. Apidol. 23 (4), 227-233.

Ifantidis, MD (1984): Parameters of the population dynamics of the Varroa mite on
honeybees. J. Apic. Res. 23(4), 227-233.

Ifantidis, MD (1990): Reexamination of reproduction parameters of the mite Varroa
jacobsoni Oudemans, .

lrndorf, A; Kilchenmann, V (1993): (Varroa invasion-a surprise for beekeepers. Apiculture
abstracts 44 No. 1, 58.

Jay, SC (1963): The development of honeybees in their cells. J. Apic. Res. 2, 117-134.

Jay, S (1966): Drifting of honeybee in commercial apiaries. II. Effect of various factors
when hives are arranged in rows. J. Apic. Res. 4 (3), 103-112.

133

Koeniger, N; Koeniger, G; Wijayagunasekaran, H (1981): Beobachtungen uber die
Anpassung von Varroa jacobsoni an ihren ursprunglichen Wirt Apis cerana in Sri
Lanka. Apidol. 12 (1),37-40.

Koeniger, N; Fuchs, S (1988): Control of Varroa jacobsoni: current status and
developments. In: African Honey Bees and Bee Mites. (Eds: Needham, GR; Page,
RE, Jr; Delfinado-Baker, M; Bowman, C) Ellis Horwood, Chichester, 360-369.

Koeniger, N, Koeniger, G; Wijayagunasekaran, H (1981): Beobachtungen uber die
Anpassung von Varroa jacobsoni an ihren ursprunglichen Wirt Apis cerana in Sri
Lanka. Apidol. 12 (1),37-40.

Korpela, S; Aarhus, A; Fries, I; Hansen, H (1992): Varroa jacobsoni Oud. in cold climates:
Population growth, winter mortality and influence on the survival of honey bee
colonies. J. Apic. Res 31(3—4), 157-164.

Kraus, B; Page, RE, Jr (1995): Population growth of Varroajacobsoni Oud in Mediterranean
climates of California. Apidol. 26, 149-157.

Kulincevic, JM; Rinderer, TE (1988): Breeding honey bees for resistance to Varroa
jacobsoni: analysis of mite population dynamics. In: African Honey Bees and Bee
Mites. (Eds: Needham, GR; Page, RE, Jr; Delfinado-Baker, M; Bowman, C) Ellis
Horwood, Chichester, 434 —443.

Kustermann, T (1990): P0pulationsstruktur der varro-milbe in Arbeiterinneneme Deutsches
Irnker- Journal 1 (1 1), 436-437.

Lodesani, M; Colombo, M; Spreafico, M (1995): Ineffectiveness of Apistan registered
treatment against the mite Varroa jacobsoni Oud in several districts of Lombardy
(Italy). Apidol. 26(1), 67-72.

Marcangeli, JA; Eguarae, MJ; Fernandez, NA (1992): Reproduction of Varroa jacobsoni
(Acari: Mesostigmatzu Varroidae) in temperate climates of Argentina. Apidol. 23, 57-
60.

Marcangeli,J; Eguaras,M; Femandez,N (1995): Population growth of Varroa jacobsoni
(Gamasida: Varroidae) in colonies of Apis mellifera (Hymenoptera:Apidae) in
temperte climates using camazine’s model. Apiacta 30 (1), 13-19.

Martin, SJ (1994): Ontogenesis of the mite Varroa jacobsoni Qud. in worker brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 18(2), 87-
100.

Martin, SJ (1995): Ontogenesis of the mite Varroa jacobsoni Oud. in drone brood of the
honeybee Apis mellifera L. under natural conditions. Exp. & Appl. Acar. 19(4), 199-
210.

134

McLellan, AR; Rowland, CM; Fawcett, and RH (1980): A monogynous eusocial insect

worker population model with particular reference to honeybees. Insects Sociaux 27,
305-31 1.

Mikityuk, V (1979): [Reproductive capacity of female Varroa mites]. Pchelovod. 9, 21.

Mikityuk, V; Korzjhova, L; Sedin, I (1976): [Experiments on the biology of the Varroa
mite]. Pchelovod. 12, 19-20 (in Russian).

Mikityuk,V (1979): [Reproductive capacity of female Varroa mites]. Pchelovod. 9, 21.

Milani, N (1994): Possible presence of fluvalinate-resistant strains of Varroa jacobsoni in
northern Italy. In: New perspectives on Varroa. (United Kingdom) (Ed: Matheson,A)
International Bee Research Association, Cardiff, 87.

Mobus, B; Connor, L (Eds.) (1988): The Varroa Handbook. Northern Bee Books, West
Yorkshire, U. K. 52 pages.

Moosbeckhofer, R (1991): Varroaverluste wahrend der uberwinterung. Bienenvater 112 (9),
300-

Moosbeckhofer, R; Fabsicz, M; Kohlich, A (1988): Untersuchungen gber die Abhangigkeit
der

Moretto, G; Concalves, S; De Jong, D (1991): Africanized bees are more efficient at
removing Varroa jacobsoni - Preliminary data. Amer. Bee J. 131, 434.

Moritz, RFA (1985): Heritability of the postcapping stage in Apis mellifera and its relation
to varroatosis resistance. J. Hered. 76(4), 267-270.

Moritz, RFA; Hanel, H (1984): Restricted development of the parasitic mite Varroa
jacobsoni Oud. in the Cape honeybee Apis mellifera capensis Esch. Z. Angewandte
Entomol. (J. Appl. Entomol.) 97(1), 91-95.

Moritz, RFA; Jordan, M (1992): Selection of resistance against Varroa jacobsoni across
caste and sex in the honeybee (Apis mellifera L., Hymenoptera: Apidae). Exp. &
Appl. Acar. 16(4), 345-353.

Moritz, RFA; Mautz, D (1990): Development of Varroa jacobsoni in colonies of Apis
mellifera capensis and Apis mellifera camica. Apidol. 21(1), 53-58.

Muller, M (1987): Befallsentwicklung der Milbe Varroa jacobsoni den Wintermonaten.
(summary). Allg. Dtsch. Irnkerztg. 2, 6-11.

Mussen, E (1993): Keeping Varroa Under control. Extension Apiculture University of
Califomia., 1-6.

Nachkommensrate von Varroa jacobsoni Ouf. vom Befallsgrad der Bienenv‘lker. Apidol.
19, 181-208.

135

Nolan, W (1928): Seasonal brood-rearing activity of the Cyprian Honeybee. J. Econ.
Entomol. 21, 392-401.

Nolan, WJ (1925): The brood-rearing cycle of the honeybee. USDA Dept. Bull. No. 1349.
(55 PP)

Omholt, S; Cralisheim, K (1991): The possible prediction of the degree of infestation of
honeybee colonies Apis mellifera by Varroa jacobsoni Oud. by means of its natural

death-rate: a dynamic model approach. Norwegian Journal of Agriculture Sciences
5, 393-400.

Otten, C (1991): Reproduction and population dynamics of Varroa jacobsoni Oud. in
colonies of Apis mellifera L. of different origin. In: Proceedings of the International
Symposium on Recent Research on Bee Pathology - Gent, Belgium, Sept. 1990.
(Eds: Ritter,W; Van Laere,O; Jacobs,F; De Wael,L) Apimondia - Intemat. Fed.
Beekeepers Assoc., Rome, 67 -69.

Peng, Y-S; Fang, Y-Z; Xu, S-Y; Ge, L-H (1987): The resistance mechanism of the Asian
honey bee, Apis cerana Fabr., to an ectoparasitic mite, Varroa jacobsoni Oudemans.
J. Invertebr. Pathol. 49(1), 54-60.

Peng, YSC; Fang, Y; Xu, S; Ge, L; Nasr, ME (1987): Response of foster Asian honeybee
(Apis cerana Fabr.) colonies to the brood of European honeybee (Apis mellifera L.)
infested with parasitic mite, Varroa jacobsoni Oudemans. J. Invertebr. Pathol. 49(3),
259-264.

Rademacher, E; Geiseler, E (1986): "Die Varroatose der BienenzGeschichte, Diagnose,
Therapie.". In: (Eds: Verlag, S; Jeep),, .

Rath, W; Drescher, W (1990): Response of Apis cerana Fabr. towards brood infested with
Varroa jacobsoni Oud. and infestations rates of colonies in Thailand. Apidol. 21,
311-321.

Rehm, S; Ritter, W (1989): Sequence of the sexes in the offspring of Varroa jacobsoni and
the resulting consequences for the calculation of the developmental period. Apidol.
20, 339- 343.

Ribbands, CR (1953): The Behavior and Social Life of Honeybees. International Bee
Research Association, London. 352 pages.

136

Ritter, W; De J ong, D (1984): Reproduction of Varroa jacobsoni O. in Europe, the Middle
East and tropical South America. Z. Angewandte Entomol. (J. Appl. Entomol.) 98(1),
55-57.

Ritter, W; Leclercq, E; Koch, W (1984): [Observations on bee and Varroa mite populations
in infested honey bee colonies] Observations des populations (1’ abeilles et de Varroa
dans les colonies a differents niveaux d’ infestation. Apidol. 15(4), 389-399.

Ritter, W (1984): Neuester Stand der diagnostischen und therapeutischen Massnahmen zur
Bekampfung der Varroatose. Tierarztliche Umschau 39, 122-127.

Ritter, W (1988): Varroa jacobsoni in Europe, the tr0pics, and subtropics. In: African Honey
Bees and Bee Mites. (Eds: Needham,GR; Page,RE,Jr; Delfinado-Baker,M;
Bowman,C) Ellis Horwood, Chichester, 349-359.

Ritter, W (1990): Development of the Varroa mite population in treated and untreated
colonies in Tunisia. Apidol. 21(4), 368-370.

Ritter, W; Leclerq,E (1987): [Honeybee and varroa population in area of high and low bee
population densities]. Tierarztliche Umschau 42(7), 548-551.

Robaux, P (1988): Varroa jacobsoni: problems with diagnosis and control in Europe. In:
African Honey Bees and Bee Mites. (Eds: Needham, GR; Page, RE;
Delfinado-Baker, M;

Rosenkranz, P; Engels, W (1994): Infertith of Varroa jacobsoni females after invasion into
Apis mellifera worker brood as a tolerance factor against varroatosis. Apidol. 25(4),
402-41 1.

Rothenbuhler, WC (1964): Behavior genetics of nest cleaning in honey bees. IV. Responses
of F1 and backcross generations to disease-killed brood. Amer. Zoo]. 4, 11-123.

Ruijter, A, De (1987): Reproduction of Varroa jacobsoni during successive brood cycles of
the honey bee. Apidol. 18, 321-326.

Ruttner, F; Hanel, H (1992): Active defense against Varroa mites in a Camiolan strain of
honeybee (Apis mellifera camica Pollmann). Apidol. 23, 173-187.

Ruttner, FH; Marx, G (1984): [Observation about a possible adaptation of Varroa jacobsoni
to Apis mellifera L. in Uruguay] Beobachtung neber cine moegliche Anpassung von
Varroa jacobsoni an Apis mellifera L. in Uruguay. Apidol. 15(1), 43-62.

Schousboe, C (1986): The duration of closed cell stage in worker brood of Danish honey-
bees (Apis mellifera L.) in relation to increased resistance to the varroa mite (Varroa
jacobsoni Oud.). Tidsskr. Planteavl. 90(4), 293-299.

137

Schulz, AE (1984): Reproduction and p0pulation dynamics of the parasitic mite Varroa
Jacobsoni Oud. and its dependence on the brood cycle of its host Apis Mellifera.
Apidol. 15 (4), 401- 420.

Sulimanovic, D; Ruttner, F; Pechhacker, H (1982): Studies on the biology of reproduction
in Varroa jacobsoni. Honeybee Sci. 3, 109-112.

Weiss, K (1984): Bienen-patholgie (Summary). Franz Ehrenwirth Verlage,

Wendel, H; Rosenkranz, P (1990): Invasions- geschwindigkeit und Fertilitat von Varroa
Weibchen in aufeinanderfolgenden Reoproduktionszyklen. Apidol. 21 (4), 372-374.

Wienands, A (1988): Varroa research makes headway in West Germany. Amer. Bee J.
128(4), 265.

Woyke, J (1987): Infestation of honeybee (Apis mellifera) colonies by the parasitic mites
Varroa jacobsoni and Tropilaelaps clareae in South Vietnam and results of chemical
treatment. J. Apic. Res. 26(1), 64-67.

DISCUSSION

Deflinado—Baker and Houck (1989) suggested a lower virulence of mites from North
America compared to mites from Europe on the basis of the hypothesis of the South America
origin of the mite in the US. Our results agree with other research from the US (Page and
Kraus, 1995; Delaplane and Hood, in press 1997) and disagree with Deflinado-Baker and
Houck. In Georgia, Delplane and Hood (1997, in press) started with new packages of bees
which contained a small incipient population of V. Jacobsoni. The mite population increased
in June to 4271-110, in August to 3172:324 and to 6662:2127 in October. Kraus and Page
(1995) reported that in the Mediterranean climate of California the initial population is
capable of increasing 300-fold during one year, they started with 50 mites at the end April
and the mite population increased to 2,367 in October. When the researchers started with 50
mites at the end of October, the mite population increased to 1,620 in April. Our results show
that over the period of one summer, the mite population increased 81, 188 and 193-fold for
the groups that were infected with 5, 10 and 25 mites respectively, when using the sticky
board method of estimating mite p0pulation. This was an average of 154-fold increase. When
estimating the population from the live bee method, there was a 352, 225, and 125-fold

increase, respectively. This method had an average increase of 234-fold.

138

139

When comparing the mite population in US with that in Europe, the increase averages
about 10—fold per year (Ritter, 1984; Fries, 1991; Korpela et al., 1992); but can increase up

to 100-fold within one summer (Fries et al., 1991).

Comparing reproduction rates obtained in this study with other studies, it is found
that the mite reproduction rate of 86.8% in worker and 93% for drone broods was very
representative of what other researchers found. Blum (1989) reported mite reproduction
rates at 88.7%, Fuchs and Langenbauch (1989) recorded 92.7% and Buchler (1990) found
86.6% in worker brood cells. In drone cells, Fuchs and Langenbach (1989) also recorded
92.2% fertility rates in drone cells.

The percent of mites that did not reproduce (11.0% in workers and 7.1% in drones)
also was similar to figures presented by other researchers. In worker cells, Sulimanovic et
al. (1982), Schultz (1984), Moosbeckhofer et al. (1988), Fuchs and Langenbach (1989),
Ifantidis (1990) and Boot et al. (1995) reported infertility rates in worker cells of 13, 16, 7,
7, 14.1, and 8-12%, respectively. Fuchs and Langenbach (1989), Ifantidis (1984), and Schultz
(1984) found lower levels of infertility in drone cells also. They reported 8, 4, and 5%,
respectively.

Some of the mother mites produced only male offspring. This is probably because
they had not mated, since haploid eggs of Varroa mites develop into males (de Ruijter and
Pappas, 1983). Martin (1995) attributes this partly to the death of the male before he is able
to fertilize his sisters. He found in earlier studies, that 20% of the males died before they
mated in worker brood (1994) and 10% in drone cells (1995).

In this study, it was found that 5.1% of the offspring were male only in worker cells.

140

This was similar to other studies. Boot et al. (1995), Schultz (1994), Moosbeckhofer et al.
(1988) and Fuchs and Langebach (1989) reported a rate of 8-10, 6, 3 and 3%, respectively.
For drone cells, a rate of 3% was observed. Fuchs and Langebach (1989) reported a rate of
1%.

The other parameter that was measured was the percent of mother mites that died in
brood cells. It was found that 2.3% died in worker cells and 2.7% in drone cells . Only one
other researcher reported this statistic ( Martin, 1994). He found a higher percentage in
drones (7.7%), and a similar rate in workers (2%). He found that 32% of the deaths in drone
cells were caused by failure of the mite to emerge from the brood food and found they were
trapped in the cell wall. The percentage rose to 50% in worker cells.

The model simulates bee and mite populations and illustrates the effect of the bees
on the mite population and the effect of the mites on the bee population. When starting with
different initial infestation, 5, 10 or 20 mites, the population peaked at 690, 1,339 and 2,521
in the first year, and as the initial infestation increased, the population peaks early in the
season in the second year. The same trend will continue the next year until the damage
threshold is reached. It seems that the threshold depends on the ratio of mites to bees. If the
number of mites is high in proportion to bees, the bee population can not support the colony
and it will die.

The model is used to evaluate some of the important traits involved in resistance or
tolerance of honey bees against Varroa colonies such as, grooming, brood removal, brood
attractiveness, infertility of mites, post-capping period, and reproductive cycle. The results
obtained indicate that the model is a valuable tool for predicting the changes in mite and bee

populations resulting from varying a single factor concerning the above mentioned traits. The

141

values used as a parameter in the model were obtained from research literature when
available. The results from the model simulations illustrate the effect of these traits on the
bee and mite populations and it showed that any thing that effects mite reproduction such as
number of offspring, fertility, reproductive cycle and phoretic period are the most important

trait for which to select.