Research Collection

Doctoral Thesis

Distributional ecology and inbreeding in bumble bees

Author(s): Gerloff-Gasser, Christine U.

Publication Date: 2001

Permanent Link: https://doi.org/10.3929/ethz-a-004217724

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

Diss. ETH No. 14194

DISTRIBUTIONAL ECOLOGY AND INBREEDING

IN BUMBLE BEES

A dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY (ETH), ZURICH

for the degree of Doctor of Natural Sciences

presented by CHRISTINE U. GERLOFF-GASSER

dipl. Natw. ETH

born December 23, 1969

in Cheverly, Maryland, USA

accepted on the recommendation of

Prof. Dr. Paul Schmid-Hempel, examiner

Prof. Dr. Pekka Pamilo, co-examiner

Zürich 2001

To my family and its academic tradition

and the bumble bees of the world

Acknowledgements

A lot of people accompanied me during my Ph.D. time and supported me in many ways. A

big thank you to all of them although they might not be listed explicitely.

Prof. Paul Schmid-Hempel, my supervisor, provided the setting for making this Ph.D.

project an interesting and enjoyable experience (most of the time), both scientifically and

socially. I particularly appreciated learning and applying a broad spectrum of scientific

methods, working independently, and his patience when it came to analyzing the data and

writing up besides starting my new job as a teacher. Not forgetting the many conferences

and TMR workshops I attended and the sometimes adventurous ways to get there...

Prof. Pekka Pamilo, my co-referee, enabled an interesting and inspiring visit of his research

group in Uppsala, Sweden, and provided useful comments on this thesis.

Mark Brown had the bad luck to be working at Experimental Ecology the entire time I was

writing up. His comments strongly improved my thesis and my writing skills - which

makes the extra time I invested due to his exacting proofreading worthwhile. Furthermore,

he was helpful with any sort of problem, from stats to translating school preparations into

(British) English to explaining Gary Larson cartoons. Most important of all was his ability

to motivate and cheer me up by being a close friend, who enjoys dancing, ice cream, and e-

mail jokes as much as I do (and in addition - hélas! - singing and whistling

enthusiastically). :-)

Klas Allander, Sebastian Bonhoeffer, Claudie Doums-Traoré, Jukka Jokela, Pia

Mutikainen, Jackie Shykoff, Jouni Taskinen, and Jürgen Wiehn gave valuable scientific

input in many discussions and - especially the Finns - never ran out of suggestions for

improving the statistical analyses, i.e., running again another test.

Birgit Ottmer, "my" Diploma student, spent uncounted hours (including the holiday season)

with me in the climate chamber rearing and sharing anecdotes about our bumble bee

colonies. Thanks for the company and the interesting data! Jürgen Gadau made good use of

(some of) my samples filling the lab's freezers and gave permission to include the resulting

collaborative manuscript in this thesis. I enjoyed these collaborations very much.

Regula Schmid-Hempel introduced me to and helped with various techniques from

catching and rearing bumble bees to running gels. Stella Koulianos taught me many

molecular techniques, but eventually Nadia Krüger did most of the molecular work.

Christine Reber Funk and Roland Loosli were ever helpful with laboratory and computer

problems, as were Renate Brunner and Rita Jenny with administration. Hanne Magro was

tremendously helpful with the bumble bee rearing. Nicole Duvoisin, Annette Sauter, and

Karin Suter helped increasing the amount of data collected even more. Madeleine Beekman

was always helpful by sharing her bumble bee breeding expertise (and her insight into

where to observe platypus).

Boris Baer, Barbara Baer-Imhoof, Elmar Benelli, Erika Bucheli, Georg Funk, Oliver Kaltz,

Alfred Köpf, Pius Korner, Kristeil Le Garrec, Yannick Moret, Sonja Negovetic, Roland

Regös, Mark Rigby, Alex Widmer, and Esther Wullschleger - my fellow Ph.D. students -

shared willingly all their experience, ideas, materials, and coffee breaks. So did Regula

Billetter, Martin Blaser, Dina Dechmann, Oliver Fischer, Barbara Gautschi, Anne

Meyerhöfer, Kerstin Huck, Cornelia König, Stefan Liersch, Senta Niederegger, Katina

Puhr, Katerina Pirounakis, Hans-Peter Stauffer, and Franca Theile.

Klas Allander, Boris Baer, Barbara Baer-Imhoof, Elmar Benelli, Claudie Doums-Traoré,

Jean-Robert Escher, Gerhard Gasser Gerloff, Thomas Gloor, Kerstin Huck, Urs Lengwiler,

Pia Mutikainen, Katina Puhr, Christine Reber Funk, Regula Schmid-Hempel, Sven Seitz,

Hans-Peter Stauffer, Frank Sunder, Jouni Taskinen, and Anna-Barbara Utelli ran with me

across meadows, insect nets in our hands, enjoying the beautiful scenery of my (Alpine)

field sites. Simon Bieri, Bernhard Merz, and Andreas Müller from the Entomological

Collection of ETH Zürich, and Felix Amiet helped when it came to bumble bee

identification.

My family was very supportive and patient, knowing from own experience what writing a

Ph.D. thesis involves. Especially my sister Dietlind encouraged me constantly by intense e-

mail correspondence. I am very grateful for the time she sacrificed for her motivating help

with the finishing touches of this thesis. Less appreciated were her black and yellow striped

gifts - also named "Gräuels" - from all over the world...

Geri was always there to remind me of the world outside my lab. He shared all my ups and

downs as a Ph.D. student, accepting that the bumble bees outnumbered him and therefore

often had priority, or that the reward for his weekend pizza delivery service was a thankful

smile and a report about the latest stats result. He married me nevertheless.

Last, but not least, numerous bumble bees were (sometimes too) cooperative, although they

had not deliberately chosen to advance science.

1

Table of Contents

Summary i

Zusammenfassung iii

1 General Introduction and Thesis Outline 1

The Study Organisms - Bumble Bees 3

Thesis Outline 4

References 7

2 Abundance and Distribution of the Subalpine and Alpine

Bumble Bee Species in the Swiss Alps, with Notes on a

Gynandromorphic Bombus pratorum (L.) 13

Abstract 13

Introduction 14

Materials and Methods 15

Results 20

Discussion 26

Acknowledgements 31

References 31

Inbreeding Depression and Family Variation in a Social

Insect, Bombus terrestris (L.) (Hymenoptera: Apidae) 37

Abstract 37

Introduction 38

Materials and Methods 42

Results 50

Discussion 61

Acknowledgements 66

References 67

Effects of Inbreeding and Ploidy on Offspring Quality in a

Social Insect, Bombus terrestris (L.) (Hymenoptera: Apidae) 77

Abstract 77

Introduction 78

Materials and Methods 82

Results 87

Discussion 96

Acknowledgements 100

References 101

11

5 A Linkage Analysis of Sex Determination in

Bombus terrestris (L.) (Hymenoptera: Apidae) 113

Abstract 113

Introduction 114

Materials and Methods 117

Results 123

Discussion 128

Acknowledgements 132

References 132

6 Bibliography 137

Curriculum Vitae 163

Appendix I

Chapter 2 II

Chapter 3 VII

Chapter 4 XI

Ill

Summary

In small and isolated populations the likelihood of inbreeding is increased. In

many cases, inbreeding leads to inbreeding depression, i.e., lower fitness of inbred

offspring as compared to non-inbred offspring. In addition to the negative effects on

fitness, inbreeding in bumble bees leads to the production of a new caste, the diploid

male. These males develop from eggs that should have produced workers or young

queens but instead develop into diploid males because they are homozygous at the sex

determining locus (or, perhaps, several loci). The production of diploid males is costly

for the colony because it decreases the work force, and furthermore, the diploid males

do not sire viable colonies.

The aim of this thesis was to study the potential occurrence of inbreeding in

natural populations of bumble bees and to test the effects of inbreeding experimentally.

First, I quantitatively surveyed the bumble bee species of a central alpine region

(Valais, southwestern Switzerland). I studied the abundance and distribution of the

subalpine and alpine Bombus species at 18 study sites. Many of the 21 bumble bee

species found seem to have population sizes that are large enough not to impose

problems of small population size, such as inbreeding. However, a few bumble bee

species were so rare as to potentially face these problems; these were B. subterraneus,

B. argillaceus, B. sylvarum, and, to a lesser degree, B. gerstaeckeri and B. jonellus. As

in previous studies, abundant species were widely distributed (present at all or almost

all of the study sites), while species occurring in low numbers had a restricted

distribution (present only at a few sites). In addition, I briefly describe a

gynandromorphic specimen of B. pratorum.

Second, I studied the effect of inbreeding on several traits of the colony cycle and

offspring quality in the bumble bee, B. terrestris (L.). In a laboratory experiment, I

IV

compared colonies of queens mated to their brother to colonies of queens mated to an

unrelated male (33 inbred and 43 outbred colonies out of 15 maternal families). For all

colony traits measured, family lines differed in inbreeding depression. In some family

lines the inbred colonies were even fitter than the outbred colonies. In contrast,

offspring quality (level of immune response and body size) was not affected by

inbreeding but only by family and colony of origin. These results suggest that as a

population, B. terrestris can cope with a short inbreeding episode, provided that the

diversity of family lines is large enough to include family lines that do not suffer from

strong inbreeding depression. An additional finding of this experiment is that both sex

and ploidy affect offspring quality: diploid males had a significantly lower immune

response than haploid males, who in turn had a significantly lower immune response

than workers of the same colony. The body size of diploid males was intermediate

between the body size of workers and haploid males.

Finally, a linkage analysis of sex determination in B. terrestris is presented. The

sex determining locus mapped to a single position on the linkage map generated by this

analysis, supporting a single locus complementary sex determination system for this

species.

V

Zusammenfassung

In kleinen und isolierten Populationen ist die Inzuchtswahrscheinlichkeit erhöht.

In vielen Fällen führt Inzucht zu Inzuchtsdepression, d.h. zu geringerer Fitness in

ingezüchteten Nachkommen als in nicht ingezüchteten Nachkommen. Neben den

negativen Auswirkungen auf Fitness hat Inzucht bei Hummeln die Entstehung einer

neuen Kaste, den diploiden Männchen, zur Folge. Solche Männchen entwickeln sich

aus Eiern, die Arbeiterinnen oder junge Königinnen hätten hervorbringen sollen, sich

aber stattdessen zu diploiden Männchen entwickeln, weil sie am

geschlechtsbestimmenden Locus (oder vielleicht mehreren Loci) homozygot sind. Das

Hervorbringen diploider Männchen ist kostspielig für die betreffende Kolonie, weil ihre

Arbeiterinnenzahl verringert wird und diploide Männchen ausserdem keine

lebensfähigen Kolonien zeugen.

Das Ziel dieser Dissertation war es, das potentielle Auftreten von Inzucht in

freilebenden Hummelpopulationen zu untersuchen und die Auswirkungen von Inzucht

experimentell zu testen.

Als Erstes habe ich die Hummelarten eines Gebietes der Zentralpen (Wallis,

Südwestschweiz) quantitativ erfasst. Ich untersuchte die Häufigkeit und Verbreitung

subalpiner und alpiner Bombus-Arten an 18 Untersuchungsstandorten. Viele der 21

nachgewiesenen Hummelarten scheinen Populationsgrössen zu haben, die gross genug

sind, keine Probleme aufkommen zu lassen, die mit kleiner Populationsgrösse

einhergehen, wie z.B. Inzucht. Einige Hummelarten waren jedoch so selten, dass sie

möglicherweise von solchen Problemen betroffen sind; es waren dies B. subterraneus,

B. argillaceus, B. sylvarum sowie, in geringerem Ausmass, B. gerstaeckeri und B.

jonellus. Wie in früheren Untersuchungen zeigten häufige Arten eine weite Verbreitung

(an allen oder fast allen Untersuchungsstandorten nachgewiesen), wohingegen Arten

VI

mit zahlenmässig geringem Vorkommen eine begrenzte Verbreitung aufwiesen (nur an

wenigen Standorten festgestellt). Ausserdem beschreibe ich kurz ein gynandromorphes

Tier der Art B. pratorum.

Als Zweites habe ich die Auswirkungen von Inzucht auf verschiedene Merkmale

des Koloniezyklus und der Nachkommensqualität in der Dunklen Erdhummel (B.

terrestris (L.)) untersucht. In einem Laborexperiment verglich ich Kolonien von

Königinnen, die sich mit einem Bruder verpaart hatten, mit Kolonien von Königinnen,

die sich mit einem nichtverwandten Männchen verpaart hatten (33 ingezüchtete und 43

ausgezüchtete Kolonien aus 15 mütterlichen Familien). Sämtliche Familien

unterschieden sich bezüglich der Inzuchtsdepression voneinander, die sie für die

Koloniemerkmale zeigten. In einigen Familien waren die Inzuchtskolonien sogar fitter

als die ausgezüchteten. Im Gegensatz dazu wurde die Qualität des Nachwuchses (Stärke

der Immunantwort und Körpergrösse) nicht von Inzucht beeinflusst sondern von der

Herkunftsfamilie und -kolonie. Diese Ergebnisse deuten darauf hin, dass eine

Population von B. terrestris eine kurze Inzuchtsepisode verkraften kann, sofern die

Familienvielfalt gross genug ist, um Familien mit einzuschliessen, die keine starke

Inzuchtsdepression aufweisen. Eine zusätzliche Erkenntnis aus diesem Versuch ist, dass

sowohl Geschlecht als auch Ploidiegrad die Qualität des Nachwuchses beeinflussen:

diploide Männchen hatten eine signifikant schwächere Immunantwort als haploide

Männchen, welche wiederum eine signifikant schwächere Immunantwort als

Arbeiterinnen derselben Kolonie hatten. Die Körpergrösse diploider Männchen lag

zwischen derjenigen von Arbeiterinnen und haploiden Männchen.

Schliesslich wurde eine Kopplungsanalyse des Hummel-Genoms durchgeführt,

um den Locus für die Geschlechtsbestimmung bei B. terrestris zu lokalisieren. Nur ein

Locus für die Geschlechtsbestimmung wurde gefunden und in die erstellte Genomkarte

integriert. Dieses Ergebnis stützt die Vermutung, dass das komplementäre

Geschlechtsbestimmungssystem dieser Art auf einem einzelnen Locus basiert.

1

1 General Introduction and Thesis Outline

Rare species are a major concern of conservation biology, the "science of scarcity and

diversity" (Soulé 1986). Rarity of a species can be defined on different (geographic) scales

(Soulé 1986, section II). In any case, however, rarity involves small population size at least

on one of the scales. Small population size is associated with several problems: (i)

increased risk of inbreeding and its negative effects on fitness, (ii) loss of genetic variability

due to genetic drift that deprives a population of its potential for adaptation and can lead to

fixation of deleterious alleles, and (iii) increased risk that stochastic environmental

perturbations affect all members of a small population at once (for reviews, see Ellstrand &

Elam 1993; Frankham 1995; Hedrick & Kalinowski 2000). All of these processes - usually

in combination with each other - can eventually lead to extinction of a species or a

population by setting off or powering an extinction vortex (Gilpin & Soulé 1986; for

evidence, see e.g., Frankham & Ralls 1998; Saccheri et al. 1998).

In my thesis, I focus on inbreeding - the mating of two related individuals (Roff

1997) - as a possible consequence of small population size. It has long been recognized that

inbreeding can lead to inbreeding depression, i.e., lower fitness of inbred offspring as

compared to non-inbred offspring (e.g., Charlesworth & Charlesworth 1987; Thornhill

1993; Frankham 1995; Lacy 1997; Roff 1997). At present, the actual mechanisms behind

inbreeding depression remain controversial (Charlesworth & Charlesworth 1987; Thornhill

1993; Roff 1997). The overdominance hypothesis assumes that heterozygotes have superior

fitness to homozygotes. Thus, inbreeding could take effect by increasing homozygosity.

Alternatively, the partial dominance hypothesis states that inbreeding depression is caused

2

by an increase in frequency of recessive or partially recessive deleterious alleles following

inbreeding. While the understanding of the actual genetic mechanism is of great

importance, in my thesis I am concerned with the ecological implications, i.e., for which

species is inbreeding likely to occur and what kinds of effects on fitness parameters are to

be expected.

Social Hymenoptera, i.e., ants, bees, and wasps, are a particularly interesting group of

insects with respect to the effects of inbreeding. Due to their mode of sex determination

inbreeding can result in the production of a new caste, the diploid male (e.g., Cook 1993;

Cook & Crozier 1995). Normally, males develop from unfertilized eggs, while females

(workers and queens) develop from fertilized eggs (Crozier 1975). Sex is determined by a

single, or perhaps several, multi-allelic locus, the "sex locus" ("complementary sex

determination", CSD; Whiting 1940). Haploid individuals are hemizygous at the sex locus

and develop into males. Diploid individuals that are heterozygous at the sex locus develop

into females. In contrast, diploid individuals that are homozygous at the sex locus develop

into males. By increasing homozygosity, inbreeding increases the incidence of diploid

males in social Hymenoptera.

The production of diploid males is costly for a colony, (i) Diploid males develop

from eggs that should have given rise to workers or young queens. Thus, a colony's worker

force or its reproductive output are reduced, resulting in a fitness reduction, (ii) Diploid

males rarely sire viable offspring or else have infertile triploid offspring (Page 1980;

Ratnieks 1990; Stouthamer et al. 1992; Cook 1993; Cook & Crozier 1995).

3

The Study Organisms - Bumble Bees

Bumble bees are eusocial Hymenoptera that live mainly in temperate regions and

have an annual life cycle (e.g., Alford 1975; Morse 1982). In spring, queens emerge from

hibernation and individually found a colony. During the summer, the colony grows in

worker numbers and reproduces by raising sexuals, the males and daughter queens, in late

summer. After mating, the young queens enter hibernation while all the members of the old

colony die.

Bumble bees are important pollinators in many natural and agricultural ecosystems

(Kearns et al. 1998; Delaplane & Mayer 2000; Sommeijer & de Ruijter 2000). Both natural

and commercial bumble bee populations can be faced with restricted population sizes and

subsequent inbreeding. In island populations of bumble bees, significant levels of

inbreeding have been revealed by molecular analyses (Widmer et al. 1998) and the frequent

occurrence of diploid males (Buttermore et al. 1998). The biology of bumble bees indicates

the potential occurrence of - at least episodic - inbreeding also in non-island populations.

Each year, only a few, successful colonies reproduce (Donovan & Weir 1978; Müller &

Schmid-Hempel 1992; Imhoof & Schmid-Hempel 1999). This results in a reproduction bias

within a population. Furthermore, population sizes fluctuate strongly due to environmental

conditions (Harder 1986; Sladen 1989; Pamilo & Crozier 1997). In addition, the natural

habitats of bumble bees and other pollinators suffer from the increasing impact of human

activities, e.g., habitat fragmentation and the loss of flower resources due to monocultures

(Delaplane & Mayer 2000). Buchmann & Nabhan (1996) even warn of "the impending

pollination crisis" (their p. 15). Bumble bees reared for pollination purposes are also likely

to suffer from increasing levels of inbreeding because of the closed breeding systems of

commercial breeders (Duchateau et al. 1994; Duchateau 2000).

4

In bumble bees, diploid males are reared to adulthood (Garöfalo 1973; Duchateau et

al. 1994; Duchateau & Marien 1995) in contrast to, for example, the honey bee, in which

the diploid male brood is selectively killed by the workers (Woyke 1963). The rearing of

mostly infertile diploid males (Duchateau & Marien 1995) can impose costs on a colony

such as reduced colony growth (Plowright & Pallett 1979, but see Duchateau et al. 1994).

Thesis Outline

In my thesis, I study the potential occurrence of inbreeding in natural populations of

bumble bees and experimentally test the effects of inbreeding on bumble bees under

laboratory conditions.

First, I give an overview of the abundance and distribution of the bumble bee species

in the Swiss Alps (Chapter 2). The Alps are known for their high species diversity of

bumble bees (Rasmont et al. 2000). I focussed on the higher altitudes, i.e, the subalpine and

alpine zones. This allowed me to sample a habitat type with restricted distribution on a

larger geographical scale, i.e., Central Europe. Based on the natural history data collected I

identified rare species that potentially face the problems of small population size, such as

inbreeding.

Then, I describe an inbreeding experiment with the bumble bee, Bombus terrestris

(L.). While this species is not present at high altitudes in the Alps, it is abundant at lower

altitudes across Europe. For the European continent, molecular data suggest one large

population of B. terrestris without detectable genetic differentiation or inbreeding (Estoup

et al. 1996). In contrast, island populations of this species can show significant amounts of

inbreeding (Buttermore et al. 1998; Widmer et al. 1998). Thus, the use of B. terrestris from

5

the Swiss lowlands ensured that the bumble bees tested were probably not inbred before the

experiment. Furthermore, B. terrestris can be reared well in the laboratory, which allowed

for standardized experimental conditions.

In a first part of the inbreeding experiment, I studied the effects of brother-sister

matings on several traits of the colony life cycle (Chapter 3). In contrast to a previous study

(Duchateau et al. 1994), the experimental design was based on several family lines. Recent

studies have found differences in inbreeding depression among family lines in a variety of

organisms (e.g., Stevens et al. 1997; Mutikainen & Delph 1998; Byers & Waller 1999).

Indeed, I found large among-family variation for hibernation survival and colony

foundation success of the bumble bee queens. Furthermore, family lines differed in their

reaction to inbreeding with respect to colony size and reproductive output. In some family

lines the inbred colonies were even fitter than the outbred colonies. This latter finding was

unexpected considering the costs diploid male production can impose on inbred colonies

(Plowright & Pallett 1979).

In a second part of the inbreeding experiment, I studied the effects of brother-sister

matings on offspring quality (Chapter 4). For this, I compared the level of immune

response and the body size of workers and males originating from inbred and outbred

colonies. One generation of inbreeding did not affect offspring quality. However, there was

again large among-family and among-colony variation for the traits measured. Furthermore,

the occurrence of diploid males allowed me to investigate the effects of sex and ploidy on

offspring quality in more detail. Diploid males had a weaker immune response than haploid

males, who in turn had a weaker immune response than workers of the same colony. In

contrast, diploid males were smaller than haploid males but larger than workers of the same

colony. Obviously, the unusual combination of male sex and diploidy decreased offspring

quality in bumble bees.

6

In addition, the occurrence of diploid males enabled further insight into the sex

determination system of bumble bees. A linkage analysis of sex determination was

performed with a sample of workers and diploid males originating from a single colony

(Chapter 5). According to the complementary sex determination system (Whiting 1940),

the only consistent genotypic difference between these workers and diploid males was,

respectively, the hetero- and homozygosity at the sex determining locus (or maybe several

loci). Indeed, the sex determining locus mapped to a single position on the linkage map

generated for B. terrestris. This result supports a single locus complementary sex

determination system for this species.

The results of this thesis suggest that the populations of many Alpine bumble bee

species are presumably large enough not to be threatened by inbreeding or other

consequences of small population size. However, the biology of bumble bees indicates that

even in large populations short episodes of inbreeding might occur. The large among-

family variation for inbreeding depression observed in B. terrestris suggests that

populations of - at least this - bumble bee species are able to tolerate short episodes of

inbreeding. A possible reason for this tolerance to inbreeding as a population might be

regular purging of deleterious alleles either through the effects of haplodiploidy (leading to

expression of recessive deleterious alleles in males) or by frequent inbreeding episodes that

expose family lines with large loads of deleterious alleles. However, in the case of small

populations, the diversity of family lines might be too low to include lines that successfully

purged deleterious alleles or their genetic load in the past, or which will succeed in doing so

in the future. In this case, inbreeding is likely to decrease the fitness of a bumble bee

population and thus to increase its risk of extinction.

7

References

Alford, D.V. 1975. Bumblebees. Davis-Poynter, London, UK.

Buchmann, S.L. & Nabhan, G.P. 1996. The forgotten pollinators. Island Press, Washington

D.C., USA.

Buttermore, R.E., Pomeroy, N., Hobson, W., Semmens, T. & Hart, R. 1998. Assessment of

the genetic base of Tasmanian bumble bees (Bombus terrestris) for development as

pollination agents. J. Apic. Res. 37:23-25.

Byers, D.L. & Waller, D.M. 1999. Do plant populations purge their genetic load? Effects of

population size and mating history on inbreeding depression. Annu. Rev. Ecol. Syst.

30:479-513.

Charlesworth, D. & Charlesworth, B. 1987. Inbreeding depression and ist evolutionary

consequences. Annu. Rev. Ecol. Syst. 18:237-268.

Cook, J.M. 1993. Sex determination in the Hymenoptera: a review of models and evidence.

Heredity 71:421-435.

Cook, J.M. & Crozier, R.H. 1995. Sex determination and population biology in the

Hymenoptera. Trends Ecol. Evol. 10:281-286.

Crozier, R.H. 1975. Insecta 7. Hymenoptera. Gebrüder Borntraeger, Berlin, Germany.

Delaplane, K.S. & Mayer, D.F. 2000. Crop pollination by bees. CABI Publishing, Oxon,

UK.

Donovan, B.J. & Weir, S.S. 1978. Development of hives for field population increase, and

studies on the life cycle of four species of introduced bumble bees in New Zealand.

N. Z. J. Agric. Res. 21:733-756.

8

Duchateau, M.J., Hoshiba, H. & Velthuis, H.H.W. 1994. Diploid males in the bumble bee

Bombus terrestris: sex determination, sex alleles and viability. Entomol. Exp. Appl.

71:263-269.

Duchateau, M.J. & Marien, J. 1995. Sexual biology of haploid and diploid males in the

bumble bee Bombus terrestris. Insect. Soc. 42:255-266.

Duchateau, M.J. 2000. Biological aspects of rearing bumble bees for pollination, pp. 25-29

in M.J. Sommeijer & A. de Ruijter, eds. Insect pollination in greenhouses:

proceedings of the specialists' meeting held in Soesterberg, The Netherlands, 30

September to 2 October 1999. University of Utrecht, The Netherlands.

Ellstrand, N.C. & Elam, D.R. 1993. Population genetic consequences of small population

size: implications for plant conservation. Annu. Rev. Ecol. Syst. 24:217-242.

Estoup, A., Solignac, M., Cornuet, J.M., Goudet, J. & Scholl, A. 1996. Genetic

differentiation of continental and island populations of Bombus terrestris

(Hymenoptera: Apidae) in Europe. Mol. Ecol. 5:19-31.

Frankham, R. 1995. Conservation genetics. Annu. Rev. Genet. 29:305-27.

Frankham, R. & Ralls, K. 1998. Inbreeding leads to extinction. Nature 392:441-442.

Garöfalo, C.A. 1973. Occurence of diploid drones in a neotropical bumblebee. Experientia

29:726-727.

Gilpin, M.E. & Soulé, M.E. 1986. Minimum viable populations: processes of species

extinction, pp. 19-34 in M.E. Soulé, ed. Conservation biology: the science of scarcity

and diversity. Sinauer Associates Inc., Sunderland, MA, USA.

Harder, L.D. 1986. Influences on the density and dispersion of bumble bee nests

(Hymenoptera: Apidae). Holarctic Ecol. 9:99-103.

Hedrick, P.W. & Kalinowski, S.T. 2000. Inbreeding depression in conservation biology.

Annu. Rev. Ecol. Syst. 31:139-162.

9

Imhoof, B. & Schmid-Hempel, P. 1999. Colony success of the bumble bee, Bombus

terrestris, in relation to infections by two protozoan parasites, Crithidia bombi and

Nosema bombi. Insect. Soc. 46:233-238.

Kearns, C.A., Inouye, D.W. & Waser, N.M. 1998. Endangered mutualisms: the

conservation of plant-pollinator interactions. Annu. Rev. Ecol. Syst. 29:83-112.

Lacy, R.C. 1997. Importance of genetic variation to the viability of mammalian

populations. J. Mammal. 78:320-335.

Morse, D.H. 1982. Behavior and ecology of bumble bees. pp. 245-322 in H.R. Hermann,

ed. Social Insects. Academic Press, New York, NY, USA.

Müller, C.B. & Schmid-Hempel, P. 1992. Correlates of reproductive success among field

colonies of Bombus lucorum: the importance of growth and parasites. Ecol. Entomol.

17:343-353.

Mutikainen, P. & Delph, L.F. 1998. Inbreeding depression in gynodioecious Lobelia

siphilitica: among-family differences override between-morph differences. Evolution

52:1572-1582.

Page, R.E., Jr. 1980. The evolution of multiple mating behavior by honey bee queens {Apis

mellifera L.). Genetics 96:263-274.

Pamilo, P. & Crozier, R.H. 1997. Population biology of social insect conservation. Mem.

Museum Victoria 56:411-419.

Plowright, R.C. & Pallett, MJ. 1979. Worker-male conflict and inbreeding in bumble bees

(Hymenoptera: Apidae). Can. Entomol. 111:289-294.

Rasmont, P., Verhaeghe, J.C., Rasmont, R. & Terzo, M. 2000. West-Palaearctic

bumblebees, pp. 93-99 in M.J. Sommeijer & A. de Ruijter, eds. Insect pollination in

greenhouses: proceedings of the specialists' meeting held in Soesterberg, The

10

Netherlands, 30 September to 2 October 1999. University of Utrecht, The

Netherlands.

Ratnieks, F.L.W. 1990. The evolution of polyandry by queens in social Hymenoptera: the

significance of the timing of removal of diploid males. Behav. Ecol. Sociobiol.

26:343-348.

Roff, D.A. 1997. Evolutionary quantitative genetics. Chapman & Hall, New York, NY,

USA.

Saccheri, I., Kuussaari, M., Kankare, M., Vikman, P., Fortelius, W. & Hanski, I. 1998.

Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494.

Sladen, F.W.L. 1989. The humble-bee. Its life history and how to domesticate it. Logaston

Press, Little Logaston, UK.

Sommeijer, M.J. & de Ruijter, A. 2000. Insect pollination in greenhouses: proceedings of

the specialists' meeting held in Soesterberg, The Netherlands, 30 September to 2

October 1999. University of Utrecht, The Netherlands.

Soulé, M.E. 1986. Conservation biology: the science of scarcity and diversity. Sinauer

Associates Inc., Sunderland, MA, USA.

Stevens, L., Yan, G. & Pray, L.A. 1997. Consequences of inbreeding on invertebrate host

susceptibility to parasitic infection. Evolution 51:2032-2039.

Stouthamer, R., Luck, R.F. & Werren, J.H. 1992. Genetics of sex determination and the

improvement of biological control using parasitoids. Environ. Entomol. 21:427-435.

Thornhill, N.W. 1993. The natural history of inbreeding and outbreeding: Theoretical and

empirical perspectives. The University of Chicago Press, Chicago, IL, USA.

Whiting, P.W. 1940. Multiple alleles in sex determination of Habrobracon. J. Morph.

66:323-355.

11

Widmer, A., Schmid-Hempel, P., Estoup, A. & Scholl, A. 1998. Population genetic

structure and colonization history of Bombus terrestris s.l. (Hymenoptera: Apidae)

from the Canary Islands and Madeira. Heredity 81:563-572.

Woyke, J. 1963. Drone larvae from fertilized eggs of the honeybee. J. Apic. Res. 2:19-24.

12

Seite Leer /Blank leaf

13

2 Abundance and Distribution of the Subalpine and Alpine

Bumble Bee Species in the Swiss Alps, with Notes on a

Gynandromorphic Bombus pratorum (L.)

(GerloffC.U. & Schmid-Hempel P.)

Abstract

The Alps are known for their high species diversity of bumble bees. Here, we

studied the abundance and distribution of bumble bee species of the subalpine and

alpine zones, seeking to identify rare species that potentially face the problems imposed

by small population size, such as inbreeding and loss of genetic variability. Across one

season, we caught a total of 2252 bumble bees of 21 species (excluding the subgenus

Psithyrus) at 18 sites in the Swiss Alps. The data suggest that many Alpine bumble bee

species have population sizes that are presumably large enough not to impose problems.

However, Bombus subterraneus, B. argillaceus, B. sylvarum, and, to a lesser degree, B.

gerstaeckeri and B. jonellus were so rare as to be potentially threatened by the

consequences of small population size. The mean species richness of the sites was 13.0

± 0.6 (SE; range 9-17 species). Bumble bee species that were abundant locally tended to

be present in many sites, while species that occurred in low densities tended to have a

restricted distribution. Our data did not show the distributional pattern predicted under

the core and satellite species hypothesis (Hanski 1982c). In addition, we briefly describe

a bilateral gynandromorph of B. pratorum, of which the head and thorax exhibited a

mosaic of female and male tissue.

14

Keywords

abundance; alpine; Alps; Bombus; bumble bees; gynandromorph; subalpine.

Introduction

Species differ in abundance and distribution. Generally, common species tend to

occur locally in high densities and to be widely distributed, while rare species tend to

occur locally in low densities and to be distributed patchily (e.g., Hanski 1982c; Brown

1984; Gaston & Lawton 1989; Hanski et al. 1993; and references therein). This pattern

goes along with large differences in population sizes between common and rare species.

The typically small population size of rare species makes them vulnerable to inbreeding,

loss of genetic variability due to genetic drift, and eventually (local) extinction

(Ellstrand & Elam 1993; Frankham 1995; Frankham & Ralls 1998; Saccheri et al. 1998;

Hedrick & Kalinowski 2000).

Bumble bees are important pollinators in many natural and rural ecosystems

(Kearns et al. 1998; Delaplane & Mayer 2000). Worldwide, there are 239 bumble bee

species (including the subgenus Psithyrus, the cuckoo bumble bees) that are distributed

mainly in temperate regions (Williams 1998). In contrast to other pollinators, bumble

bee species diversity is high in the boreal and alpine regions of the northern latitudes

and at high altitudes in mountain ranges (Pekkarinen & Teräs 1993; Rasmont et al.

2000b). In Europe, the Alps are known for their high species diversity of bumble bees

(Rasmont et al. 2000b), but studies of the bumble bee communities of this area are

scarce (Stupf 1992; Amiet 1996).

The aim of this study was to survey quantitatively the bumble bee species of a

representative region of the Swiss Alps, focussing on higher altitudes. This type of

15

quantitative study has added value as a reference for future studies if bumble bee

populations are declining (Williams 1982). Here, we studied the abundance and

distribution of the subalpine and alpine bumble bee species at 18 sites. Our results

confirm the high bumble bee species diversity of the Alps. Based on the natural history

data collected we identified rare species that potentially face the problems of small

population size. Furthermore, we compared the distributional pattern observed in this

study to the predictions of the core and satellite species hypothesis (Hanski 1982c). In

addition, we briefly describe a gynandromorph specimen of Bombus pratorum.

Materials and Methods

Study sites

Across the Valais, a large Alpine valley in southwestern Switzerland, we

randomly chose 20 grid squares of 10 x 10 km (national maps of Switzerland, scale

1:25000, Federal Office of Topography, Wabern, Switzerland). Within each of these

grid squares, we chose a site in the subalpine and alpine zone within the

phytogeographic region "Central Alps" (Landolt 1984), and which was accessible at an

altitude of 1700 m. We discarded two of the chosen grid squares, because they did not

contain sites that fullfilled our criteria, leaving a total of 18 sites with various

expositions (Fig. 2.1; Table A2.1). At each site, the sampling covered open areas

(subalpine and alpine meadows and pastures) and forest above the altitude of 1700 m.

(The natural tree line is between 2200 and 2400 m (Landolt 1984).) The proportion of

forest varied among sites. At Gletsch (site 2, Fig. 2.1), the "forest" consisted exclusively

of alder shrubs (Alnus viridis).

46°40'N

-allens

r

2^(

'(I

45°49'N

-i

6°38'E

8°30'E

Fig

2.1.Study

siteswithintheValais,southwesternSwitzerland.Localities

are:

1:Gnmentz,

2:Gletsch,

4:Rothwald,

5:Tasch,

6:Mayensde

My,

7:Ulnchen,

8:La

Forc

laz,

9:Belalp,

10:

Verb

ier,

11:Vispertermmen,

12:Bo

urg-

St.-

Pierre,

13:Sa

as-A

lmag

ell,

14:Bellwald,

15:

Ovronnaz,

16:Leukerbad,

17:Mauvoism,

18:Thyon,20:Verconn.For

siteandsamplingch

arac

teri

stic

s,seeTableA2.1.

17

Sampling

In bumble bees, population size and the local distribution pattern can change

during the course of a season due to local extinction and recolonization events (Bowers

1985). Therefore, we sampled each site twice to assess the abundance and distribution

of bumble bee species. The first time, we sampled in spring when the queens had

emerged from hibernation (collecting dates May 14 to June 5, 1997). The second time,

we sampled in summer when colonies were established (July 16 to August 8, 1997). We

kept the temporal order, in which we visited the sites during the first and the second

sampling round, as constant as possible. The average timespan between the two visits at

each site was 63.2 ± 0.5 days (mean ± SE, range 60-67 days). Following a hiking trail or

a road one (14 visits) or two people (22 visits) caught each bumble bee (Bombus spp.)

or cuckoo bumble bee (subgenus Psithyrus) spotted. To avoid collecting many bumble

bees from the same colony, we followed a trail continously in the same direction while

collecting. We set the maximum number of individuals collected to 130 per site based

on species saturation curves calculated from previous data from the same area, albeit

from lower altitudes (A. Widmer et al., unpublished data). Ideally, we collected 65

bumble bees both the first and the second time we visited a site. If we caught less than

65 bumble bees the first time we increased the sample size the second time to reach a

total of 130 bumble bees collected per site. The sampling covered a total altitudinal

range of 1700-2350 m and took place between 09:30 (at the earliest) and 20:30 (at the

latest). We interrupted the sampling during heavy rain.

We pooled the data from different castes in order to base our measures on a

broader sample. Therefore, our abundance measure includes the non-reproducing

worker caste. However, the abundance of bumble bee workers can give an indirect

indication of the effective population size: within species, worker number correlates

with the number of sexuals a colony produces (Müller & Schmid-Hempel 1992a). We

could not correct for among-species differences in colony size because the necessary

18

information is not available for Alpine bumble bee populations. On the other hand,

including the workers, which are more numerous and forage during a longer time

window than the sexuals, increased the accuracy of our distribution estimate.

Furthermore, sampling a site twice levelled out to some extent the variation among sites

induced by the sampling scheme (e.g., date, weather, and time of day). For details on

the sampling at individual sites, see Table A2.1.

Identification

The collected specimens were freeze-killed in liquid nitrogen and stored at -80° C

for later identification. For identification, we used Amiet (1996) and von Hagen (1994).

The taxonomy follows Williams (1998). For a few species, the worker caste cannot be

identified with certainty. Therefore, we could not separate the species of the Bombus

terrestris-group but pooled them as "5. lucorum", the predominant species of this

species group at this altitude (Amiet 1996). Likewise, we pooled the species B.

hortorum and B. ruderatus as "B. hortorum". A total of 20 individuals could not be

identified because of missing body parts. The cuckoo bumble bees (subgenus Psithyrus)

were not identified to the species level. A subsample of 46 specimens (one randomly

chosen specimen of each Bombus species (21 specimens) and 25 specimens of uncertain

identification) were re-identified by Dr. A. Müller, ETH Zürich, and F. Amiet,

Solothurn, Switzerland.

Size measurements of Bombus pratorum

Conspecific workers and males differ in wing size (Alford 1975; Chapter 4). We

tested whether the wing size of the female and male side of a bilateral gynandromorph

of B. pratorum differed from the average wing size of the respective caste. From 12

sites, we randomly chose one worker and one male of B. pratorum, each collected

during the second collecting round, i.e., when the gynandromorph was collected. To

19

match the distribution of female and male tissue in the gynandromorph, for workers we

measured the right wing (dorsal view) and, for males, the left wing. For this purpose, we

clamped the forewing between two microscopic slides and measured the radial cell

length (Bertsch 1984; Owen 1988, 1989; Müller & Schmid-Hempel 1992b) under a

dissecting microscope in units of 0.04 mm.

Statistics

All sample sizes, species abundances, richnesses, and diversities were calculated

without the cuckoo bumble bees (subgenus Psithyrus; not identified to the species level)

and the unidentifiable specimens. Species abundance was defined as the number of

individuals of a given species at a site (e.g., Hanski 1982c). Because sample sizes varied

among sites we based all calculations and comparisons on the relative abundances, i.e.,

a species' proportion of the total sample size at a given site. The mean relative

abundance of a species was based only on the sites at which a species was present

(Hanski et al. 1993). The distribution of a species was expressed as the percentage of

sites in which this species was found (e.g., Hanski 1982c). We calculated the diversity

D of a site as the reciprocal of Simpson's index following Calow (1998): D = /£(n)2,

where p, is the relative abundance of the i* species at the respective site. D ranges from

1 to S, the number of species (the species richness) found in a given site. High values of

D indicate high levels of diversity. D increases with increasing equitability (sometimes

referred to as "evenness"), i.e., the more balanced the species abundances are at a given

site (Calow 1998). We tested for normal distribution of the data before using parametric

tests and report values of two-tailed tests. Statistical analyses were performed with

SPSS 6.1.1 (Norusis 1994).

20

Results

At the 18 sites, we caught a total of 2252 bumble bees of 21 species (Table 2.1).

The local sample size, i.e., the sample size at each site, was 125.1 ± 2.9 bumble bees

(mean ± SE, range 78-133). During the first sampling round in spring, we caught 48.7 ±

4.1 bumble bees per site (range 11-64). During the second sampling round in summer,

we caught 76.4 ± 3.7 bumble bees per site (range 53-112). The number of species

sampled across all sites increased from the first to the second sampling round from 17 to

21, respectively. The same was observed for the distribution of individual species, i.e.,

the number of sites at which a species was present. In spring, species were on average

found in 9.1 ± 1.3 sites (50.6 ± 7.2 % of the sites, range 2-16 sites (11.1-88.9 %), n = 17

species), in summer 11.1 ± 1.4 sites (61.7 ± 7.8 %, range 1-18 sites (5.6-100 %), n = 21

species) (Table 2.1). Cuckoo bumble bees (subgenus Psithyrus) were present at 9 (50.0

%) and 13 (72.2 %) sites, respectively.

The workenqueen ratio varied widely among species (range 0.7-7.8, Table 2.1)

The large workenqueen ratios for B. sichelii (7.8), B. monticola (6.7), and B. pyrenaeus

(5.3) indicate that the abundance measured is based mainly on the worker caste. The

workenqueen ratio of bumble bee species did not correlate with their total abundance

(Spearman's r = -0.090, n=\l,P = 0.732).

Table

2.1.Abundance,

distribution,andworkerrqueen

ratioofbumble

bee

species

inthe

Valais,southwestern

Switzerland.Numbers

in

bracketsindicatevaluesfrom

the

firstsa

mpli

nground.Thedataforindividual

sitesandsampling

roundsaregiven

inTablesA2.2andA2.3.

Species

Queens

Workers

Males

Total

Distribution*

Workers:

Queens

Bombuspratorum(L

)139

(138)

193

(1)

46

379t

(139)

18

(16)

1.4

Blucorum(L

)176

(175)

124

(35)

22

322

(210)

18

(16)

0.7

Bsoroeensis(Fabncius)

143

(139)

173

4320

(139)

18

(16)

1.2

Bwu

rfle

nuRadoszkowski

117

(115)

162

8287

(115)

17

(15)

1.4

BmonticolaSmith

22

(22)

147

(67)

18

(1)

187

(90)

17

(16)

6.7

BpyrenaeusPerez

24

(24)

126

(1)

10

160

(25)

18

(10)

5.3

Bruderarius(Müller)

47

(47)

95

(1)

11

153

(48)

18

(11)

2.0

BmesomelasGerstaecker

35

(33)

73

108

(33)

17

(14)

2.1

BsichelnRadoszkowski

8(7)

62

70

(7)

17

(5)

7.8

BmendaxGerstaecker

19

(19)

49

68

(19)

13

(5)

2.6

BhumilisIl

hger

12

(11)

39

455

(11)

10

(5)

3.3

Bhortorum(L

)9

(8)

37

551

(8)

9(5)

4.1

BmucidusGerstaecker

6(6)

14

121

(6)

6(2)

2.3

Blapida

rius

(L

)7

(7)

12

120

(7)

9(4)

1.7

Bjonell

usKirby

9(9)

716

(9)

10

(9)

0.8

Bpascuorum(Scopoh)

4(4)

11

15

(4)

7(4)

2.8

Bhypnorum(L

)4

(3)

32

9(3)

3(2)

0.8

Bgerstaeckeri

Morawitz

66

4

Bsubterraneus(L

)2

22

Bsylvarum(L

)2

22

Bargdlaceus(Scopoh)

11

1

Total:

Mean

(SE):

21Bombusspp

(17)

787

(767)

1331

(105)

133

(1)

2252

(873)

11.1±1.4

(91+1

3)2.76±0.51

SubgenusPsithyrus

22

(16)

16

38

(16)

13

(9)

*numberofsitesaspecieswaspresent

tincludesonegynandromorph

22

The total abundance (Table 2.1) and the mean relative abundance of bumble bee

species (Fig. 2.2) were highly correlated (Spearman's r = 0.976, n = 21, P < 0.0005),

meaning that a species that occurred at high densities where it was found was also

highly abundant overall. Similarly, the mean relative abundance and the distribution,

expressed as the percentage of sites a species was present, of the species was correlated

(Spearman's r = 0. 898, n = 21, P < 0.0005) (Fig. 2.2). This means that a species that

was abundant locally also occurred at many sites and that no patches were found where

rare species were abundant. Within the more widely distributed species, relative

abundance varied dramatically among sites (Fig. 2.2). The largest relative abundances

were observed in B. pratorum and B. soroeensis: up to half of the individuals sampled

at a site belonged to these species (55.6 %, site 20, and 49.2 %, site 7, respectively; see

Fig. 2.1. for the location of the sites).

Ou

se«

73S

3£>«

>

a«

60-i

50-

40-

30-

20-

10-

0 _ arü O Shyrjjsu/sy

mut5

Dpa ÏB h ï

0 20 40

—j—

60

wu

mot

si 3ÜL

pr

so/lu

ru/py

80 100

Distribution (percentage of sites present)

Fig. 2.2. The relationship between mean relative abundance and distribution in

subalpine bumble bee species. The bars indicate the range of relative abundance. The

exact data are given in Table A2.4. ar: Bombus argillaceus, ge: B. gerstaeckeri, ho: B.

hortorum, hu: B. humilis, hy: B. hypnorum, jo: B. jonellus, la: B. lapidarius, lu: B.

lucorum, mn: B. mendax, ms: B. mesomelas, mo: B. monticola, mu: B. mucidus, pa: B.

pascuorum, pr: B. pratorum, py: B. pyrenaeus, ru: B. ruderarius, si; B. sichelii, so: B.

soroeensis, su: B. subterraneus, sy: B. sylvarum, wu: B. wurflenii.

23

The four most common bumble bee species in the subalpine and alpine areas of

the Valais were B. pratorum, B. lucorum, B. soroeensis, and B. wurflenii (based on the

mean relative abundance, listed in descending order; Fig. 2.2). Each of these species had

a mean relative abundance of more than 13 % and was present at 94 % or more of the

sites. The four rarest bumble bee species in the study area were B. subterraneus, B.

argillaceus, B. sylvarum, and B. gerstaeckeri (based on the mean relative abundance,

listed in ascending order; Fig. 2.2). Each of these species had a mean relative abundance

of less than 1.2 % and was present at less than 25 % of the sites. B. jonellus was

remarkebly widespread (present at 56 % of the sites) but not very abundant throughout

(mean relative abundance 1.3 %). In contrast, B. hypnorum had the eighth lowest mean

relative abundance (2.3 %) but was present only at 17 % of the sites (Fig. 2.2, Table

2.1).

The local species richness S of the bumble bee communities ranged from 9 to 17

species, with an average of 13.0 ± 0.6 (SE) species per site (Fig. 2.3). The mean local

diversity D was 6.72 ± 0.46 (Fig. 2.3). The most diverse site was Grimentz (site 1, Fig.

2.1) with D = 9.97, the least diverse site was Vercorin (site 20, Fig. 2.1) with D = 2.98.

Species rich sites had a high diversity (Spearman's r = 0.520, n = 18, P = 0.027), as can

be expected for a diversity measure that incorporates species richness as a factor (Calow

1998).

24

tu

12-1

10-

6-

4-

2-

0-

D13

D8

D2

D7

10

Dl5

D12D10

D9 D6 D18D15

B14 DU

D20

12

i

14

1—

16

D16

D17

Species richness S

Fig. 2.3. The relationship between diversity D (Simpson's index) and species richness S

in bumble bee communities (n = 18 sites). For the site legend and the site location, see

Fig. 2.1. The exact data are given in Table A2.5.

The study sites differed in exposition. Due to practical constraints, our sampling

method introduced further variation. However, none of the following factors affected or

correlated with local species richness S or diversity D: exposition (based on the four

major compass directions), maximal elevation of sampling, mean date of collection,

mean number of collectors, total collecting time, total number of bumble bees collected,

and proportion of bumble bees collected during the first visit (Table 2.2; only the

statistics for diversity D are given, the results for species richness S are qualitatively

similar).

25

Table 2.2. Effects of various site and sampling characteristics on, or their correlation

with the diversity D of the bumble bee communities (n = 18 sites). The raw data are

given in Table A2.1.

Factor Statistic* P

Exposition! F3>u =1.06 0.399

Elevation (max.) r = -0.035 0.891

Mean date of collection r = -0.271 0.278

Mean number of collectors $ tf = 1.036, df = 2 0.596

Total collecting time r = 0.187 0.457

Total number of bumble bees

collected §r = -0.279 0.262

Proportion of bumble bees

collected during the 1st visit 1r = -0.314 0.204

* If not indicated otherwise, the Pearson correlation coefficient is given.

f one-way ANOVA with fixed effect

$ Kruskal-Wallis one-way ANOVA

§ Spearman rank correlation

% data transformation f(x) = 2arcsin(V.x:)

A gynandromorph of the species Bombus pratorum was collected in Täsch (site 5,

Fig. 2.1) on July 17, 1997. The head and the thorax of this specimen were bilateral

mosaics: morphologically, the left side (dorsal view) was male, exhibiting typically

male characteristics such as, for example, an antenna with 13 segments, yellow facial

pile, a reduced mandible, and a hindleg without a pollen basket. The right side, on the

other hand, had typically female characteristics, such as an antenna with 12 segments,

black facial pile, a fully developed mandible, and a hindleg with a pollen basket. In

26

contrast, the gaster exhibited only female external morphology such as, for example, six

dorsal plates (tergites) and a sting.

Workers of B. pratorum (radial cell: 2.46 ± 0.06 mm, mean ± SE) were smaller

than conspecific males of this species (2.79 ± 0.04 mm; paired f-test: t = 3.77, df = 11, P

= 0.003). The gynandromorph's left wing (2.60 mm, male side) was intermediate in size

between males and workers (respectively, one-sample f-test: t = 4.45, df = 11, P =

0.001, and t = -2.37, df = 11, P = 0.037). However, the latter difference was no longer

significant after Bonferroni correction for multiple testing. The gynandromorph's right

wing (2.52 mm, female side) had the size of a worker wing and differed from male wing

size (respectively, t = -1.05, df = 11, P = 0.317, and t = 6.33, df = 11, P < 0.0005).

Discussion

Bumble bee species diversity in the Alps is high (Rasmont et al. 2000b): we found

more than a fourth (21 (28 %) out of 75 species) of the West Palaearctic Bombus

species (Rasmont et al. 2000b, excluding the subgenus Psithyrus) and two fifths (21 (40

%) out of 53 species) of the European Bombus species (Reinig 1981) in the subalpine

and alpine zones of an area of about 40 x 80 km. While this number is higher than the

one reported from northern Spain (24 bumble bee species over an altitudinal range from

sea level to 2200 m; Obeso 1992) it does not reach the extraordinary diversity reported

from Eyne, eastern Pyrenees (32 species within an area of 2000 ha over an altitudinal

range of 1400 to 2900 m; Rasmont et al. 2000a). There are 31 bumble bee species in

Switzerland (Amiet 1996), excluding the subgenus Psithyrus), of which 22 can be

expected to be found in the subalpine and alpine areas of the Valais. (B. cryptarum and

B. lucorum were counted as one species because we did not distinguish between these

two species.) The only expected species that we did not sample was B. alpinus, which is

27

classified as being rare and exhibits large fluctuations in population size across years

(Amiet 1996).

Local species richness, i.e., the number of species present at a site, ranged from 9-

17 bumble bee species (excluding the subgenus Psithyrus). This is higher than the local

species richnesses found in most previous studies of bumble bee communities (Ranta &

Vepsäläinen 1981, their Table 1; Hanski 1982a, his Table 1; Hanski 1982b, his Figure

3; Bowers 1985; Williams 1988; Obeso 1992; Stupf 1992; Dürrer & Schmid-Hempel

1995). However, here, the single study sites often covered a more heterogenous and

larger area, and a broader altitudinal range as compared to the studies cited above.

Ranta & Vepsäläinen (1981) pointed out that bumble bee communities are on average

more diverse in Europe than in North America. Our data corroborate this observation

for the bumble bee communities of the subalpine areas (Inouye 1977; Bowers 1985).

The large workenqueen ratios in our samples of B. sichelii and B. monticola, and

B. pyrenaeus (Table 2.1) indicate that the abundances found for these species are based

strongly on the non-reproductive worker caste. This is likely to result in an

overestimation of the effective population size relative to other species with smaller

workenqueen ratios. The three species concerned do not produce disproportionally large

colonies (von Hagen 1994). In the case of B. monticola, the large worker:queen ratio

can be explained by a relatively early phenology, as we already sampled many workers

during the first sampling round (Table 2.1).

According to our survey, the four most common bumble bee species in the

subalpine and alpine areas of southwestern Switzerland were B. pratorum, B. lucorum,

B. soroeensis, and B. wurflenii (Fig. 2.2). The four rarest bumble bee species in the

study area were B. subterraneus, B. argillaceus, B. sylvarum, and B. gerstaeckeri. These

observations correspond well with Amiet (1996). He labels only the three species that

had the lowest mean relative abundance in our study as "rare", and B. jonellus (fifth

lowest mean relative abundance) as "quite rare". B. gerstaeckeri is present throughout

28

the Swiss Alps (Amiet 1996; Erhardt 1996), however, it often occurs in low densities, is

patchily distributed (C. Gerloff, pers. obs.), and queens emerge from hibernation only in

summer (von Hagen 1994; Erhardt 1996; this study, Table 2.1). This pattern of

distribution reflects the spatial and temporal distribution pattern of flowering

monkshood and wolfsbane aconites (Aconitum lycoctonum and A. napellus), the flower

resources on which B. gerstaeckeri is specialized (Erhardt 1996; Rasmont et al. 2000a;

Utelli & Roy 2000). The restricted distribution of B. hypnorum, which is classified as

being "not rare" (Amiet 1996), can be explained by the fact that often only a minor part

of each site was covered by the species' preferred habitats (light forests and forest

edges, Amiet 1996). Based on the relative abundance found in this study (Fig. 2.2,

Table A2.4), the bumble bee species that potentially face the problems imposed by

small population size are B. subterraneus, B. argillaceus, B. sylvarum, and, to a lesser

degree, B. gerstaeckeri and B. jonellus.

Generally, abundant bumble bee species were present at all or almost all sites,

while species occuring in low numbers were present only at a few sites (Fig. 2.2). This

pattern has been observed in many species communities (e.g., Hanski 1982c; Brown

1984; Gaston & Lawton 1989; Hanski et al. 1993; Durrer & Schmid-Hempel 1995; and

references therein). The core and satellite species hypothesis (Hanski 1982c) assumes

that such a distribution results from extinction-colonization dynamics in a

metapopulation. However, the distribution of the bumble bee species in the Swiss Alps

as found here is not distributed bimodally as predicted by this hypothesis (Fig. 2.2,

Table 2.1). While a group of nine mainly abundant species with wide distribution could

be classified as "core species", no distinct group represented the "satellite species" with

a restricted distribution and low abundance. At a medium distribution level (presence at

about 40-60 % of the sites) locally abundant and rare species even overlapped. Previous

tests of the distributional pattern expected under the core and satellite species

29

hypothesis have yielded ambiguous results (Hanski 1982a, b; Williams 1988; Obeso

1992; Dürrer & Schmid-Hempel 1995; Rasmont et al. 2000a).

Several site and sampling characteristics, such as exposition, elevation, date of

collection, and sample size did not explain the large variation in species richness and

diversity found among sites (Fig. 2.3, Table 2.2). Possible reasons for the among-site

variation could be differences in habitat diversity, or in diversity and availability of

flower resources (Bowers 1985; Williams 1989). We did not measure these variables.

However, we did not observe any obvious differences in habitat and flower resource

diversity across sites during sampling. Although species richness and diversity were

positively correlated, the most and the least diverse site had a medium level of species

richness. The low diversities of sites 20 and 7 reflect low equitability (evenness) as a

consequence of a single species being strongly dominant.

Gynandromorphs, i.e., individuals that show a mosaic of female and male body

parts, are known from several insect orders (Coleoptera: Deuve 1992, Dermaptera:

Bailleul 1993, Diptera: Hadi et al. 1994, Ephemeroptera: da Silva & Pereira 1993,

Homoptera: Miller & Williams 1995, Lepidoptera: Josephrajkumar et al. 1998,

Odonata: Papazian 1997, Orthoptera: Barranco et al. 1995, and Psocoptera: Smithers

1996). For the aculeate Hymenoptera, the reports on gynandromorphs are reviewed in

Hoop (1964) and Nilsson (1987), and the possible mechanisms leading to

gynandromorphs are discussed in Roseler (1962), Milne (1985, 1986), and Nilsson

(1987). Generally, gynandromorphism is a rare phenomenon. However, in honey bees, a

gynandromorph-producing line has been selected (Rothenbuhler 1958; Rothenbuhler et

al. 1968), and in a natural ant population, high frequencies of gynandromorphs have

been found (Kinomura & Yamauchi 1994).

In bumble bees, six gynandromorphic specimens have been reported to date (two

each of Bombus lapidarius and B. pascuorum, and one each of B. wurflenii and B.

sylvarum, reviewed in Röseler 1962). The gynandromorph of B. pratorum presented in

30

this study adds a fifth bumble bee species in which gynandromorphs have been

described. All the bumble bee gynandromorphs, with the exception of one B. lapidarius

specimen (Stöckhert 1921, cited in Röseler 1962), were bilateral gynandromorphs, i.e.,

they exhibited female characteristics on one body side and male characteristics on the

other body side. Two specimens - one of B. lapidarius (Sichel 1858) and one of B.

wurflenii (Ritsema 1881) - exhibited a similar mosaic of female and male tissue, except

that their gasters were male. The classification of female and male body parts has been

based only on external morphology (with the exception of one B. pascuorum

gynandromorph (Röseler 1962)). However, the sex characteristics of the internal tissue

do not always correspond to the sex characteristics of the external cuticle (Milne 1986).

While the B. lapidarius gynandromorph described by Sichel (1858) had the size

of a small queen and was notably larger than a typical male, the B. pascuorum

gynandromorph described by Röseler (1962) had the size of a very small worker. The B.

pratorum gynandromorph presented here had the size of a worker and was smaller than

a typical male. In the two gynandromorphs, in which the wing sizes of the female and

male side were measured, the sizes of the two wings were too similar as to assign each

wing to a different caste according to wing size (Röseler 1962; this study). The intra-

individual wing size difference of the B. pratorum gynandromorph described here was

smaller than the maximal intra-individual wing size difference observed in B. terrestris

workers (0.08 mm and 0.12 mm, respectively; B. Ottmer, unpubl. data).

Our study confirmed the high diversity of bumble bee species in the subalpine and

alpine zones in the Alps. The distributional data presented suggest that many Alpine

bumble bee species have population sizes that are presumably large enough so as not to

impose problems, such as inbreeding. However, a few species were so rare as to be

potentially threatened by the consequences of small population size.

31

Acknowledgements

We thank D. Frey for computation of the species saturation curve, B. Baer, E.

Benelli, J.-R. Escher, G. Gasser Gerloff, U. Lengwiler, S. Liersch, P. Mutikainen, C.

Reber Funk, J. Taskinen, and A.-B. Utelli for help with the field work, F. Amiet and A.

Müller for the re-identification of several specimens, and M. Brown for comments on

the manuscript. The study was financially supported by a grant from the Swiss Federal

Institute of Technology (ETH) Zürich (no. 0-20-010-95 to PSH).

References

Alford, D.V. 1975. Bumblebees. Davis-Poynter, London, UK.

Amiet, F. 1996. Hymenoptera Apidae, 1. Teil. Schweizerische Entomologische

Gesellschaft, Neuchâtel, Switzerland.

Bailleul, S. 1993. Observations on the earwigs from the Cher region (Dermaptera).

Entomologiste 49:268.

Barranco, P., Cabrero, J., Camacho, J.P.M. & Pascual, F. 1995. Chromosomal basis for

a bilateral gynandromorph in Pycnogaster inermis (Rambur, 1838) (Orthoptera,

Tettigoniidae). Contrib. Zool. 65:123-127.

Bertsch, A. 1984. Foraging in male bumblebees {Bombus lucorum L.): maximizing

energy or minimizing water load? Oecologia 62:325-336.

Bowers, M.A. 1985. Bumble bee colonization, extinction, and reproduction in subalpine

meadows in northeastern Utah. Ecology 66:914-927.

Brown, J.H. 1984. On the relationship between abundance and distribution of species.

Am. Nat. 124:255-279.

32

Calow, P. 1998. The encyclopedia of ecology & environmental management. Blackwell

Science, Oxford, UK.

da Silva, E.R. & Pereira, S.M. 1993. Mayflies of Serra dos Orgaos, Rio de Janeiro State,

Brazil: I. Description of a gynandromorph of the genus Baetis Leach, 1815

(Ephemeroptera: Baetidae). Anais Acad. Brasil. Ciencias 65:77-79.

Delaplane, K.S. & Mayer, D.F. 2000. Crop pollination by bees. CABI Publishing,

Oxon, UK.

Deuve, T. 1992. Segmentary origin of male and female ectodermic genitalia in insects:

New data from a gynandromorph of Coleoptera. Compt. Rend. Acad. Sei. Ser. Ill

Sei. Vie 314:305-308.

Dürrer, S. & Schmid-Hempel, P. 1995. Parasites and the regional distribution of bumble

bee species. Ecography 18:114-122.

Ellstrand, N.C. & Elam, D.R. 1993. Population genetic consequences of small

population size: Implications for plant conservation. Annu. Rev. Ecol. Syst.

24:217-242.

Erhardt, G.M. 1996. Blütenökologische Untersuchungen an Aconitum lycoctonum

(Ranunculaceae) und Bombus gerstaeckeri (Hymenoptera, Apidae). Diploma

thesis. University of Zürich, Switzerland.

Frankham, R. 1995. Conservation genetics. Annu. Rev. Genet. 29:305-27.

Frankham, R. & Ralls, K. 1998. Inbreeding leads to extinction. Nature 392:441-442.

Gaston, K.J. & Lawton, J.H. 1989. Insect herbivores on bracken do not support the

core-satellite hypothesis. Am. Nat. 134:761-777.

Hadi, U.K., Aoki, C, Saito, K. & Kanayama, A. 1994. Sexual mosaics in two blackfly

species (Diptera: Simuliidae) collected from Shizuoka Prefecture, Japan. Jpn. J.

Sanit. Zool. 45:297-299.

Hanski, I. 1982a. Communities of bumblebees: testing the core-satellite species

hypothesis. Ann. Zool. Fennici 19:65-73.

33

—. 1982b. Structure in bumblebee communities. Ann. Zool. Fennici 19: 319-326.

—. 1982c. Dynamics of regional distribution: the core and satellite species hypothesis.

Oikos 38:210-221.

Hanski, I., Kouki, J. & Halkka, A. 1993. Three explanations of the positive relationship

between distribution and abundance of species, pp. 108-116 in R.E. Ricklefs & D.

Schlüter, eds. Species diversity in ecological communities: historical and

geographical perspectives. University of Chicago Press, Chicago, IL, USA.

Hedrick, P.W. & Kalinowski, S.T. 2000. Inbreeding depression in conservation biology.

Annu. Rev. Ecol. Syst. 31:139-162.

Hoop, M. 1964. Ein Gynander von Ancistrocerus ichneumonideus Ratzeburg. Sehr.

Naturw. Ver. Schlesw.-Holst. 35:28-32.

Inouye, D.W. 1977. Species structure of bumblebee communities in North America and

Europe. Pp. 35-40 in W.J. Mattson, ed. The Role of Arthropods in Forest

Ecosystems. Springer, New York, NY, USA.

Josephrajkumar, A., Subrahmanyam, B. & Ramamurthy, V.V. 1998. Gynandromorph of

Helicoverpa armigera (Lepidoptera: Noctuidae). Entomol. News 109:288-292.

Kearns, CA., Inouye, D.W. & Waser, N.M. 1998. Endangered mutualisms: the

conservation of plant-pollinator interactions. Annu. Rev. Ecol. Syst. 29:83-112.

Kinomura, K. & Yamauchi, K. 1994. Frequent occurrence of gynandromorphs in the

natural population of the ant Vollenhovia emeryi (Hymenoptera: Formicidae).

Insect. Soc. 41:273-278.

Landolt, E. 1984. Unsere Alpenflora. 5th edn. Verlag Schweizer Alpen-Club,

Wallisellen, Switzerland.

Miller, D.R. & Williams, DJ. 1995. Systematic revision of the family Micrococcidae

(Homoptera: Coccoidea), with a discussion of its relationships, and a description

of a gynandromorph. Boll. Laborat. Entomol. Agrar. Filippo Silvestri 50:199-247.

34

Milne, C.P., Jr. 1985. Estimating primordial cell numbers giving rise to honeybee adult

structures. J. Apic. Res. 24:7-12.

—. 1986. Cytology and Cytogenetics, pp. 205-233 in T.E. Rinderer, ed. Bee genetics

and breeding. Academic Press, Orlando, FL, USA.

Müller, C.B. & Schmid-Hempel, P. 1992a. Correlates of reproductive success among

field colonies of Bombus lucorum: the importance of growth and parasites. Ecol.

Entomol. 17:343-353.

—. 1992b.Variation in life-history pattern in relation to worker mortality in the bumble¬

bee, Bombus lucorum. Funct. Ecol. 6:48-56.

Nilsson, G.E. 1987. A gynandromorphic specimen of Evylaeus albipes (Fabricius)

(Hymenoptera, Halictidae) and a discussion of possible causes of

gynandromorphism in haplodiploid insects. Notulae Entomol. 67:157-162.

Norusis, MJ. 1994. SPSS Advanced statistics 6.1. SPSS, Chicago, IL, USA.

Obeso, J.R. 1992. Geographic distribution and community structure of bumblebees in

the northern Iberian Peninsula. Oecologia 89:244-252.

Owen, R.E. 1988. Body size variation and optimal body size of bumble bee queens

(Hymenoptera: Apidae). Can. Entomol. 120:19-28.

—. 1989. Differential size variation of male and female bumblebees. J. Hered. 80:39-43.

Papazian, M. 1997. A morphological anomaly with gynandromorphic aspect in

Gynacantha kirbyi Kruger, 1898 (Odonata, Aeshnidae). Bull. Soc. Entomol.

France 102:103-109.

Pekkarinen, A. & Teräs, I. 1993. Zoogeography of Bombus and Psithyrus in

northwestern Europe (Hymenoptera, Apidae). Ann. Zool. Fennici 30:187-208.

Ranta, E. & Vepsäläinen, K. 1981. Why are there so many species? Spatio-temporal

heterogeneity and northern bumblebee communities. Oikos 36:28-34.

Rasmont, P., Durieux, E.-A., Iserbyt, S. & Baracetti, M. 2000a. Why are there so many

bumblebee species in Eyne (France, Pyrénées-Orientales, Cerdagne)? pp. 83-92 in

35

M.J. Sommeijer & A. de Ruijter, eds. Insect pollination in greenhouses.

University of Utrecht, The Netherlands.

Rasmont, P., Verhaeghe, J.C., Rasmont, R. & Terzo, M. 2000b. West-Palaearctic

bumblebees, pp. 93-99 in M.J. Sommeijer & A. de Ruijter, eds. Insect pollination

in greenhouses. University of Utrecht, The Netherlands.

Reinig, W.F. 1981. Synopsis der in Europa nachgewiesenen Hummel- und

Schmarotzerhummelarten (Hymenoptera, Bombidae). Spixiana 4:159-164.

Ritsema, C. 1881. Verslag van de veertiende Wintervergadering der Nederlandsche

Entomologische Vereeniging, gehouden te Leiden op 19 December 1880.

Tijdschr. Entomol. 24:CXI.

Röseler, P.-F. 1962. Über einen Fall von Gynandromorphismus bei der Hummel

Bombus agrorum Fabr. Mitt. bad. Landesver. Naturkunde u. Naturschutz. N. F.

8:289-303.

Rothenbuhler, W.C. 1958. Progress and problems in the analyses of gynandromorphic

honey bees. Proc. Tenth Int. Congr. Entomol. 2:867-873.

Rothenbuhler, W.C, Kulincevic, J.M. & Kerr, W.E. 1968. Bee genetics. Annu. Rev.

Genet. 2:413-438.

Saccheri, I., Kuussaari, M., Kankare, M., Vikman, P., Fortelius, W. & Hanski, I. 1998.

Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494.

Sichel, J. 1858. Note sur un cas d'hermaphrodisme observé chez un Bombus lapidarius.

Ann. Soc. Entomol. France 6:CCXLVII-CCXLIX.

Smithers, C.N. 1996. A gynandromorph of Ectopsocus australis Schmidt and Thornton

(Psocoptera: Ectopsocidae) from Australia. Aust. Entomol. 23:93-95.

Stockhert, F.K. 1921. Über einen Fall von frontaler Gynandromorphie bei Bombus

lapidarius L. (Hym.). Z. wiss. Ins.-biol. 16:132-135.

36

Stupf, R. 1992. Vorkommen und Verbreitung von Bombus-Arten (Apidae,

Hymenoptera) und ihrer Parasiten in einer Alpenregion. Diploma thesis. Swiss

Federal Institute of Technology (ETH), Zürich, Switzerland.

Utelli, A.-B. & Roy, B.A. 2000. Pollinator abundance and behavior on Aconitum

lycoctonum (Ranunculaceae): An analysis of the quantity and quality components

of pollination. Oikos 89:461-470.

von Hagen, E. 1994. Hummeln: bestimmen, ansiedeln, vermehren, schützen. Naturbuch

Verlag, Augsburg, Germany.

Williams, P.H. 1982. The distribution and decline of British bumble bees (Bombus

Latr.). J. Apic. Res. 21:236-245.

—. 1988. Habitat use by bumble bees (Bombus spp.). Ecol. Entomol. 13:223-237.

—. 1989. Why are there so many species of bumble bees at Dungeness? Biol. J. Linn.

Soc. 101:31-44.

—. 1998. An annotated checklist of bumble bees with an analysis of patterns of

description (Hymenoptera: Apidae, Bombini). Bull. Nat. Hist. Mus. Lond. (Ent.)

67:79-152.

37

3 Inbreeding Depression and Family Variation in a Social

Insect, Bombus terrestris (L.) (Hymenoptera: Apidae)

(GerloffC.U. & Schmid-Hempel P., in revisionfor Evolution)

Abstract

If reproduction within a population is dominated by a few, successful family lineages,

and if dispersal is limited, these lineages face a higher risk of inbreeding as they become

more successful. This could be the case in bumble bees, where only a few colonies

contribute almost all of the offspring at the end of a season. Because mating typically

occurs within the local population, successful families are likely to experience episodic

inbreeding in a given year, i.e., an increased likelihood of mating among relatives. To

understand how bumble bees cope with these inbreeding episodes we compared the fitness

of colonies resulting from brother-sister matings to that of outbred colonies - among and

within families of Bombus terrestris (L.). From a total of 371 mated queens (188 inbred and

183 outbred matings), we raised 76 colonies (33 inbred and 43 outbred, respectively). We

found large among-family variation to be a key feature of all traits measured at different

stages of the life cycle. For example, the probability of surviving hibernation and colony

foundation varied significantly among maternal lines (the hibernating queens' mother

colonies). In contrast, mating with a related male did not affect a queen's hibernation

survival but it did reduce her success in founding a colony. Hibernation success of queens

was significantly affected by the family of their mates and the date of mating; queens that

38

mated and hibernated early in the season survived hibernation more frequently than those

that mated later. Maternal family lines differed significantly in their response to inbreeding

with respect to colony size, caloric investment into sexuals, and the Shaw-Mohler fitness

index. In some family lines the inbred colonies were even fitter than the outbred colonies

(outbreeding depression). Consequently, averaged over the entire experimental population,

there was no significant inbreeding depression. In bumble bees, the strong variation among

family lines is likely to reflect differences in genetic background, in carried-over maternal

effects, or an interaction between progeny genotype and maternal effects. Thus, as a

population, B. terrestris can cope with episodic inbreeding. However, among-family

variation in the effects of episodic inbreeding for important life-history traits intensifies the

differential representation of families in the population.

Key words

among-family variation; Bombus terrestris; diploid males; Hymenoptera; inbreeding

depression; outbreeding depression.

Introduction

Within any one year, insect populations are probably often dominated by a few,

successful family lineages who happen to be adapted best to the current environmental

conditions. If environmental conditions remain similar over time, or if mating within a

season is indiscriminate to lineage, successful family lineages face the dilemma that, as

they become more frequent, inbreeding becomes more likely.

39

In bumble bees, as in many other social insects, this situation must be an important

characteristic of their evolutionary ecology. Effective population sizes of bumble bees are

in the order of a few hundred (A. Widmer, unpubl. data) in the best of cases, and island

populations show significant departure from random mating and strong genetic

differentiation (Estoup et al. 1996; Widmer et al. 1998). In any given year, although spring

queens of many family lines will be present at the start of the season, reproduction in the

local population appears to be dominated by a few, very successful colonies that produce

most, or all of the sexuals, especially as far as females (the daughter queens) are concerned

(e.g., Donovan & Weir 1978; Müller & Schmid-Hempel 1992b; Imhoof, 1999). Because

bumble bees mate within the local population (Svensson 1979), the consequence of such

local, biased reproduction could be an increased likelihood of inbreeding within the

successful families, especially when some of the spring queen lines are more successful

than others (B. Baer et al., unpubl. data). Hence, episodic inbreeding is expected to be

common in bumble bees. This local structure is diluted by migration after mating, such that

little genetic structure of the entire population at a larger scale results (Estoup et al. 1996;

Widmer & Schmid-Hempel 1999). Whether or not episodic inbreeding leads to significant

inbreeding depression in these species is unknown but it is important to understand how

organisms are able to cope with inbreeding episodes.

Bumble bees are annual, primitively eusocial Hymenoptera that live mainly in

temperate regions (e.g., Alford 1975; Morse 1982). Queens emerge from hibernation in

spring and found colonies on their own. During colony foundation, mortality of incipient

colonies is high and unfavorable weather conditions then can reduce the bumble bee

population dramatically (Harder 1986; Sladen 1989; Pamilo & Crozier 1997). Workers are

the first offspring, and they proceed to take over the tasks of brood care and foraging.

Towards the end of the colony cycle in late summer, sexuals (young queens and males) are

40

produced. After mating, the young queens enter hibernation while the founding queen, the

workers, and the males of a colony die. The following spring, the queens which survived

hibernation give rise to the next generation (Free & Butler 1959; Alford 1975).

The insect order Hymenoptera is a particularly interesting group to investigate the

effects of inbreeding because of the complementary sex determination system (e.g.,

Whiting 1940; Cook & Crozier 1995). Under this system, sex is determined by the alleles at

a single (or perhaps several) multi-allelic locus, the sex locus. Heterozygous individuals

develop into females, hemizygous (haploid) individuals into haploid males, and

homozygous individuals into diploid males. Thus, males develop from unfertilized eggs,

while females or diploid males arise from fertilized eggs (Crozier 1975), a system also

referred to as "haplodiploidy". Haplodiploidy is associated with smaller effective

population size (Crozier 1985; Pamilo & Crozier 1997) and perhaps reduced heterozygosity

as compared to diplodiploid species (e.g., Pamilo & Crozier 1981; Berkelhamer 1983;

Graur 1985). This could make haplodiploids more likely to experience inbreeding (but see

Crozier 1975; Crozier 1985; Adamson 1989). In addition, with the complementary sex

determination system, inbreeding can lead to the production of diploid males because of

homozygosity at the sex locus. Diploid males are indeed known to occur in a large number

of hymenopteran species (for reviews see Stouthamer et al. 1992; Cook 1993; Cook &

Crozier 1995; Crozier & Pamilo 1996, their Table 1.4).

The problem of diploid male production is probably most obvious in the social

Hymenoptera, such as in bumble bees, because diploid males develop from eggs that

should have turned into workers or young queens. This leads to a reduction in the worker

force with potentially disastrous consequences for colony growth and reproduction (e.g.,

Plowright & Pallett 1979; Ross & Fletcher 1986; but see Duchateau et al. 1994). Diploid

males usually do not sire viable offspring or else have inviable or infertile triploid offspring

41

(Stouthamer et al. 1992; Cook 1993; Cook & Crozier 1995). In bumble bees, the production

of diploid males is considered a major cost of inbreeding, because diploid males are reared

to adulthood (e.g., Garöfalo 1973; Duchateau et al. 1994) in contrast to, for example, the

honey bee where the diploid male larvae are selectively removed by the workers (Woyke

1963).

When assessing inbreeding depression, family variation has to be considered, as

estimates based on the family level differ from those based on the population level

(Holsinger 1988, 1991; Uyenoyama et al. 1993). The difference between these two

inbreeding depression estimates might be substantial in social insects, where colony and

worker characteristics typically vary among colonies within a population (for ants: Pamilo

& Rosengren 1984; Johnston & Wilson 1985; Porter & Tschinkel 1985; Gordon 1991;

honey bees: Page et al. 1991; Breed et al. 1995; and wasps: Queller et al. 1993). Some of

these characteristics can be selected for, e.g., disease resistance (Kulincevic 1986), or show

genetic variance (e.g., Page et al. 1989; Oldroyd et al. 1991). In bumble bees also, there is

considerable variation among colonies in important parameters, such as various colony life-

history traits (Webb 1961 cited in Michener 1964; Donovan & Weir 1978; Owen et al.

1980; Owen & Plowright 1982; Duchateau & Velthuis 1988; Müller & Schmid-Hempel

1992b; Beekman & van Stratum 1998), individual morphometries (Owen 1988, 1989;

Owen & Harder 1995), and susceptibility and defence to parasite infections (Konig&

Schmid-Hempel 1995; Imhoof & Schmid-Hempel 1999; Schmid-Hempel & Loosli 1998;

Schmid-Hempel et al. 1999). The sources of this variation are both genetic (Schmid-

Hempel & Loosli 1998; Schmid-Hempel et al. 1999) and environmental (Owen 1988;

Beekman & van Stratum 2000), or an interaction of the two (Owen 1989; Owen & Harder

1995). In spite of this large variation, previous studies on inbreeding in bumble bees did not

take among-family variation into account (Plowright & Pallett 1979; Duchateau et al. 1994;

42

Duchateau & Marien 1995), except for Beekman et al. (1999), who, however, did not take

into account the colonies that produced diploid males.

We studied variation among family lines for important life-history traits in the

bumble bee, Bombus terrestris (L.). In addition, we were interested in how episodic

inbreeding affects these measures. Episodic inbreeding is mimicked in our experiment by

brother-sister matings. Because the amount of inbreeding depression typically varies among

different life stages (for discussion, see Uyenoyama et al. 1993), we measured the effect of

inbreeding on several important traits of the colony life cycle. These include hibernation

success and colony foundation success for the queens mated to a related or unrelated male,

and colony size, quantity, and quality of reproductive output for the resulting colonies.

Bumble bees play an important role in commercial breeding for greenhouse pollination of

crops (e.g., Beekman et al. 1998) and effects of inbreeding are also interesting from this

applied perspective.

Materials and Methods

Experimental design

To generate inbred and outbred colonies, females and males of known pedigree were

mated. For this purpose, spring queens of B. terrestris were collected in northern

Switzerland in 1997 and they were allowed to establish a colony in the laboratory ("mother

colonies"; Fig. 3.1). Between June 18 and August 9, 1997, we mated sexual offspring of

these mother colonies in the laboratory. The queens used for mating originated from 19

different mother colonies ("maternal family lines"), while the males came from 28 different

mother colonies ("paternal family lines"). To produce inbred colonies we mated daughter

43

queens with their brothers. To produce outbred colonies we mated daughter queens singly

with unrelated males. For those pairings, males were allocated randomly with respect to

paternal family line. For matings, we placed each queen together with two males in a clear

plastic box (12 x 9.5 x 7 cm). The two males offered were always brothers. During

copulation, we removed the non-copulating male. After copulation, we freeze-killed the

mate to take individual size measures later. From mating until hibernation, we kept the

queens individually in the plastic boxes under a natural light regime, and provided them

with sugar water and pollen ad libitum. On average, 19.5 ±1.2 queens (mean ± SE) per

maternal family line (range 2-12 mated sister queens per treatment, see Table 3.1) were

successfully mated. Two days after mating we transferred the queens into a climate cabinet

for a transition period of 6 ± 0.1 days (mean ± SE, range 4-14 days) at 13°C. For

hibernation we placed the queens individually into perforated plastic tubes (length 6.2 cm,

diameter 2.2 cm) stopped at each end, and grouped them according to mating date in

wooden boxes (range 5-27 queens, usually from several maternal family lines) into a

climate cabinet at 6°C in continuous darkness. A water-filled bowl inside the climate

cabinet humidified the circulating air. We hibernated a total of 371 mated queens (188

inbred and 183 outbred matings) for a period of 107 days to control for effects of

hibernation duration (e.g., Beekman et al. 1998; Beekman & van Stratum 2000).

44

Matemal Maternal Paternal

family line 1 family line 2 family line 1

\

DDD

\ \

DDD DDD

Mother

colonies

Mated

queens

Hibernation

Experimentalcolonies

Fig. 3.1. Experimental design. "Maternal family line" refers to the mother colony of a

mated queen (pattern in the left half of the queen symbol). The "paternal family line" refers

to the mother colony of the queen's mate (pattern in the right half of the queen symbol). For

each maternal family line, half of the queens were mated to one of their brothers (inbred

treatment), the other half to an unrelated male (outbred treatment). After hibernation the

mated queens produced the experimental colonies.

Rearing of experimental colonies

The queens surviving hibernation were placed individually into a rearing box (acrylic

glass, 12.5 x 7.5 x 5.5 cm) in a climate chamber set at standard breeding conditions (29°C,

60 % relative humidity, red light). We provided sugar water and pollen daily ad libitum so

that queens could start their colonies. We discarded queens that did not lay eggs within

eight weeks or that did not successfully rear at least one offspring within eight weeks after

egg laying. This limit was chosen because the probability of a queen laying eggs more than

six weeks after hibernation is very low (Beekman et al. 1999). When the incipient colony

had produced five offspring, we tranferred it into an observation hive (Pomeroy &

Plowright 1980) with a separate feeding box attached. Colonies that produced less than five

45

offspring were kept in the rearing box until the end of the colony cycle. We culled 10 % of

the workers in regular intervals of eight days to mimic worker mortality in the field (Brian

1952; Rodd et al. 1980; Schmid-Hempel & Heeb 1991).

To measure reproductive success, we removed the sexuals, i.e., males and young

queens daily when they were seen foraging in the feeding box. This mimics the natural

process of sexuals leaving the colony at the end of the season (e.g., Alford 1975). We used

the sexuals for further matings or freeze-killed them to take individual measures later. We

terminated the colony cycle 26 days after the queen's death, i.e., after the average

developmental time of a B. terrestris male under rearing conditions similar to ours

(Duchateau & Velthuis 1988). Thus, our measure of reproductive success takes the queen's

life span into account. In total, we raised 76 colonies (33 inbred and 43 outbred,

respectively) representing 15 maternal and 22 paternal family lines.

Variables measured

Hibernation success.—We classified queens as having successfully hibernated when

they survived at least the first day after the end of hibernation. A total of 42 (11.3 %) of the

queens survived hibernation but died within the first day of rearing and thus were not

considered successful. Results remain qualitatively the same when these queens are

included in the data.

Body size.—Wing size correlates with fresh weight in bumble bees (Bertsch 1984;

Müller & Schmid-Hempel 1992a). Wing size of hibernating queens and their mates was

measured as length of the radial cell of the right forewing (Owen 1988, 1989; Müller &

Schmid-Hempel 1992a). To assess the average size of the sexuals produced by a colony, we

measured, where possible, the wing size of the first 20 young queens and the first 20 males

produced (on average, 13.2 ± 1.5, mean ± SE, range 5-20 queens per colony; and 19.3 ± 0.3

46

haploid males, range 15-20 per colony). Because of the generally positive effect of body

mass on hibernation survival (Horber 1961; Holm 1972; Owen 1988; Beekman et al. 1998),

our measure of size reflects sexual quality and therefore provides a way to assess whether

colonies trade off quantity for quality of sexuals.

Colonyfoundation.—We considered that a queen founded a colony successfully if

she reared at least one offspring within eight weeks after egg laying.

Colony size.—We counted the number of workers a colony produced as the number

of workers present at the end of the colony cycle (dead and alive), plus the workers

removed at the regular cullings.

Caloric investment into sexuals.—Due to the larger body size, the production of a

queen is substantially more costly than the production of a male. The numerical sex ratio

does therefore not reflect the reproductive investment. We calculated the caloric investment

per colony as the product of the number of sexuals produced and the average caloric

content of a sexual offspring (for B. terrestris: 7.83 kJ per queen and 2.35 kJ per haploid

male; Beekman & van Stratum 1998). Based on caloric content, the queen : male-ratio was

3.33, in contrast to 2.11 based on the average dry weight of sexuals (Duchateau & Velthuis

1988). While 65 colonies (85.5 %) produced males, only 33 colonies (43.4 %) produced

queens. Queen production is comparable to other reports for B. terrestris in the field and in

the laboratory (e.g., 22.2 % (Shykoff & Müller 1995), and 69.2 % (Duchateau & Velthuis

1988), respectively). Colonies that produce diploid males are characterized by simultaneous

worker and male production at about equal rates from the first brood on (Duchateau et al.

1994), and hereafter are referred to as "diploid male colonies". For practical reasons, it was

impossible to screen the genotype of all males in the diploid male colonies. Therefore, we

derived the number of haploid males as follows: the sum of workers and young queens

produced was calculated. Then, based on the complementary sex determination system, we

47

assumed that diploid males would arise at an equal rate to the diploid females (Duchateau

et al. 1994). In B. terrestris, the diploid male larvae are as viable as the diploid female

larvae (Duchateau et al. 1994). Hence, the number of haploid males is given by the

difference of total males (haploid plus diploid) minus the number of females (workers plus

young queens).

Shaw-Mohler fitness index.—The relative fitness of a colony (W) is defined as W =

f/F + m/M (Charnov 1982). /and m are the number of queens and males produced by the

colony under consideration, respectively. F and M are the total numbers of queens and

males produced in the experimental population, respectively. Here, the experimental

population, i.e., the breeding group of interest, consists of the colonies included in the

statistical analysis. For each colony, m equals all the males collected until 26 days after the

queen's death. In diploid male colonies we calculated the number of haploid (functional)

males as described above.

Statistical analyses

Statistical analyses were performed with SPSS 6.1.1 (Norusis 1994). Due to the large

variation among maternal family lines we calculated the population means ± SE on the

basis of family line averages. For colony size, caloric investment into sexuals, and the

Shaw-Mohler index, we considered only the maternal family lines included in the statistical

analyses to estimate the means. Eight family lines produced too few colonies for

meaningful analysis and thus were excluded. The breeding scheme allowed discrimination

of the effects of both maternal family line (i.e., colonies arising from sister queens) and

paternal family line (i.e., colonies arising from brothers, mated with unrelated queens) only

for hibernation survival. For the other traits we were able to analyze the effects of the

maternal family line only. In logistic regressions, we used hibernation survival or colony

48

foundation success as dependent variables (yes/no) and included only family lines with at

least four queens per crossing treatment. We removed variables from the full model (Table

3.2) by the backward (stepwise) likelihood ratio method. The three colony performance

measures are dependent on each other: colony size influences the number of sexuals

produced, which is the basis for caloric investment and Shaw-Mohler index calculations.

For these measures we calculated a MANOVA in which we included family lines with at

least two colonies per crossing treatment. We log-transformed the data, as recommended by

Johnston and Schoen (1994) for testing family line differences in inbreeding depression.

We calculated a mixed-model MANOVA with single colonies nested within the interaction

between maternal family line and crossing treatment and determined the error terms

according to Zar (1996). We checked the data for normality and heteroscedasticity of

variances.

Queens that start their colonies early in the season typically have larger colonies and

more sexuals than colonies started later in the season (Müller & Schmid-Hempel 1992b).

Thus, we included the date of mating as a covariate in all analyses except for the

correlations of number and size of the sexuals produced. As the effect of the mating date

was significant (P < 0.05) only for hibernation survival we excluded this covariate from the

analyses not concerning hibernation survival. Recall that duration of hibernation was kept

constant. Thus, the date of mating was the day our experimental protocol started, reflecting

roughly the date a queen had emerged from a mother colony and was ready to mate. Queen

size (mass) affects hibernation survival (Horber 1961; Holm 1972; Owen 1988; Beekman et

al. 1998). Hence we included size as a covariate in the analyses of hibernation survival.

Due to insufficient sample size, we did not include the interaction between maternal family

line and size of hibernating queen and between paternal family line and size of hibernating

queen's mate in the logistic regression models. The size of the hibernating queen and the

49

interaction between size and crossing treatment, and the size of the hibernating queen's

mate had no significant effect (P > 0.05) when analyzing the maternal and the paternal

family lines, respectively. These terms were excluded from the models. Nevertheless, the

size of the hibernating queens differed significantly among maternal family lines (Kruskal-

Wallis one-way ANOVA, f = 34.965, df = 18, n = 336, P < 0.0001).

For the variables hibernation survival, colony foundation, colony size, caloric

investment into sexuals, and Shaw-Mohler index, we calculated a value for the fitness

reduction due to the crossing treatment according to Agren and Schemske (1993). For

simplicity, we use the term "inbreeding depression" for all these estimates, although the

queens considered for hibernation survival and colony foundation were not inbred

themselves but had mated either with a related or unrelated male. In the following, the

variable w is used to denote any fitness component such as percentage of queens surviving

hibernation, percentage of surviving queens founding a colony, mean number of workers,

mean caloric investment into sexuals, or mean Shaw-Mohler index. Index i refers to the

inbred treatment, index o to the outbred treatment. Inbreeding depression was estimated as

5=1- Wi/w0. In case of queens or colonies in the outbred treatment performing better than

the ones in the inbred treatment, Ô ranges from 0 to 1. If the inbred treatment performed

better, outbreeding depression was calculated as <5 = w0/ w, - 1, resulting in values for S

ranging from -1 to 0. Overall inbreeding depression was calculated by averaging the

estimates for <5 at the family level.

We report significance values of two-tailed tests. We used the step-up sequential

Bonferroni correction (Hochberg 1988) for multiple comparisons of the hibernation

survival data (maternal and paternal family line), and the correlations of number and size in

sexuals.

50

Results

Hibernation survival

Half of the mated queens survived hibernation (186 queens, 50.1 %), but

independently of crossing treatment: 49.6 ± 7.2 % (mean ± SE) and 53.1 ± 6.6 % for the

inbred and the outbred matings, respectively. (Classifying the queens that died within one

day after the end of hibernation as having hibernated successfully, raised the values to 61.2

± 6.7 % and 63.7 ± 7.2 %, respectively.) However, the probability of surviving hibernation

varied significantly from 0 to 100 % among maternal family lines, i.e., the hibernating

queens' mother colonies (Tables 3.1, 3.2 A). In addition, the logistic regression model

showed a clear trend for queens that emerged, mated, and hibernated early in the season to

survive hibernation more frequently than queens that mated later (Julian date of mating;

Fig. 3.2 A, Table 3.2 A). This effect was very strong as the odds of a queen surviving

hibernation decreased by 10 % every day (odds ratio e = 0.902). Significances remained

after step-up sequential Bonferroni correction. Although the interaction between maternal

family line and crossing treatment was not significant and thus was not included in the final

logistic regression model (Table 3.2 A), the range of inbreeding depression 8 from -1 to 1

for hibernation survival reflects large variation among maternal family lines in their

reaction to the crossing treatment (Table 3.4).

51

Table 3.1. Hibernation survival and colony foundation success of the bumblebee, B.

terrestris, according to family line and crossing treatment.

Maternal Number of Hibernation survival Colony foundation

family queens mated [%] [%]

line inbred outbred inbred outbred inbred outbred

1 10 11 50.0 81.8 20.0 11.1

2 11 10 90.9 90.0 20.0 33.3

3 11 9 0.0 0.0 — —

4 11 11 100.0 63.6 72.7 42.9

5 9 7 22.2 28.6 50.0 0.0

6 2 2 100.0 100.0 0.0 0.0

7 11 11 54.5 72.7 50.0 62.5

8 11 12 81.8 58.3 11.1 71.4

9 11 11 36.4 45.5 0.0 40.0

10 11 11 9.1 0.0 — —

11 11 11 72.7 63.6 87.5 100.0

12 5 5 20.0 80.0 0.0 25.0

13 11 11 0.0 9.1 — —

14 12 11 58.3 63.6 14.3 14.3

15 11 11 45.5 54.5 40.0 33.3

16 7 6 28.6 33.3 50.0 100.0

17 11 11 63.6 72.7 42.9 75.0

18 11 11 54.5 45.5 0.0 20.0

19 11 11 54.5 45.5 50.0 80.0

mean 9.89 9.63 49.61 53.07 31.78 44.30

(SE) (0.58) (0.61) (7.16) (6.62) (6.88) (8.33)

52

(A) Maternal family line

c3

3

o

jo

O

43O

ex.

1-1

0.75-

0.5-

0.25'

0-

—i 1 1 1 1 1

170 180 190 200 210 220 230

Julian date ofmating

(B) Paternal family line

>

3

o

O

•8

2

H

0.75-

0.5-

0.25-

0i 1 1 1 1 f 1 1

160 170 180 190 200 210 220 230

Julian date of mating

Fig. 3.2. (A) Probability of hibernation survival of bumble bee queens in relation to Juliandate of mating (days starting on January 1) for maternal lines. Lines show the predictions ofthe logistic regression models for different maternal family lines with at least four differentdates of mating (cf. Table 3.2). (B) Probability of hibernation survival for paternal lines.Lines show the predictions of logistic regression models for outbred matings of differentpaternal family lines (mother colony of the queen's mate). Only paternal family lines withat least four different dates of mating are shown.

53

Table 3.2. Effects of family line (maternal or paternal), crossing treatment, and mating date

on hibernation survival (A) and colony foundation success (B) in the bumble bee, B.

terrestris. The raw data are given in Table 3.1.

Source* tf df

A. HIBERNATION SURVIVAL

Maternal family line (n = 367)!

Maternal family line f

Crossing treatment

Maternal family line X crossing treatment

Julian date of mating |

Paternal family line (n = 162)$

Paternal family line t

Julian date of mating t

B. COLONY FOUNDATION (n = 167)+

Maternal family line t

Crossing treatment f

Maternal family line X crossing treatment

* Variables in logistic regression. Variables included in the final logistic regression model

based on backward (stepwise) method are denoted with t-

$ n is number of queens tested

95.651 17 < 0.0001

0.757 1 0.384

13.801 17 0.681

11.529 1 0.0007

29.549 16 0.030

16.293 1 0.0001

47.978 11 < 0.0001

4.236 1 0.040

12.254 11 0.345

In the inbred treatment the maternal family line is identical with the paternal family

line. Thus, when testing for paternal family line effects, we analyzed queens from the

outbred treatment only to avoid confounding effects of the maternal family line. The effect

of the paternal family line, i.e., the mother colony of the queen's mate, and the date of

mating on hibernation survival of bumble bee queens followed the same pattern as the

effect of the maternal family lines (Fig. 3.2 B, Table 3.2 A). The odds of a queen surviving

hibernation decreased by 7 % every day (odds ratio eB = 0.929). Significances remained

54

after step-up sequential Bonferroni correction. However, the effect of paternal family line

on hibernation success was not significant when classifying the queens that died within one

day after the end of hibernation as having hibernated successfully (P = 0.240).

Colony foundation success

Overall colony foundation success was 41.1 % (33.5 % if queens that died within one

day after the end of hibernation were classified as having hibernated successfully). Again,

the percentage of queens successfully founding a colony varied from 0 to 100 % among

maternal family lines (Table 3.1). Hence, maternal family line significantly predicted a

queen's ability to raise offspring (Table 3.2 B). In contrast to hibernation survival, colony

foundation success was significantly reduced by crossing treatment but not by a late mating

date (Tables 3.1, 3.2 B). Inbreeding depression S in maternal family lines ranged from -1 to

1 (Table 3.4), reflecting the large among-family line variation.

Colony life-history traits

For the three colony performance measures, colony size, caloric investment into

sexuals, and the Shaw-Mohler index, the MANOVA revealed a uniform pattern: maternal

family lines varied strongly in their response to inbreeding, indicated by the highly

significant interaction term between maternal family line and crossing treatment (Table

3.3). The crossing treatment per se did not have a consistent effect. Neither did the maternal

family line when analyzing the three colony measures individually (univariate P-values;

Table 3.3). However, when considering the colony measures in combination, the effect of

the maternal family line was close to significance at the 5 %-level.

55

Table 3.3. MANOVA for three colony performance measures (colony size, caloric

investment into sexuals, and the Shaw-Mohler index) in the bumble bee, B. terrestris. n =

58 colonies tested. The raw data are given in Table A3.1.

Source of variation* Wilks A df F P Univariate F

effect, (£)

error Colony Caloric Shaw-

size investment Mohler

sexuals index

Maternal family line 0.532 18, 1.658 0.057 1.086 1.593 0.861

119.3 (0.386) (0.172) (0.531)

Crossing treatmentf 0.524 3,4 1.212 0.413 0.021 0.522 1.050

(0.889) (0.497) (0.345)

Maternal family line 0.390 18, 2.616 0.001 4.143 4.180 3.914

X crossing treatment 119.3 (0.002) (0.002) (0.003)

* Maternal family line is considered a random effect, crossing treatment a fixed effect.

Single colonies are nested within the interaction between maternal family line and

crossing treatment.

t Maternal family line X crossing treatment is error term

Colony size (number of workers).—Average colony size was 72.8 ± 9.5 workers

(mean ± SE, range 0-322), 66.1 ± 14.4 workers in inbred colonies and 78.3 ± 23.1 workers

in outbred colonies. While in some maternal family lines the inbred colonies grew larger

than the outbred colonies (e.g., family lines #7 and #19; Fig. 3.3, Table A3.2), it was the

opposite in other maternal family lines (e.g., #2 and #17; Fig. 3.3, Table A3.2). Inbreeding

depression ô varied from -0.86 to 0.85 among maternal family lines (Table 3.4). Of the

inbred colonies, 51.5 % (17 colonies) did not produce diploid males, 45.5 % (15) produced

diploid males, and 3 % (1) could not be assigned because they produced too few offspring.

This ratio is in line with expectations from complementary sex determination at one locus

56

(Cook & Crozier 1995). Colonies that did not produce diploid males (inbred: 95.3 ± 22.1

workers; outbred: 78.3 ± 23.1) had roughly double the number of workers compared to

colonies that produced diploid males (42.4 ± 12.9). The colonies that produced diploid

males therefore did not seem to compensate for the loss of worker force due to diploid male

production. When we accounted for the potential colony size of the diploid male colonies

by doubling the number of workers they produced, results of the MANOVA (Table 3.3)

were qualitatively similar with one exception: the maternal family line effect was highly

significant in the multivariate test (Wilks A = 0.450, Faill93 = 2.164, P = 0.007) while still

remaining insignificant in the univariate tests.

Colony size

250-i

SS 200-j

r

of

workerso

o

1

1

1 I T 0 inbred

outbred

a 50-

10J

1

I::::

:::: t i:

-T T

T ::: t

n

2 4 7 11 15 17 19

Maternal family line

Fig. 3.3. Average number of workers in bumble bee colonies from different maternal

family lines. Within each family line, the colony founding queens were mated to their

brother (inbred) or to an unrelated male (outbred). The exact data are given in Table A3.2.

57

Caloric investment into sexuals.—The experimental colonies produced 8.2 ± 2.4

young queens (mean ± SE, range 0-94) and 184.5 ± 34.6 haploid males (range 0-590)

(Table A3.1). The two colonies with the largest output of sexuals (316 queens and 634

haploid males, respectively) did not belong to a maternal family line producing enough

colonies to be included in the final analyses. Inbred colonies invested on average 528.70 ±

111.42 kJ into their sexuals, outbred colonies 453.85 ± 140.72 kJ (range for both treatments

combined 0-1463 kJ; Fig. 3.4, Table A3.2). Inbreeding depression ôin maternal family

lines ranged from -0.94 to 0.90 (Table 3.4).

1/5

x

in

a

t/3

>

o• 1—(

JO

U

Caloric investment into sexuals

1500-,

1000-

500-

TT

T

II

S inbred

outbred

2 4 7 11 15 17 19

Maternal family line

Fig. 3.4. Mean caloric investment into sexuals in bumble bee colonies from different

maternal family lines with crossing treatments as in Fig. 3.3. For details on calculation of

the caloric investment, see text. The exact data are given in Table A3.2.

58

Shaw-Mohler index.—The Shaw-Mohler fitness index can differ from a numeric

measure such as the caloric investment into sexuals (e.g., family lines #11 and #17, Figs.

3.4, 3.5, Table A3.2; magnitude of inbreeding depression 8, Table 3.4). However, the main

overall pattern for this fitness measure is the same as for the preceding variables (Fig. 3.5).

The inbred colonies had a mean fitness of 0.044 ± 0.014 (SE), the outbred colonies one of

0.031 ± 0.013 (Table A3.2). Inbreeding depression S ranged from -0.97 to 0.89 (Table 3.4).

Shaw-Mohler fitness index

0.21^-v,

wGO

X0.15-

<D

T3G

S-H

JO0.1- I

:•

S I;

o •'.

s0.05-

:

l T T£

t : ': t : —

c5 T^"1 ;' T ;

^=3 jITT" :| :| T

GO

0- F jjsa. l_i n

E3 inbred

outbred

2 4 7 11 15 17 19

Maternal family line

Fig. 3.5. Mean colony fitness (Shaw-Mohler index) in bumble bees from different maternal

family lines with crossing treatments as in Fig. 3.3. For details on calculation of the Shaw-

Mohler index, see text. The exact data are given in Table A3.2.

59

Inbreeding depression

For all variables measured, among-family variation of inbreeding depression Ô was

large and ranged from strong inbreeding depression (relative fitness of the outbred

treatment higher than the fitness of the inbred treatment, Ô > 0) to strong outbreeding

depression (relative fitness of the inbred treatment higher than the fitness of the outbred

treatment, ô < 0). Within maternal family line, the values of inbreeding depression ô for the

three colony measures, colony size, caloric investment into sexuals, and the Shaw-Mohler

index, were intercorrelated (range of Pearson correlation coefficients r 0.897-0.962, all P-

values < 0.001 and significant after step-up sequential Bonferroni correction). Averaged

over all the family lines, inbreeding had no significant effect on any of the traits measured

as the mean inbreeding depression values did not differ from zero (Table 3.4).

Size of sexuals

To estimate the quality of the sexuals produced, we measured the wing size of the

young queens and haploid males. Within a colony, the number of daughter queens

correlated negatively with the mean size of the daughter queens (Pearson's r = -0.518, n =

16, P = 0.040). At the same time, the worker number was positively correlated with the

queen number (r = 0.572, n = 48, P < 0.001) but negatively correlated with queen size (r =

-0.622, n = 16, P = 0.010). In males, there was also a trade-off between number and size (r

= -0.523, n = 28, P = 0.004). However, neither male number nor male size correlated with

worker number (r = 0.354, n = 28, P = 0.064 and r = -0.100, n = 28, P = 0.612,

respectively). To avoid diploid males in the sample, inbred colonies were excluded from

the analysis of the males. Significance levels reported remained after step-up sequential

Bonferroni correction. The raw data are given in Table A3.1.

60

Table 3.4. Estimated inbreeding depression, S, for five fitness measures in maternal familylines of the bumble bee, B. terrestris. Mean inbreeding depression was calculated from

values at the level of the family line. The raw data for hibernation survival and colonyfoundation success are given in Table 3.1, for the other traits in Table A3.2.

Maternal Hibernation Colony Colony Caloric Shaw-

family line survival foundation size* investment Mohler

sexuals* index*

1 0.39 -0.45

2 -0.01 0.40 0.75 0.61 0.89

3 0.00 —

4 -0.36 -0.41 0.34 0.005 0.27

5 0.22 -1.00 —

6 0.00 0.00 —

7 0.25 0.20 -0.83 -0.94 -0.97

8 -0.29 0.84

9 0.20 1.00 —

10 -1.00 —

11 -0.13 0.13 0.09 -0.20 -0.32

12 0.75 1.00 —

13 1.00 —

14 0.08 0.00

15 0.17 -0.17 -0.08 -0.86 -0.96

16 0.14 0.50

17 0.13 0.43 0.85 0.90 0.60

18 -0.17 1.00 —

19 -0.17 0.38 -0.86 -0.82 -0.81

mean 5 0.06 0.24 0.04 -0.19 -0.19

(SE) (0.10) (0.14) (0.26) (0.28) (0.29)

t 0.66 1.67 0.14 -0.67 -0.63

P 0.519 0.115 0.891 0.529 0.550

Only the values for the maternal family lines included in the statistical analyses are

given.

61

Discussion

Large among-family variation appears to be a key feature that characterizes all main

steps in the life cycle of the bumble bee, Bombus terrestris, in particular, hibernation

survival, colony foundation success, colony size, and reproductive output. With respect to

the effects of inbreeding, the among-family variation is even more pronounced.

We found that the effect of inbreeding on colony performance varied substantially

among maternal family lines, both with respect to the direction of the effect (inbreeding or

outbreeding depression) and its magnitude. While this study might overestimate this

among-family variation in bumble bees to some extent because of small sample sizes in

some family lines, similar patterns have recently been found in a variety of plant and

animal systems, including haplodiploid ones. Typically, some families show less

inbreeding depression and thus are better able to tolerate inbreeding than others (e.g.,

Stevens et al. 1997; Koelewijn 1998; Mutikainen & Delph 1998; Rodrigânez et al. 1998;

Byers & Waller 1999; Jarne et al. 2000; Saito et al. 2000). This could be due to differences

in prior inbreeding history (Lande & Schemske 1985; Charlesworth & Charlesworth 1987),

random variation in mutational load due to deleterious alleles (Schultz & Willis 1995), or

maternal effects (del Castillo 1998). In some families in this study, inbred colonies

performed better than outbred ones. Therefore, averaged over the entire experimental

population, the inbreeding and outbreeding depressions cancelled each other out. Thus, our

findings are consistent with Duchateau et al. (1994) who did not find inbreeding depression

for colony growth rate at the level of the experimental population in bumble bees.

However, due to the large among-family variation in inbreeding depression, estimates

based on a small number of family lines are associated with large standard errors. These

can conceal inbreeding depression found with more powerful analyses as was the case for

62

colony foundation success. Variation among bumble bee families has also been found after

four generations of inbreeding: only in one out of four isofemale lines was the percentage

of queens laying eggs significantly reduced after repeated brother-sister matings (Beekman

et al. 1999).

While hibernation survival was not reduced after brother-sister mating colony

foundation success was. The latter result is in constrast to the one found by Duchateau et al.

(1994) which was based on a threefold smaller sample size. An idiosyncratic consequence

of inbreeding in organisms with complementary sex determination system is the production

of diploid males. Previous studies on single-generation inbreeding in bumble bees have

found that colonies producing diploid males grow at a slower or similar rate to colonies not

producing diploid males (Plowright & Pallett 1979 and Duchateau et al. 1994,

respectively). To our surprise, we found that inbred colonies, despite the production of

diploid males, sometimes performed better than outbred colonies. As for mating itself, we

did not observe any obvious inbreeding avoidance behavior: queens mated as readily with

their brother as with an unrelated male (Duvoisin 1998; C. Gerloff, pers. obs.). Of course,

our mating conditions might have been too artificial for natural mating behavior (e.g.,

Svensson 1979; Hovorka et al. 1998) and inbreeding avoidance mechanisms (Motro 1991;

Foster 1992) either to be observed or to be effective. However, as this study shows, the

costs of episodic inbreeding may not be high and consistent enough to facilitate the

evolution of inbreeding avoidance, since inbreeding advantage is as likely to occur as

inbreeding depression.

All findings on inbreeding in bumble bees stem from laboratory studies. The effects

of inbreeding under natural conditions can only be extrapolated, as no field studies exist.

However, inbreeding depression is generally expected to be stronger in the field as

compared to the laboratory (for evidence, see e.g., Dudash 1990; del Castillo 1998;

63

Koelewijn 1998). Hence, Plowright & Pallett (1979) and Duchateau et al. (1994) argued

that their laboratory breeding conditions were too benign for inbreeding depression to be

expressed fully, and that some (diploid male) colonies probably would not have survived in

nature. This might also apply to our study. However, benign laboratory conditions cannot

explain the outbreeding depression we observed in certain family lines. Thus, some family

lines might also exhibit outbreeding depression in the field. This could help to explain why

bumble bees have been successful at colonizing new areas even from extremely small

founder populations: the introduction and establishment of bumble bees in New Zealand in

1885 (Fair 1889, cited in Macfarlane & Gurr 1995) and in Tasmania in 1992 (Semmens et

al. 1993) are well documented examples where only a few queens invaded and founded a

population. Although highly inbred - half the colonies produce diploid males (Buttermore

et al. 1998) - B. terrestris is spreading at a rate of 12.5 km/year across Tasmania

(Buttermore 1997). The dispersal rate of bumble bees in New Zealand was even faster,

about 90 km/year (Buttermore 1997). On the other hand, the conditions in the new habitats

might have been favorable with little interspecific competition, few parasites, and a benign

climate that allows for, at least partially, bivoltine life cycles (Donovan & Weir 1978;

Buttermore 1997). Nevertheless, our results shed some light on this issue - if queens from a

"good" family happen to invade, then the invasion can be successful, for example, because

of high hibernation or colony foundation success and little inbreeding depression.

Due to insufficient sample size, the effect of the paternal family line could be

analyzed only for hibernation success. To our surprise, hibernation success of queens was

significantly affected by the family line of their mates and the mating date (discussed

below). In bumble bees, the sexuals interact only during mating. The queen stores the

sperm during hibernation for subsequent colony foundation. Thus, the male can potentially

affect a queen's hibernation survival by whatever the male transfers during mating. The

64

male bumble bee transfers sperm as well as a mating plug during copulation (Duvoisin et

al. 1999). The mating plug reduces queen (re-)mating behavior but seems to have few other

effects (Sauter et al. 2001). In other insects, for example, in Drosophila melanogaster,

toxins in male seminal fluid secretions reduce a female's life span (reviewed in Wolfner

1997). The other potential influence of the male on the queen's hibernation survival might

be behaviorally, e.g., harassment prior or during mating. In bumble bees, we currently do

not know the mechanism behind how the male affects the hibernation success of its mate.

As in all studies that investigate the influence of family lineage, genetic and

environmental effects need to be separated. In social Hymenoptera, this is straightforward

for the effects of paternal lineage as males only donate sperm but have died long before the

queen starts her colony. However, how widespread the influence of the mate's family

lineage on the female's postmating performance is (e.g., on hibernation survival or life

span, see above) and what basis it has, remains open. The effect of the maternal family line,

in contrast, is typically confounded, as the environmental effects include the environment

outside and inside a colony, as well as maternal effects (e.g., Schaal 1984). In this study,

hibernation and breeding conditions were kept constant and therefore environmental effects

are randomized over maternal family lines. To reduce possible maternal effects the "mother

colonies" (Fig. 3.1), from which the experimental queens originated, were reared under

standardized breeding conditions. Thus, the strong variation among family lines is likely to

reflect real differences either in genetic background, in carried-over maternal effects, or an

interaction between progeny genotype and maternal effect (Roach & Wulff 1987).

Hibernation success is a crucial step in the life cycle of diapausing species, such as B.

terrestris. Hibernation was the only step influenced by the Julian date of mating, that is, the

date when hibernation started (see Materials and Methods): the first queens to mate were

much more likely to survive hibernation than the queens that mated later. As the slope of

65

hibernation survival probability is roughly the same when analyzing different subsets of

hibernated queens (maternal or paternal family lines), there seems to be a strong and

generally negative temporal effect. This effect cannot be explained by the grouped storage

of the hibernating queens because the groups contained queens that mated on the same day

but that usually belonged to different family lines. As a result, early and late queens - with

respect to the colony cycle - hibernated in the same group. Higher survival rates and fitness

of individuals that are born and reproduce early in a season as compared to individuals that

are born and reproduce later have been reported for a variety of organisms, such as plants

(Roach & Wulff 1987; Uyenoyama et al. 1993), birds (e.g., Martin 1987; Korpimäki &

Wiehn 1998), fish (Schultz 1993), and insects (Ohgushi 1991; Cushman et al. 1994). In

bumble bees, early colonies grow larger and produce more sexuals than late-starting

colonies (Müller & Schmid-Hempel 1992b). The temporal decrease of hibernation success

therefore reflects a dual advantage of early timing: early colonies produce more queens

which in turn hibernate more successfully. Remarkably, however, and in contrast to

previous studies (Horber 1961; Holm 1972; Owen 1988; Beekman et al. 1998), hibernation

success in this study was not affected by queen size. Rather, some family lines simply were

better at hibernating than others. Admittedly, the hibernation duration applied in this study

might have been too short (3.5 as compared to 8-9 months in the field; Holm 1972) for size

effects to show.

Not surprisingly, our study also showed patterns reported previously. For example,

the number of young queens produced is correlated with worker number of the colony (e.g.,

Owen et al. 1980; Owen & Plowright 1982; Müller & Schmid-Hempel 1992b). In both

sexes, number and size were negatively correlated within colony. In queens, this trade-off

might have been caused partly by an experimental artifact: large colonies might have been

space limited in the standard sized hives or food limited inspite of daily feeding and

66

therefore might have produced smaller queens (Sutcliffe & Plowright 1988). While size

(mass) can be important for daughter queens (for example, heavier queens have a higher

hibernation success (Horber 1961; Holm 1972; Owen 1988; Beekman et al. 1998), are more

successful in usurpation, and more efficient in foraging (discussed in Owen 1988), small

size seems not to decrease male fitness, neither by reducing mating opportunities (Foster

1992; C. Gerloff, pers. obs.) nor hibernation success of the queen with which it mated (this

study). We conclude that large among-family variation in important life-history traits and in

the effects of episodic inbreeding are important for the evolutionary ecology of a social

insect like B. terrestris. Episodic inbreeding, as shown by the results of our experiment, is

likely to intensify the differential representation of families in the population.

Acknowledgements

We thank N. Duvoisin, H. Magro, B. Ottmer, R. Schmid-Hempel, and K. Suter for

help with the experiments, J. Jokela, P. Mutikainen, and J. Wiehn for help with the

statistics, and M. Brown, P. Mutikainen, and J. Wiehn for comments on the manuscript.

The study was financially supported by grants from the Swiss National Science Foundation

(no. 3100-049040.95 to PSH) and the Swiss Federal Institute of Technology (ETH) Zürich

(no. 0-20-010-95 to PSH).

67

References

Brian, A.D. 1952. Division of labor and foraging in Bombus agrorum Fabricius. J. Anim.

Ecol. 2:223-240.

Adamson, M.L. 1989. Evolutionary biology of the Oxyurida (Nematoda): Biofacies of a

haplodiploid taxon. Adv. Parasitai. 28:175-228.

Àgren, J. & Schemske, D.W. 1993. Outcrossing rate and inbreeding depression in two

annual monoecious herbs, Begonia hirsuta and B. semiovata. Evolution 47:125-135.

Alford, D.V. 1975. Bumblebees. Davis-Poynter, London, UK.

Beekman, M. & van Stratum, P. 1998. Bumblebee sex ratios: why do bumblebees produce

so many males? Proc. R. Soc. Lond. B 265:1535-1543.

Beekman, M., van Stratum, P. & Lingeman, R. 1998. Diapause survival and post-diapause

performance in bumblebee queens (Bombus terrestris). Entomol. Exp. Appl. 89:207-

214.

Beekman, M., van Stratum, P. & Veerman, A. 1999. Selection for non-diapause in the

bumblebee Bombus terrestris, with notes on the effect of inbreeding. Entomol. Exp.

Appl. 93:69-75.

Beekman, M. & van Stratum, P. 2000. Does the diapause experience of bumblebee queens

Bombus terrestris affect colony characteristics? Ecol. Entomol. 25:1-6.

Berkelhamer, R.C. 1983. Intraspecific genetic variation and haplodiploidy, eusociality, and

polygyny in the Hymenoptera. Evolution 37:540-545.

Bertsch, A. 1984. Foraging in male bumblebees (Bombus lucorum L.): maximizing energy

or minimizing water load? Oecologia 62:325-336.

68

Breed, M.D., Page, R.E., Jr., Hibbard, B.E. & Bjostad, L.B. 1995. Interfamily variation in

comb wax hydrocarbons produced by honey bees. J. Chem. Ecol. 21:1329-1338.

Buttermore, R.E. 1997. Observations of successful Bombus terrestris (L.) (Hymenoptera:

Apidae) colonies in Southern Tasmania. Aust. J. Entomol. 36:251-254.

Buttermore, R.E., Pomeroy, N., Hobson, W., Semmens, T. & Hart, R. 1998. Assessment of

the genetic base of Tasmanian bumble bees {Bombus terrestris) for development as

pollination agents. J. Apic. Res. 37:23-25.

Byers, D.L. & Waller, D.M. 1999. Do plant populations purge their genetic load? Effects of

population size and mating history on inbreeding depression. Annu. Rev. Ecol. Syst.

30:479-513.

Charlesworth, D. & Charlesworth, B. 1987. Inbreeding depression and ist evolutionary

consequences. Annu. Rev. Ecol. Syst. 18:237-268.

Charnov, E.L. 1982. The theory of sex allocation. Princeton University Press, Princeton,

NJ, USA.

Cook, J.M. 1993. Sex determination in the Hymenoptera: a review of models and evidence.

Heredity 71:421-435.

Cook, J.M. & Crozier, R.H. 1995. Sex determination and population biology in the

Hymenoptera. Trends Ecol. Evol. 10:281-286.

Crozier, R.H. 1975. Insecta 7. Hymenoptera. Gebrüder Borntraeger, Berlin, Germany.

—. 1985. Adaptive consequences of male-haploidy. pp. 201-222 in W. Helle & M.W.

Sabelis, eds. Spider mites. Their biology, natural enemies and control. Elsevier

Science Publishers B. V., Amsterdam, The Netherlands.

Crozier, R.H. & Pamilo, P. 1996. Evolution of social insect colonies: sex allocation and kin

selection. Oxford University Press, Oxford, UK.

69

Cushman, J.H., Boggs, CL., Weiss, S.B., Murphy, D.D., Harvey, A.W. & R., E.P. 1994.

Estimating female reproductive success of a threatened butterfly: influence of

emergence time and hostplant phenology. Oecologia 99:194-200.

del Castillo, R.F. 1998. Fitness consequences of maternal and nonmatemal components of

inbreeding in the gynodioecious Phacelia dubia. Evolution 52:44-60.

Donovan, BJ. & Weir, S.S. 1978. Development of hives for field population increase, and

studies on the life cycle of four species of introduced bumble bees in New Zealand.

N. Z. J. Agric. Res. 21:733-756.

Duchateau, M.J. & Velthuis, H.H.W. 1988. Development and reproductive strategies in

Bombus terrestris colonies. Behaviour 107:186-207.

Duchateau, M.J., Hoshiba, H. & Velthuis, H.H.W. 1994. Diploid males in the bumble bee

Bombus terrestris: Sex determination, sex alleles and viability. Entomol. Exp. Appl.

71:263-269.

Duchateau, M.J. & Marien, J. 1995. Sexual biology of haploid and diploid males in the

bumble bee Bombus terrestris. Insect. Soc. 42:255-266.

Duvoisin, N. 1998. Mating behaviour in the bumblebee Bombus terrestris L.

(Hymenoptera: Apidae): Mate choice, sperm transfer and plugging. Diploma thesis.

Swiss Federal Institute of Technology (ETH) Zürich, Switzerland.

Duvoisin, N., Baer, B. & Schmid-Hempel, P. 1999. Sperm transfer and male competition in

a bumblebee. Anim. Behav. 58:743-749.

Estoup, A., Solignac, M., Cornuet, J.M., Goudet, J. & Scholl, A. 1996. Genetic

differentiation of continental and island populations of Bombus terrestris

(Hymenoptera: Apidae) in Europe. Mol. Ecol. 5:19-31.

Farr, S.C. 1889. Introduction of the humble bee into New Zealand. N. Z. Country J. 13:284-

287.

70

Foster, R.L. 1992. Nestmate recognition as an inbreeding avoidance mechanism in bumble

bees (Hymenoptera: Apidae). J. Kansas Entomol. Soc. 65:238-243.

Free, J.B. & Butler, CG. 1959. Bumblebees. Collins, London, UK.

Gordon, D.M. 1991. Behavioral flexibility and the foraging ecology of seed-eating ants.

Am. Nat. 138:379-411.

Graur, D. 1985. Gene diversity in Hymenoptera. Evolution 39:190-199.

Harder, L.D. 1986. Influences on the density and dispersion of bumble bee nests

(Hymenoptera: Apidae). Holarctic Ecol. 9:99-103.

Hochberg, Y. 1988. A sharper Bonferroni procedure for multiple tests of significance.

Biometrika 75:800-802.

Holm, S.N. 1972. Weight and life length of hibernating bumble bee queens (Hymenoptera:

Bombidae) under controlled conditions. Entomol. Scand. 3:313-320.

Holsinger, K.E. 1988. Inbreeding depression doesn't matter: the genetic basis of mating-

system evolution. Evolution 42:1235-1244.

—. 1991. Inbreeding depression and the evolution of plant mating systems. Trends Ecol.

Evol. 6:307-308.

Horber, E. 1961. Beitrag zur Domestikation der Hummeln. Vierteljahresschr. Naturf. Ges.

Zürich 106:424-447.

Hovorka, O., Urbanovâ, K. & Valterovâ, I. 1998. Premating behavior of Bombus confusus

males and analysis of their labial gland secretion. J. Chem. Ecol. 24:183-193.

Imhoof, B. & Schmid-Hempel, P. 1999. Colony success of the bumble bee, Bombus

terrestris, in relation to infections by two protozoan parasites, Crithidia bombi and

Nosema bombi. Insect. Soc. 46:233-238.

71

Jarne, P., Perdieu, M.A., Pernot, A.F., Delay, B. & David, P. 2000. The influence of self-

fertilization and grouping on fitness attributes in the freshwater snail Physa acuta:

Population and individual inbreeding depression. J. Evol. Biol. 13:645-655.

Johnston, A.B. & Wilson, E.O. 1985. Correlates of variation in the major/minor ratio of the

ant Pheidole dentata (Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 78:8-11.

Johnston, M.O. & Schoen, D.J. 1994. On the measurement of inbreeding depression.

Evolution 48:1735-1741.

Koelewijn, H.P. 1998. Effects of different levels of inbreeding on progeny fitness in

Plantago coronopus. Evolution 52:692-702.

Korpimäki, E. & Wiehn, J. 1998. Clutch size of kestrels: Seasonal decline and experimental

evidence for food limitation under fluctuating food conditions. Oikos 83:259-272.

Kulincevic, J.M. 1986. Breeding accomplishment with honey bees. pp. 391-414 in T.E.

Rinderer, ed. Bee genetics and breeding. Academic Press, Orlando, FL, USA.

Lande, R. & Schemske, D.W. 1985. The evolution of self-fertilization and inbreeding

depression in plants: I. Genetic models. Evolution 39:24-40.

Macfarlane, R.P. & Gurr, L. 1995. Distribution of bumble bees in New Zealand. N. Z.

Entomol. 18:29-36.

Martin, T.E. 1987. Food as a limit on breeding birds: a life-history perspective. Annu. Rev.

Ecol. Syst. 18:453-487.

Michener, CD. 1964. Reproductive efficiency in relation to colony size in hymenopterous

societies. Insect. Soc. 11:317-342.

Morse, D.H. 1982. Behavior and ecology of bumble bees. pp. 245-322 in H.R. Hermann,

ed. Social Insects. Academic Press, New York, NY, USA.

Motro, U. 1991. Avoiding inbreeding and sibling competition: The evolution of sexual

dimorphism for dispersal. Am. Nat. 137:108-115.

72

Müller, C.B. & Schmid-Hempel, P. 1992a. Variation in life-history pattern in relation to

worker mortality in the bumble-bee, Bombus lucorum. Funct. Ecol. 6:48-56.

—. 1992b. Correlates of reproductive success among field colonies of Bombus lucorum: the

importance of growth and parasites. Ecol. Entomol. 17:343-353.

Mutikainen, P. & Delph, L.F. 1998. Inbreeding depression in gynodioecious Lobelia

siphilitica: among-family differences override between-morph differences. Evolution

52:1572-1582.

Ohgushi, T. 1991. Lifetime fitness and evolution of reproductive pattern in the herbivorous

lady beetle. Ecology 72:2110-2122.

Oldroyd, B., Rinderer, T. & Buco, S. 1991. Heritability of morphological characteristics

used to distinguish European and Africanized honeybees. Theor. Appl. Genet.

82:499-504.

Owen, R.E., Rodd, F.H. & Plowright, R.C. 1980. Sex ratios in bumble bee colonies:

complications due to orphaning? Behav. Ecol. Sociobiol. 7:287-291.

Owen, R.E. & Plowright, R.C. 1982. Worker-queen conflict and male parentage in bumble

bees. Behav. Ecol. Sociobiol. 11:91-99.

Owen, R.E. 1988. Body size variation and optimal body size of bumble bee queens

(Hymenoptera: Apidae). Can. Entomol. 120:19-28.

—. 1989. Differential size variation of male and female bumblebees. J. Hered. 80:39-43.

Owen, R.E. & Harder, L.D. 1995. Heritable allometric variation in bumble bees:

Opportunities for colony-level selection of foraging ability. J. Evol. Biol. 8:725-738.

Page, R.E., Jr., Robinson, G.E. & Fondrk, M.K. 1989. Genetic specialists, kin recognition

and nepotism in honey-bee colonies. Nature 338:576-579.

73

Page, R.E., Jr., Metcalf, R.A., Metcalf, R.L., Erickson, E.H., Jr. & Lampman, R.L. 1991.

Extractable hydrocarbons and kin recognition in honeybee (Apis mellifera L.). J.

Chem. Ecol. 17:745-756.

Pamilo, P. & Crozier, R.H. 1981. Genie variation in male haploids under deterministic

selection. Genetics 98:199-214.

Pamilo, P. & Rosengren, R. 1984. Evolution of nesting strategies of ants: Genetic evidence

from different population types of Formica ants. Biol. J. Linn. Soc. 21:331-348.

Pamilo, P. & Crozier, R.H. 1997. Population biology of social insect conservation. Mem.

Museum Victoria 56:411-419.

Plowright, R.C. & Pallett, M.J. 1979. Worker-male conflict and inbreeding in bumble bees

(Hymenoptera: Apidae). Can. Entomol. 111:289-294.

Pomeroy, N. & Plowright, R.C. 1980. Maintenance of bumble bee colonies in observation

hives (Hymenoptera: Apidae). Can. Entomol. 112:321-326.

Porter, S.D. & Tschinkel, W.R. 1985. Fire ant polymorphism: the ergonomics of brood

production. Behav. Ecol. Sociobiol. 16:323-336.

Queller, D.C., Negron Sotomayor, J.A., Strassmann, J.E. & Hughes, CR. 1993. Queen

number and genetic relatedness in a neotropical wasp, Polybia occidentalis. Behav.

Ecol. 4:7-13.

Roach, D.A. & Wulff, R.D. 1987. Maternal effects in plants. Annu. Rev. Ecol. Syst.

18:209-235.

Rodd, F.H., Plowright, R.C. & Owen, R.E. 1980. Mortality rates of adult bumble bee

workers (Hymenoptera, Apidae). Can. J. Zool. 58:1718-1721.

Rodrigânez, J., Toro, M.A., Rodriguez, M.C. & Silio, L. 1998. Effect of founder allele

survival and inbreeding depression on litter size in a closed line of Large White pigs.

Anim. Sei. 67:573-582.

74

Ross, K.G. & Fletcher, D.J.C. 1986. Diploid male production: A significant colony

mortality factor in the fire ant Solenopsis invicta (Hymenoptera: Formicidae). Behav.

Ecol. Sociobiol. 19:283-292.

Saito, Y., Sahara, K. & Mori, K. 2000. Inbreeding depression by recessive deleterious

genes affecting female fecundity of a haplo-diploid mite. J. Evol. Biol. 13:668-678.

Sauter, A., Brown, M.J.F., Baer, B. & Schmid-Hempel, P. 2001. Males may prevent the

evolution of super-maters in social insects. Proc. R. Soc. Lond. B, in press.

Schaal, B.A. 1984. Life-history variation, natural selection, and maternal effects in plant

populations, pp. 188-206 in R. Dirzo & J. Sarukhan, eds. Perspectives on plant

population ecology. Sinauer Associates, Sunderland, MA, USA.

Schmid-Hempel, P. & Heeb, D. 1991. Worker mortality and colony development in

bumblebees, Bombus lucorum (L.) (Hymenoptera, Apidae). Mitt. Schweiz. Entomol.

Ges. 64:93-108.

Schmid-Hempel, P. & Loosli, R. 1998. A contribution to the knowledge of Nosema

infections in bumble bees, Bombus spp. Apidologie 29:525-535.

Schmid-Hempel, P., Puhr, K., Krüger, N., Reber, C. & Schmid-Hempel, R. 1999. Dynamic

and genetic consequences of variation in horizontal transmission for a microparasitic

infection. Evolution 53:426-434.

Schultz, E.T. 1993. The effect of birth date on fitness of female dwarf perch, Micrometrus

minimus (Perciformes: Embiotocidae). Evolution 47:520-539.

Schultz, S.T. & Willis, J.H. 1995. Individual variation in inbreeding depression: the roles of

inbreeding history and mutation. Genetics 141:1209-1223.

Semmens, T.D., Turner, E. & Buttermore, R. 1993. Bombus terrestris (L.) (Hymenoptera:

Apidae) now established in Tasmania. J. Aust. Entomol. Soc. 32:346.

75

Shykoff, J.A. & Müller, C.B. 1995. Reproductive decisions in bumble-bee colonies: The

influence of worker mortality in Bombus terrestris (Hymenoptera, Apidae). Funct.

Ecol. 9:106-112.

Sladen, F.W.L. 1989. The humble-bee. Its life history and how to domesticate it. Logaston

Press, Little Logaston, UK.

Stevens, L., Yan, G. & Pray, L.A. 1997. Consequences of inbreeding on invertebrate host

susceptibility to parasitic infection. Evolution 51:2032-2039.

Stouthamer, R., Luck, R.F. & Werren, J.H. 1992. Genetics of sex determination and the

improvement of biological control using parasitoids. Environm. Entomol. 21:427-

435.

Sutcliffe, G.H. & Plowright, R.C. 1988. The effects of food supply on adult size in the

bumble bee Bombus terricola Kirby (Hymenoptera: Apidae). Can. Entomol.

120:1051-1058.

Svensson, B.G. 1979. Patrolling behaviour of bumble bee males (Hymenoptera, Apidae) in

a supalpine/alpine area, Swedish Lapland. Zoon 7:67-94.

Uyenoyama, M.K., Holsinger, K.E. & Waller, D.M. 1993. Ecological and genetic factors

directing the evolution of self-fertilization. Oxf. Surv. Evol. Biol. 9:327-381.

Webb, M.C. 1961. The biology of the bumblebee of a limited area in eastern Nebraska.

Ph.D. thesis. University of Nebraska, Lincoln, NE, USA.

Whiting, P.W. 1940. Multiple alleles in sex determination of Habrobracon. J. Morph.

66:323-355.

Widmer, A., Schmid-Hempel, P., Estoup, A. & Scholl, A. 1998. Population genetic

structure and colonization history of Bombus terrestris s.l. (Hymenoptera: Apidae)

from the Canary Islands and Madeira. Heredity 81:563-572.

76

Widmer, A. & Schmid-Hempel, P. 1999. The population genetic structure of a large

temperate pollinator species, Bombus pascuorum (Scopoli) (Hymenoptera: Apidae).

Mol. Ecol. 8:387-398.

Wolfner, M.F. 1997. Tokens of love: functions and regulation of Drosophila male

accessory gland products. Insect Biochem. 27:179-192.

Woyke, J. 1963. Drone larvae from fertilized eggs of the honeybee. J. Apic. Res. 2:19-24.

Zar, J.H. 1996. Biostatistical Analysis. 3rd edn. Prentice-Hall, Upper Saddle River, NJ,

USA.

77

4 Effects of Inbreeding, Sex, and Ploidy on Offspring Quality

in a Social Insect, Bombus terrestris (L.) (Hymenoptera:

Apidae)

(Gerloff C.U., Ottmer B.K. & Schmid-Hempel P., in revision for Journal of Evolutionary

Biology)

Abstract

Inbreeding can negatively affect offspring quality. Here, we studied how the immune

response of a social insect, measured as the degree of encapsulation of a novel antigen, and

body size are affected by inbreeding, sex, and ploidy. In the bumble bee, Bombus terrestris

(L.), we compared offspring of colonies resulting from brother-sister matings to that of

outbred colonies. Inbreeding affected neither immune response nor body size in either

workers or haploid males under laboratory conditions. However, offspring quality varied

significantly among families and colonies. Diploid males had a significantly lower immune

response than haploid males, who in turn had a significantly lower immune response than

workers of the same colony. The body size of diploid males was intermediate between the

body size of workers and haploid males. Our results suggest a strong genetic influence on

immunity while they corroborate a strong environmental effect on body size.

78

Keywords

among-family variation; body size; Bombus terrestris; bumble bee; diploid males;

encapsulation; Hymenoptera; inbreeding; sex differences; social insects.

Introduction

It has long been recognised that inbreeding can negatively affect a number of fitness

components in a variety of organisms (e.g., Thornhill 1993; Frankham 1995; Lacy 1997; Roff

1997), although the actual mechanisms behind inbreeding depression remain controversial

(Thornhill 1993; Roff 1997). However, in some cases, inbred offspring can have higher

fitness than outbred offspring (Thornhill 1993; Roff 1997). This is usually true only for some

lineages within a population (e.g., Mutikainen & Delph 1998; Rodrigânez et al. 1998; Byers

& Waller 1999; Jarne et al. 2000).

In a world where parasites abound, immune defence is arguably one of the most

important fitness components for any organism, as parasites can have a large impact on host

fitness (e.g., Chapman & George 1991; Lehmann 1993; Brown et al. 2000). Among the many

factors that influence an organism's immune function genotype plays an important role (e.g.,

Thompson & Burdon 1992; Schmid-Hempel & Koella 1995; Wakelin & Apanius 1997).

Inbreeding affects an individual's genotype and, therefore, inbreeding can be expected to

influence immune function. To date, the effect of inbreeding on the immune system's capacity

to respond has been studied almost exclusively in vertebrates. In most cases, an increased

degree of homozygosity - a consequence of inbreeding (Falconer 1989) - was found to be

associated with increased susceptibility to parasitism (e.g., Allendorf & Leary 1986; O'Brien

&Evermann 1988; Lively et al. 1990; Mitton 1995; M0ller 1996; Coltman et al. 1999; Ochoa

& Jaffe 1999; Penn & Potts 1999, but see Burt & Bell 1992; Strauss & Karban 1994;

79

Ouborg et al. 2000). In contrast, the few studies using invertebrate systems have yielded

ambiguous results for this relationship (Plass 1953 cited in Cornuet 1986; Stevens et al.

1997). Although the invertebrate immune system shows functional analogies to the vertebrate

immune system (Gupta 1986a; Karp 1990; Hoffmann et al. 1999) there is a major difference

between the two types of immune system: invertebrates do not possess an adaptive immune

response based on the proliferation of antibodies but rely on constitutive defences. These

consist, on the one hand, of the inducible release of anti-microbial peptides (humoral reaction)

and, on the other hand, of haemocytes aggregating and encapsulating large foreign objects

(cellular reaction) (Brehélin 1986; Gupta 1986b; Lackie 1988b; Pathak 1993; Hoffmann &

Reichhart 1997). Hence, the findings in vertebrates might not always be transferable to

invertebrates (Stevens et al. 1997).

Social Hymenoptera, i.e., ants, bees, and wasps, are a particularly interesting group of

insects with respect to the effects of inbreeding. Firstly, within colonies, castes differ in the

level of inbreeding. Workers are female and arise from fertilised, diploid eggs and hence are

inbred if they are the offspring of two related parents. In contrast, males arise from

unfertilised, haploid eggs (Crozier 1975). By definition, they cannot carry a pair of alleles

identical by descent (Falconer 1989) and therefore cannot be inbred in the traditional sense.

Nevertheless, haploid males can be affected indirectly by inbreeding, for example, by maternal

effects (Moritz 1985) or by being reared by inbred workers. Secondly, due to the mode of sex

determination in social Hymenoptera ("complementary sex determination", CSD, where sex

is determined by a single, or perhaps several, multi-allelic locus, the "sex locus", Whiting

1940) inbreeding can result in the production of a new caste, the diploid males (e.g., Cook

1993; Cook & Crozier 1995). Haploid individuals are hemizygous at the sex locus and

develop into males. Diploid individuals that are heterozygous at the sex locus develop into

females (queens and workers). In contrast, diploid individuals that are homozygous at the sex

locus develop into males. By increasing homozygosity, inbreeding increases the incidence of

diploid males in social Hymenoptera.

80

The existence of diploid males offers an elegant way to investigate the phenomenon of

sex differences in immune response and susceptibility to parasites. Diploid males combine

male sex and the female characteristics of diploidy and early development during the colony

cycle (as in workers). This allows the partitioning of the influences of ploidy and sex on

immune response on the one hand, and genetic and environmental effects on the other hand.

In vertebrates, many cases of sex differences in immune response and prevalence of infections

are documented (e.g., Roberts et al. 1995; Sankilampi et al. 1997; McCurdy et al. 1998; Saino

et al. 2000). So are many cases where no differences between females and males were found

(e.g., Goh et al. 1990; Theis & Schwab 1992; Kukovetz et al. 1997; Johnsen et al. 2000).

Among the suggested reasons for these sex differences are differential exposure to parasites

and sex hormone levels (McCurdy et al. 1998), possibly mediated by sexual selection (M0ller

et al. 1998), and social environment (Saino et al. 2000). However, most studies focussed on

the inducible immune reaction and might therefore not adequately describe the situation in

invertebrates. Studies about sex differences in immune responses in invertebrates are scarce

and results obtained are ambiguous. While Carton and co-workers (1992) did not find a sex-

related difference in encapsulation response in Drosophila melanogaster, parasite infections

differed in prevalence, intensity, and the fitness changes they induced in their female and male

flour beetle hosts, Tribolium spp. (Stevens et al. 1997; Yan 1997). No reasons for these sex-

related differences observed were suggested.

Here, we use the bumble bee, Bombus terrestris (L.), as a system to study the effects of

inbreeding, sex, and ploidy on offspring quality in insects. Bumble bees are annual eusocial

Hymenoptera where only the fertilised queens overwinter to found a colony the next spring.

The colony then grows in worker numbers and finally reproduces by raising sexuals, the

males and daughter queens, that mate in late summer and disperse. Their biology indicates the

potential occurrence of inbreeding in bumble bees. Each year, only a few, successful colonies

reproduce, leading to a heavy reproduction bias within a population (Donovan & Weir 1978;

Müller & Schmid-Hempel 1992a; Imhoof & Schmid-Hempel 1999). Furthermore, population

81

sizes fluctuate strongly due to environmental conditions (Harder 1986; Sladen 1989; Pamilo

& Crozier 1997). Indeed, in island populations of the study species, B. terrestris, molecular

analyses and high frequencies of diploid males revealed significant levels of inbreeding

(Widmer et al. 1998 and Buttermore et al. 1998, respectively). In large, connected populations

inbreeding can be expected to be episodic, i.e., to last only for one or a few generations,

because migration dilutes the local population structure (Estoup et al. 1996). Inbreeding

depression in laboratory-reared bumble bees has been demonstrated for brood viability,

colony size, and egg laying in queens (Plowright & Pallett 1979; Beekman et al. 1999).

Gerloff & Schmid-Hempel have found large among-family variation for the effect of

inbreeding on colony fitness (Chapter 3, but see Duchateau et al. 1994).

We measured offspring quality in terms of non-specific immune response (Frank

2000) and body size. Both traits have strong effects on fitness in bumble bees. Bumble bees

have a large number of parasites (Macfarlane & Gurr 1995; Schmid-Hempel 1998;

Sommeijer & de Ruijter 2000) which can reduce survival (Brown et al. 2000) and are

associated with reduced reproduction in bumble bee colonies (Müller & Schmid-Hempel

1992a, b). Here, we measured the efficiency of the cellular immune response as the degree of

encapsulation of a novel antigen. In nature, this immune response is triggered by antigens that

are too large to be phagocytosed, such as parasitoid eggs (Carton et al. 1992, Schmid-Hempel

& Schmid-Hempel 1996). The encapsulation process is associated with a visible melanization

of the formed capsule (reviewed in Ratcliffe et al. 1985; Götz 1986; Lackie 1988a; Karp

1990). The strength of this immune reaction depends on a number of factors, such as species

(discussed in Karp 1990), age and temperature (Blumberg & DeBach 1981), and genotype

(Carton & Boulétreau 1985; Carton et al. 1992; Carton & Nappi 1993). In bumble bees, the

encapsulation response varies among colonies and in response to demanding activities, such

as foraging (König & Schmid-Hempel 1995; Doums & Schmid-Hempel 2000). Large body

size is also advantageous in each of the bumble bee castes (discussed in Owen 1988; Sutcliffe

& Plowright 1988, and references therein). For example, large size increases hibernation

82

survival in queens, foraging efficiency in workers, and spermatozoa production in males.

Body size in bumble bees is strongly influenced by environmental effects (Owen 1988, 1989),

especially by larval food supply (Sutcliffe & Plowright 1988; Ribeiro 1994).

Here, we report on how inbreeding, sex, and ploidy affect the immune response and

body size in a social insect. We show that episodic inbreeding, as is likely for natural

populations, has no effect on the encapsulation response and body size of bumble bees, but

that family lineage, colony of origin, and caste all affect these measures of offspring quality.

Our results suggest a strong genetic influence on immune response while they corroborate a

strong environmental effect on body size.

Materials and Methods

Experimental design

We generated episodic inbreeding by brother-sister matings. For this purpose, young

queens and males of the bumble bee, Bombus terrestris (L.), were mated in the laboratory

between June 18 and August 9, 1997. These sexuals originated from laboratory colonies

established by spring queens collected in northern Switzerland in 1997. We applied two

treatments: inbred and outbred. To produce inbred colonies we mated daughter queens with

their brothers, resulting in diploid offspring (female and male) with an inbreeding coefficient

F = 0.25. To produce outbred colonies we mated daughter queens singly with unrelated males

originating from randomly allocated colonies. The queens used for mating originated from 19

different colonies ("maternal families"; Fig. 4.1), while the males originated from 28

different colonies. On average 19.5 ±1.2 sister queens (mean ± SE) per maternal family

(range 2-12 queens per treatment) were successfully mated. After copulation, we freeze-killed

the male at -80°C for later molecular analyses.

83

Maternal family 1 Matemal family 2 Matemal family 3

(sister queens) (sister queens) (sister queens)

? ? $ ? ? ? ? ? 9 S ? ?°o° ü o° H °o° H °° H °°° H o°

Experimental colonies

Fig. 4.1. Breeding scheme and experimental design. Maternal families are defined by colonies

arising from sister queens. For each maternal family, half of the queens were mated to one of

their brothers (inbred colonies; shaded rectangles), the other half were mated to an unrelated

male (outbred colonies; open rectangles). In each colony, measurements were obtained from a

sample of individual workers and males (circles).

We hibernated a total of 371 mated queens (188 inbred and 183 outbred matings) in a

climate cabinet at 6°C for 107 days. After hibernation, these queens were allowed to establish

the experimental colonies in a climate chamber under standard breeding conditions (29°C, 60

% r.H., red light). We provided pollen and sugar water (50 %-Apiinvert® solution) ad libitum.

When a queen had produced five offspring, we transferred the entire colony into a perlite

observation hive (Pomeroy & Plowright 1980). For further details on the mating, hibernation,

and rearing conditions, see Chapter 3. Eventually, we raised 69 colonies (32 inbred and 37

outbred, respectively) representing 14 maternal families. After having excluded the families

with too few or too small colonies, we included 39 colonies belonging to 6 maternal families

in the analyses. We compared two measures of offspring quality - immune response and

body size - of workers and males from inbred and outbred bumble bee colonies. Because

inbreeding depression varies among families (e.g., Byers & Waller 1999; Jarne et al. 2000),

we compare offspring from inbred and outbred colonies within the family line, as

recommended by Holsinger (1988,1991) and Uyenoyama et al. (1993).

84

Sampling of the test animals

Colonies that produced males later in the colony cycle (i.e., during the reproductive

phase only) were classified as colonies that produce only haploid males. Inbred colonies that

produced workers and males at about equal rates starting with the first brood were classified

as colonies that produce diploid males (after Duchateau et al. 1994). Hereafter, we refer to

these two colony types as "haploid male colonies" and "diploid male colonies",

respectively. We verified and corrected the classification of diploid male colonies posthoc

based on molecular analyses (see below). After a colony reached the size of fifteen

individuals, we removed the next ten workers to emerge for the immunological tests and size

measurements. For the diploid male colonies, we also collected the next ten males to emerge,

which were considered to be diploid. Male bumble bees do not forage or help to feed the

brood (but see Cameron 1985). Thus, our sampling of diploid males was not expected to

reduce colony growth disproportionally as compared to haploid male colonies where we

removed only workers. We collected the haploid males as follows: ten (from haploid male

colonies) or fourteen males (from diploid male colonies), respectively, were collected at the

peak of a colony's male production (usually on 1 or 2 days). This timing with respect to the

colony cycle ensured that diploid male colonies had also switched to haploid male production.

To avoid worker-produced males in our sample we excluded males collected 27 days or

longer after the queen's death, i.e., after the average developmental time of a B. terrestris male

under rearing conditions similar to ours (Duchateau & Velthuis 1988). We verified and

corrected the ploidy of all males collected from the diploid male colonies posthoc with

molecular markers (see below). We kept the test animals in groups per colony and caste

(workers, haploid or diploid males) in separate wooden boxes (20 x 15 x 11 cm) in the same

climate chamber as the colonies. We provided pollen and sugar water ad libitum.

85

Variables measured

Immune response.—We marked all test animals individually and tested their immune

response at seven days of age (7.02 ± 0.01 days post-eclosure, mean ± SE; n = 841, range 6-

8 days) to control for age effects (Doums et al. in press). We anaesthetised the animals briefly

with C02 and implanted an artificial "parasite" (a piece of nylon, diameter 0.16 mm, length

0.814 ± 0.003 mm, n = 841, range 0.492-1.215 mm) between the third and fourth steraite.

This novel antigen is thereby exposed to the circulating haemolymph and readily provokes an

immune response. No visible adverse effects of this treatment are known to occur (Schmid-

Hempel & Schmid-Hempel 1998) and the bumble bees quickly resume their normal activities.

Because the encapsulation process progresses rapidly (Lackie 1988a; Allander & Schmid-

Hempel 2000), we freeze-killed the animals two hours after implantation and stored them at

-20°C for later inspection. As described elsewhere (e.g., Schmid-Hempel & Schmid-Hempel

1998), we dissected the bumble bees under the dissecting microscope, recovered the implant,

and measured its darkness with a video imaging system (NIH Image 1.52, National Institutes

of Health, USA). This arbitrary but repeatable (Allander & Schmid-Hempel 2000) measure

reflects a combination of cellular encapsulation and degree of melanization, with higher values

indicating a stronger encapsulation response.

Body size.—As a general measure for body size, we used the mean radial cell length of

the left and right forewings (Bertsch 1984; Owen 1988, 1989; Müller & Schmid-Hempel

1992b). The measures were taken in units of 0.04 mm under a dissecting microscope.

Molecular discrimination of haploid and diploid males

We verified the ploidy of the sampled males for the diploid male colonies with

polymorphic microsatellite markers (Estoup et al. 1995) and protocols as in Schmid-Hempel

& Schmid-Hempel (2000). We used diagnostic loci where the queen did not share the same

alleles with her mate (loci B10-B11 in 5 colonies, and B124-B126 in 2 colonies). We scored a

86

male as diploid if it showed two alleles at the diagnostic locus, one of them being identical

with the allele of the queen's mate.

Statistical analyses

Statistical analyses were performed with SAS 6.12. (SAS Institute Inc. 1989) and SPSS

6.1.1 (Norusis 1994). Some family lines produced too few or too small colonies for

meaningful analysis and thus were excluded. For analysing family differences in inbreeding

depression, we log-transformed the data as recommended by Johnston & Schoen (1994). We

calculated a mixed-model MANOVA with workers or haploid males nested within colonies,

and colonies nested in the interaction between maternal family and treatment. The effect of

caste on immune response and body size was also tested with a mixed model MANOVA

(body size was transformed with xA to meet the assumptions of normality and

homoscedasticity). We excluded the males whose molecularly observed ploidy did not match

the ploidy expected from the time of sampling with respect to the colony cycle. This reduced

the variance of the covariate date of eclosure within the interaction between caste and colony,

and multiple testing. The excluded males were used to test whether development early or late

in the colony cycle affected offspring quality. Males we could not score molecularly were

excluded from all the analyses.

Seasonal timing is known to affect fitness in bumble bees, either through the production

of fewer sexuals for late-starting colonies (Müller & Schmid-Hempel 1992a) or lower

hibernation survival for late-reproducing colonies (Chapter 3). Therefore, we included the test

animals' date of eclosure as a covariate in the analyses. However, we calculated the

MANOVAs without covariates because the two variables measured had to be analysed

considering different covariates (date of eclosure and implant length for the immune response,

and date of eclosure only for the body size). We tested the effect of the covariates in the

univariate tests. For worker size, the interactions between date of eclosure and family and

between date of eclosure and colony were significant and thus were included in the final

87

analysis. Likewise, for the diploid male colonies, we included the interaction between date of

eclosure and colony in the final analysis of encapsulation response. Similar to earlier studies

(König & Schmid-Hempel 1995; Allander & Schmid-Hempel 2000; Doums & Schmid-

Hempel 2000), the implant length did not influence the immune response. All significance

values reported refer to two-tailed tests.

Results

Effects in workers and haploid males

A total of 39 colonies with 377 tested workers (mean ± SE per colony: 9.67 ± 0.12,

range 6-10) and 40 colonies with 382 haploid males (9.55 ± 0.20, range 6-14) from six

maternal families entered the final analyses (for the number of males excluded based on

molecular analyses, see below). Overall, the immune response of the workers, irrespective of

treatment and colony of origin, was stronger than the response of the haploid males (56.53 ±

0.93 units, mean ± SE, and 46.43 ± 0.85, respectively; Mann-Whitney U, Z = -8.393, P <

0.0001, n = 759) (Table 4.1). On the other hand, the workers were smaller in body size than

the haploid males (length of the radial cell: 2.83 ± 0.01 mm and 3.44 ± 0.01 mm, respectively;

Mann-Whitney U, Z= -22.439, P < 0.0001, n = 759) (Table 4.1).

88

Table 4.1. Average immune response (strength of encapsulation) and body size (length of

radial cell in forewings) of workers and haploid males of inbred and outbred colonies of the

bumble bee, B. terrestris. Immune response is given in arbitrary units, see text. The detailed

data are given in Tables A4.1 and A4.2.

Trait Workers Haploid males

inbred

(n = 157)

outbred

(n = 220)

inbred

(n = 165)

outbred

(n = 217)

Immune

response (SE)

Body size (SE)

[mm]

57.14 ±1.32

2.83 ± 0.02

56.09 ± 1.28

2.84 ± 0.02

46.22 ± 1.28

3.45 ± 0.02

46.59 ± 1.14

3.43 ± 0.01

Surprisingly, inbreeding did not affect immune response or body size in either workers

or haploid males (Tables 4.1,4.2, A4.1, A4.2). However, we found significant variation among

the maternal families and among all experimental colonies for both measures of offspring

quality (Figs. 4.2, 4.3, Tables 4.2, A4.1, A4.2). Maternal family affected different measures:

body size in workers and immune response in haploid males (univariate P-values; Table 4.2).

The only exception to the generally highly significant effect of the colony of origin was the

immune response of haploid males which did not depend on colony origin (univariate P-value;

Table 4.2 B). Results remained qualitatively similar if we included the covariates "date of

eclosure X family" (F6299 = 1.66, P - 0.132, n = 377) and "date of eclosure X colony"

CF33299 = 1.83, P = 0.005, n = 377) in the univariate analysis of worker size.

89

GO

+

in

ÖoOh

<U

Öo

&OÖ

(A) Immune response workers

80-

60'

40-

20-

201C 2620 2960 40

I

4847 5710

outbred

M inbred

11 17

Maternal family

(B) Immune response haploid males

00

+

GO

a,

D

GO

» ^H

+->

O

Cw

80-1

60-

40-

20-

18191 2628 2958 |2816i 58|50| 5!mm I utai.

X

14

outbred

H inbred

h 17

Maternal family

Fig. 4.2. Immune response of (A) workers and (B) haploid males from different maternal

families of the bumble bee, B. terrestris, with treatments as in Fig. 4.1. Numbers in bars

indicate sample sizes (individuals tested). For statistics, see Table 4.2. The exact data are givenin Table A4.1.

90

Table 4.2. MANOVA for immune response and body size with respect to maternal familyand treatment (inbred / outbred). (A) workers, n = 39 colonies tested, (B) haploid males, n =

40 colonies tested. The observed data are given in Tables 4.1, A4.1, and A4.2.

(A) Workers

Source of variation* Wilks A df F P Univariate F

effect,error

(P)

Immune Bodyresponse size

Family 0.447 10,52 2.577 0.013 0.60

(0.700)3.41

(0.016)

Treatmentf 0.968 2,26 0.432 0.654 0.19

(0.669)0.24

(0.630)

Family X Treatment 0.771 10,52 0.721 0.701 0.84

(0.535)0.34

(0.885)

Colony (Family X

Treatment)

0.673 54,674 2.731 0.0001 2.46

(0.0001)3.10

(0.0001)

(B) Haploid males

Source of variation* Wilks A df F P Univariate F

effect,error

(P)

Immune Bodyresponse size

Family 0.456 10,54 2.593 0.012 2.85

(0.033)2.00

(0.109)

Treatment! 0.904 2,27 0.144 0.255 1.94

(0.175)0.92

(0.345)

Family X Treatment 0.740 10,54 0.880 0.558 1.30

(0.535)0.56

(0.731)

Colony (Family X

Treatment)

0.500 56,682 5.052 0.0001 1.66

(0.292)9.50

(0.0001)

* Family and treatment are considered fixed effects, colony a random effect. Individuals are

nested within colonies, and single colonies are nested within the interaction between familyand treatment.

t Family X treatment is the error term

91

(A) Size workers

CO

<uN

00

2-

2010 26 20 4010 4847 5710

D outbred

M inbred

h 17

CO

SI

CO

Maternal family

(B) Size haploid males

r-.3-

18S 2628 2958 28&1 5850 5814

D outbred

H inbred

11 17

Maternal family

Fig. 4.3. Size of the radial cell in the forewings of (A) worker and (B) haploid male bumble

bees from different maternal families with treatments as in Fig. 4.1. Numbers in bars indicate

sample sizes (individuals tested). For statistics, see Table 4.2. The exact data are given in

Table A4.1.

92

In contrast to Doums & Schmid-Hempel (2000), we did not find a relationship between

a colony's average immune response and colony size, i.e., the number of workers produced

(correlation for immune response of workers: r - 0.076, n = 39, P = 0.648; of haploid males:

r = -0.252, n = 40, P = 0.117). Within colonies, the mean immune response of workers and

haploid males was positively correlated (r = 0.323, n = 38, P = 0.048). However, after

Bonferroni correction for multiple testing of the data, the latter correlation was no longer

significant.

Differences among castes in diploid male colonies

In seven inbred colonies that produced diploid males, we were able to collect enough

test animals for meaningful analyses. We screened 158 males molecularly for ploidy. 69

males were diploid, 12 of which were expected to be haploid based on the time of sampling (5

and 7 "late diploid males" from 2 colonies). 84 males were haploid, 13 of which were

expected to be diploid (3 and 10 "early haploid males" from 2 colonies). 5 males could not

be scored (1 presumably diploid and 4 presumably haploid males from 2 colonies). We

excluded one colony, from which all the males tested posthoc were haploid, and all the males

whose observed ploidy differed from the expected one. This left 6 diploid male colonies for

the final analyses. Per colony, we analysed 9.7 ± 0.2 workers (mean ± SE, range 9-10), 9.8 ±

1.2 haploid males (range 6-14), and 9.5 ± 0.6 diploid males (range 7-11).

The workers and haploid males produced in the diploid male colonies had a similar level

of immune response and body size as the bumble bees originating from the haploid male

colonies (MANOVA with "presence of diploid males" as fixed effect and colony as random

effect, colonies are nested within the presence of diploid males. For the workers: Wilks X =

0.955, F236 = 0.854, P = 0.434, n = 377 from 39 colonies; for the haploid males: Wilks X =

0.980, F237 = 0.375, P = 0.690, n = 382 from 40 colonies; factor "colony" was always

93

significant, P < 0.001). Unfortunately, the final sample of six diploid male colonies

represented four maternal families and was therefore too small to test for family effects.

Workers encapsulated more strongly than haploid males, who in turn encapsulated

more strongly than diploid males (colonies 2-5, Fig. 4.4; Table 4.3). For body size, workers

were smaller than haploid males while diploid males were of intermediate size (Fig. 4.5, Table

4.3). These clear patterns are reflected in a highly significant caste effect (Table 4.4). Again,

immune response and body size were significantly affected by the colony from which a

bumble bee originated (Table 4.4). The significant interaction between caste (workers, haploid

or diploid males) and colony indicates that the relative differences among castes varied among

colonies. The only difference between the multi- and univariate tests was the insignificant

interaction between caste and colony in the univariate test for immune response. Results

remained qualitatively similar if we included the covariate "date of eclosure X colony" (F5I48

= 1.58, P = 0.169, n = 171) in the univariate analysis of encapsulation response.

Table 4.3. Average immune response (strength of encapsulation) and body size (length of

radial cell in the forewings) of workers, haploid males, and diploid males of six diploid male

colonies. Immune response is given in arbitrary units, see text. The detailed data are given in

Table A4.2.

Trait Workers Haploid males Diploid males

(n = 58) (n = 59) (n = 57)

(n = 54)

Immune response (SE) 53.55 ± 2.84 46.26 ± 3.78 35.77 ± 2.03

Body size (SE) [mm] 2.77 ± 0.09 3.50 ± 0.07 3.03 ± 0.07

94

Immune response diploid male colonies

w00

GOOhcz)

CDVh

O

-4->

o

c

ffl

80-,

60-

40-

20-

Workers

Haploid males

Diploid males

Colony

Fig. 4.4. Immune response of three castes (workers, haploid males, and diploid males) in six

colonies producing diploid males of the bumble bee, B. terrestris. For statistics, see Table 4.4.

The exact data are given in Table A4.2.

Size diploid male colonies

00

<uN

00

D Workers

Haploid males

Diploid males

12 3 4 5 6

Colony

Fig. 4.5. Size of the radial cell in the forewings of three castes (workers, haploid males, and

diploid males) in six colonies producing diploid males. For statistics, see Table 4.4. The exact

data are given in Table A4.2.

95

Table 4.4. MANOVA for immune response and body size with respect to caste (workers,haploid males, and diploid males) and colony in six diploid male colonies, n = 171 individuals

tested. The observed data are given in Tables 4.3 and A4.2.

Source of variation* Wilks A df F P Univariate F

effect, (P)error Immune Body

^___response size

Caste 0.023 4,18 25.126 <0.0005 13.46 53.14

(0.001) (<0.0005)

Colony 0.624 10,304 8.084 <0.0005 2.50 15.01

(0.033) (<0.0005)

Caste x Colony 0.721 20,304 2.695 <0.0005 1.34 4.23

(0.216) (<0.0005)

Colony is considered a random effect, caste a fixed effect.

We tested whether early or late development during the colony cycle could explain the

observed caste differences in offspring quality. Within colonies, we compared the "early" to

the late haploid males (2 colonies) and the "late" to the early diploid males (2 colonies).

Timing of development did not affect encapsulation response or body size, either for the

haploid males (Mann-Whitney U, range for Z -6.103 to -1.037, all P-values > 0.174, n = 10

and n = 20, mean difference of eclosure date between groups 15.7 and 20.9 days,

respectively), or for the diploid males (Mann-Whitney U, range for Z -1.879 to -0.333, all P-

values > 0.062, n = 17 and n = 14, mean difference of eclosure date 21 and 25.3 days,

respectively). In contrast, late diploid males were significantly smaller than late haploid males

in body size (Mann-Whitney U, Z= -3.021, P = 0.003, n = 13, and Z = -2.689, P = 0.007, n

= 12, mean difference of eclosure date 17.7 and 12 days, respectively) but did not differ in

encapsulation response (Z = -2.286, P = 0.045 (the late diploid males had a lower immune

response; not significant after Bonferroni correction) and Z = -1.056, P = 0.291). The early

haploid males did not differ from the early diploid males in offspring quality (Mann-Whitney

96

U, Z = -1.026, P = 0.305, n = 10, and Z = -0.458, P = 0.647, n = 8 for encapsulation response

and body size, respectively, mean difference of eclosure date 9.9 days; data from one colony).

Discussion

Offspring quality in bumble bees as measured by immune response and body size was

not affected by our treatment of episodic inbreeding, mimicked by a single generation of

brother-sister mating. In addition, we did not detect any indirect effects of such inbreeding.

Neither the presence of inbred workers nor of diploid males influenced the quality of the

colony members. The finding that inbreeding did not affect the level of immune response in

bumble bee workers is in contrast to studies in vertebrates where inbreeding can increase

susceptibility to parasites (Allendorf & Leary 1986; Lively et al. 1990; Mitton 1995; Coltman

et al. 1999; Penn & Potts 1999). However, our result is similar to those found in other insects,

for example, the flour beetle Tribolium castaneum, where inbreeding was not associated with

lower resistance to parasitic nematodes but large amounts of among-lineage variation were

also detected (Stevens et al. 1997). It is possible, of course, that the benign laboratory

conditions during our experiment weakened any inbreeding effect, as inbreeding depression is

generally expected to be stronger under harsh conditions (Dudash 1990; del Castillo 1998;

Koelewijn 1998). Similarly, the observation that bumble bees from inbred colonies were of the

same size as bumble bees from outbred colonies (Tables 4.1, 4.2) could indicate that the

animals might have compensated for negative effects of inbreeding in a benign laboratory

environment with unlimited food supply. However, another laboratory study, which was based

partly on the colonies used for the present study, did find inbreeding effects, albeit for

different measures (Chapter 3). In fact, it is known from other study systems, that inbreeding

has varying effects on size or body mass in field as well as in laboratory studies. While some

97

studies reported inbreeding depression for size or body mass (e.g., Analla et al. 1999;

Bradshaw et al. 2000), others did not (e.g., Baur & Baur 2000; Kittelson & Maron 2000), and

still others reported varying responses among species (Berg & Redbo-Torstensson 1999) or

populations (Weeks et al. 2000).

A persistent finding in our study was that offspring quality varied significantly among

maternal families and among colonies (Tables 4.2, 4.4). Among-colony variation is a well-

known phenomenon in bumble bees and social Hymenoptera in general (e.g., Schmid-Hempel

1998). In contrast, among-family variation in bumble bees has been investigated only recently

(Chapter 3). The finding that immune response and body size are family-level characteristics

in bumble bees corroborates recent studies of similar traits in a variety of organisms such as

susceptibility to parasites in beetles (Stevens et al. 1997) and freshwater snails (J. Wiehn,

pers. comm.), and biomass and seed characteristics in plants (Mutikainen & Delph 1998 and

Kainer et al. 1999, respectively).

Among-family and among-colony differences allow for conclusions about the

influences of genes and environment on a given trait. In bumble bees, individuals with

different genetic backgrounds with respect to the maternal lineage grow up in a common

(colony) environment. The workers' genotypes are derived half from the maternal lineage, i.e.,

the queen, and half from the paternal lineage, i.e., the queen's mate. Hence, an effect of the

genotype would be seen at the colony level in workers. In contrast, the haploid males'

genotypes are derived solely from the maternal lineage because these males develop from

unfertilised eggs laid by the queen. Therefore, genetically, haploid males can be considered

gametes of the queen. Thus, an effect of genotype would show at the colony level of the

queen, i.e., the family level for haploid males. Environmental effects, on the other hand, should

be seen at the colony level in both workers and haploid males. Here, environmental effects

include the colony environment as well as maternal effects.

The encapsulation response of the workers varied significantly among colonies but not

among maternal families while encapsulation response of the haploid males varied among

98

maternal families but not among colonies (univariate tests, Table 4.2). Thus, our results

suggest strong genetic determinism of the encapsulation response, as is the case in

Drosophila melanogaster (e.g., Carton & Boulétreau 1985; Carton et al. 1992; Carton &

Nappi 1993). In contrast, our data corroborate a strong environmental determinism of body

size in B. terrestris. In both workers and haploid males, the common colony environment

influenced body size but only worker size was influenced by maternal family (Table 4.2).

Body size in bumble bees is strongly influenced by environmental effects (Owen 1988, 1989),

namely by larval food supply (Sutcliffe & Plowright 1988; Ribeiro 1994). Thus, the

differential influence of maternal family on worker and male size might be explained by the

differential importance of queen condition on worker and male brood, respectively. Queen

hibernation survival and colony foundation success are strongly affected by maternal family

(Chapter 3). Presumably, this is true also for the queen's condition after hibernation, when she

founds her colony and rears the first worker brood on her own (Alford 1975; Sladen 1989).

In contrast, males are usually reared later during the colony cycle and are predominantly fed

by workers (Alford 1975; Sladen 1989).

Due to our experimental design and the life-history of bumble bee colonies, diploidy in

males is confounded with inbreeding and early development during the colony cycle,

respectively. Inbreeding did not affect worker quality. Diploid males show the same degree of

inbreeding as inbred workers and differ genetically from them only in the zygosity at the sex

locus. Therefore, it is unlikely that inbreeding affected diploid male quality without affecting

worker quality. Early or late development during the colony cycle did not affect diploid male

quality, either.

The comparison of workers, haploid and diploid males allows for further insights into

the role of genetic components as compared to environmental effects, and into the roles of sex

(males vs. workers) as compared to ploidy (haploid vs. diploid) in bumble bees. By

comparing the three castes produced and reared in the same colony, we controlled for

environmental and genetic among-colony variation (in contrast to previous studies, Duchateau

99

& Marien 1995). Similarity between diploid and haploid males for a given trait would indicate

that this trait is influenced strongly by sex. On the other hand, similarity between diploid

males and workers would indicate a strong influence of ploidy or of common environment

during larval development. If diploid males are different from both haploid males and workers

for a given trait, an interaction between sex and ploidy would affect this trait. This last case

applied to both our measures of offspring quality, although the interaction between sex and

ploidy differed according to the trait. We found that diploid males encapsulated less strongly

than haploid males, who in turn encapsulated less strongly than workers while the body size

of diploid males was intermediate between the body size of workers and haploid males of the

same colony (Tables 4.3,4.4; Figs. 4.4,4.5).

For encapsulation response, the intra-colony differences among castes again support a

strong genetic component because environmental effects are likely to be very small. Diploid

males and workers differed strongly in immune response although they were reared

concurrently. In addition, diploid males that were reared later than the workers ("late diploid

males"), had the same level of immune response as the diploid males that were reared

concurrently with the workers. Neglectful feeding of the diploid male brood as compared to

the worker brood is also unlikely because the diploid males were larger than the workers of

the same brood. In bumble bees, size is influenced by larval food supply (Sutcliffe &

Plowright 1988; Ribeiro 1994). Furthermore, the data suggest that the gene (or genes)

determining the level of encapsulation response does not act additively, otherwise the diploid

males would have had a higher encapsulation response than the haploid males, possibly of the

same level as the diploid workers. Stronger immune response of bumble bee workers as

compared to haploid males of the same colony has also been found in outbred colonies

(Benelli 1998).

Body size, on the other hand, is known to be affected by environmental effects, namely

by larval food supply (Sutcliffe & Plowright 1988; Ribeiro 1994), which is related to duration

of development and feeding frequency (Ribeiro 1994; Ribeiro et al. 1999). The relative body

100

sizes of the three castes (Table 4.3, Fig. 4.5), however, indicate that ploidy also influences this

trait. The diploid workers and the diploid males are smaller than the haploid males of the same

colony. This size difference cannot be explained by early (diploid individuals) and late

(haploid individuals) development during the colony cycle. Diploid males were of the same

size and significantly smaller than haploid males from the same colony, irrespective of early or

late development (Duchateau & Marien 1995; this study). A possible mechanism, by which

ploidy might influence body size would be a shorter larval development time of diploid

individuals as compared to haploid individuals. An intermediate body size of diploid males

was also found by Duchateau and Marien (1995) although the haploid males they tested

originated from different colonies than the workers and diploid males. The body size ranges

measured in the present study compare very well to the ones given by Duchateau and Marien

(1995).

In conclusion, sex and ploidy affect offspring quality more than (episodic) inbreeding

in bumble bees. While one generation of brother-sister mating did not decrease the level of

immune response or body size in workers, diploid males had a weaker immune response and

were smaller than haploid males. Thus, deviation from the usual combinations of female sex

and diploidy on the one hand, and male sex and haploidy on the other hand decreased

offspring quality significantly.

Acknowledgements

We thank N. Krüger and H. Magro for help with the experiments, J. Jokela and J.

Wiehn for help with the statistics, and M. Brown for comments on the manuscript. This study

was financially supported by grants from the Swiss National Science Foundation (no. 31-

101

49040.96 to PSH) and the Swiss Federal Institute of Technology (ETH) Zürich (no. 0-20-

010-95 to PSH).

References

Alford, D.V. 1975. Bumblebees. Davis-Poynter, London, UK.

Allander, K. & Schmid-Hempel, P. 2000. Immune defence reaction in bumble-bee workers

after a previous challenge and parasitic coinfection. Funct. Ecol. 14:711-717.

Allendorf, F.W. & Leary, R.F. 1986. Heterozygosity and fitness in natural populations of

animals, pp. 57-76 in M.E. Soulé, ed. Conservation biology: the science of scarcity and

diversity .Sinauer Associates Inc., Sunderland, MA, USA.

Analla, M., Montilla, J.M. & Serradilla, J.M. 1999. Study of the variability of the response to

inbreeding for meat production in Merino sheep. J. Anim. Breeding Genet. 116:481-

488.

Baur, B. & Baur, A. 2000. Social facilitation affects longevity and lifetime reproductive

success in a self-fertilizing land snail. Oikos 88:612-620.

Beekman, M., van Stratum, P. & Veerman, A. 1999. Selection for non-diapause in the

bumblebee Bombus terrestris, with notes on the effect of inbreeding. Entomol. Exp.

Appl. 93:69-75.

Benelli, E.F. 1998. Ecological and adaptive aspects of immunocompetence in a social insect.

Ph.D. thesis, Swiss Federal Institute of Technology (ETH), Zürich, Switzerland.

Berg, H. & Redbo-Torstensson, P. 1999. Offspring performance in three cleistogamous Viola

species. Plant Ecol. 145:49-58.

Bertsch, A. 1984. Foraging in male bumblebees {Bombus lucorum L.): maximizing energy or

minimizing water load? Oecologia 62:325-336.

102

Blumberg, D. & DeBach, P. 1981. Effects of temperature and host age upon the

encapsulation of Metaphycus stanleyi and Metaphycus helvolus eggs by brown soft

scale Coccus hesperidum. J. Invertebr. Pathol. 37:73-79.

Bradshaw, J.E., Todd, D. & Wilson, R.N. 2000. Use of tuber progeny tests for genetical

studies as part of a potato {Solanum tuberosum subsp. tuberosum) breeding

programme. Theor. Appl. Genet. 100:772-781.

Brehélin, M., ed. 1986. Immunity in invertebrates: cells, molecules, and defense reactions.

Springer-Verlag, Berlin, Germany.

Brown, M.J.F., Loosli, R. & Schmid-Hempel, P. 2000. Condition-dependent expression of

virulence in a trypanosome infecting bumblebees. Oikos 91:421-427.

Burt, A. & Bell, G. 1992. Tests of sib diversification theories of outcrossing in Impatiens

capensis: effects of inbreeding and neighbour relatedness on production and infestation.

J. Evol. Biol. 5:575-588.

Buttermore, R.E., Pomeroy, N., Hobson, W., Semmens, T. & Hart, R. 1998. Assessment of

the genetic base of Tasmanian bumble bees (Bombus terrestris) for development as

pollination agents. J. Apic. Res. 37:23-25.

Byers, D.L. & Waller, D.M. 1999. Do plant populations purge their genetic load? Effects of

population size and mating history on inbreeding depression. Annu. Rev. Ecol. Syst.

30:479-513.

Cameron, S.A. 1985. Brood care by male bumble bees. Proc. Natl. Acad. Sei. USA 82:6371-

6373.

Carton, Y. & Boulétreau, M. 1985. Encapsulation ability of Drosophila melanogaster. a

genetic analysis. Dev. Comp. Immunol. 9:211-220.

Carton, Y., Frey, F. & Nappi, A. 1992. Genetic determinism of the cellular immune reaction in

Drosophila melanogaster. Heredity 69:393-399.

103

Carton, Y. & Nappi, A. 1993. Methods for genetic investigation of cellular immune reaction in

insects, with the parasitic wasp-Drosophila system as a model, pp 91-101 in J.P.N.

Pathak, ed. Insect immunity. Kluwer Academic Publishers, Dordrecht, The Netherlands.

Chapman, B.R. & George, J.E. 1991. The effects of ectoparasites on cliff swallow growth and

survival, pp. 69-92 in J.E. Loye & M. Zuk, eds. Bird-parasite interactions: ecology,

evolution, and behaviour, Vol. 2. Oxford University Press, Oxford, UK.

Christe, P., M0ller, A.P. & de Lope, F. 1998. Immunocompetence and nestling survival in the

house martin: the tasty chick hypothesis. Oikos 83:175-179.

Clayton, D.H. & Moore, J., eds. 1997. Host-parasite evolution: general principles and avian

models. Oxford University Press, Oxford, UK.

Coltman, D.W., Pilkington, J.G., Smith, J.A. & Pemberton, J.M. 1999. Parasite-mediated

selection against inbred Soay sheep in a free-living, island population. Evolution

53:1259-1267.

Cook, J.M. 1993. Sex determination in the Hymenoptera: a review of models and evidence.

Heredity 71:421-435.

Cook, J.M. & Crozier, R.H. 1995. Sex determination and population biology in the

Hymenoptera. Trends Ecol. Evol. 10:281-286.

Cornuet, J.-M. 1986. Population genetics, pp. 235-262 in T.E. Rinderer, ed. Bee genetics and

breeding. Academic Press, Orlando, FL, USA.

Crozier, R.H. 1975. Insecta 7. Hymenoptera. Gebrüder Borntraeger, Berlin, Germany,

de Lope, F., M0ller, A.P. & de la Cruz, C. 1998. Parasitism, immune response and

reproductive success in the house martin Delichon urbica. Oecologia 114:188-193.

del Castillo, R.F. 1998. Fitness consequences of maternal and nonmatemal components of

inbreeding in the gynodioecious Phacelia dubia. Evolution 52:44-60.

Donovan, B.J. & Weir, S.S. 1978. Development of hives for field population increase, and

studies on the life cycle of four species of introduced bumble bees in New Zealand. N.

Z. J. Agric. Res. 21:733-756.

104

Doums, C., Moret, Y., Benelli, E.F. & Schmid-Hempel, P. Senescence of immune defence in

Bombus workers. Ecol. Entomol. In press.

Doums, C. & Schmid-Hempel, P. 2000. Immunocompetence in workers of a social insect,

Bombus terrestris L., in relation to foraging activity and parasitic infection. Can. J.

Zool. 78:1060-1066.

Duchateau, M.J., Hoshiba, H. & Velthuis, H.H.W. 1994. Diploid males in the bumble bee

Bombus terrestris: sex determination, sex alleles and viability. Entomol. Exp. Appl.

71:263-269.

Duchateau, M.J. & Marien, J. 1995. Sexual biology of haploid and diploid males in the

bumble bee Bombus terrestris. Insect. Soc. 42:255-266.

Duchateau, M.J. & Velthuis, H.H.W. 1988. Development and reproductive strategies in

Bombus terrestris colonies. Behaviour 107:186-207.

Dudash, M.R. 1990. Relative fitness of selfed and outcrossed progeny in a self-compatible,

protandrous species, Sabatia angularis L. (Gentianaceae): a comparison in three

environments. Evolution 44:1129-1139.

Estoup, A, Scholl, A., Pouvreau, A. & Solignac, M. 1995. Monoandry and polyandry in

bumble bees (Hymenoptera: Bombinae) as evidenced by highly variable microsatellites.

Mol. Ecol. 4:89-93.

Estoup, A., Solignac, M., Cornuet, J.M., Goudet, J. & Scholl, A. 1996. Genetic differentiation

of continental and island populations of Bombus terrestris (Hymenoptera: Apidae) in

Europe. Mol. Ecol. 5:19-31.

Falconer, D.S. 1989. Introduction to quantative genetics, 3rd edn. Longman Scientific &

Technical, Harlow, UK.

Frank, S.A. 2000. Specific and non-specific defense against parasitic attack. J. Theor. Biol.

202:283-304.

Frankham, R. 1995. Conservation genetics. Annu. Rev. Genet. 29:305-27.

105

Goh, K.T., Ooi, P.L., Miyamura, K., Ogino, T. & Yamazaki, S. 1990. Acute hemorrhagic

conjunctivitis: seroepidemiology of coxsackievirus A24 variant and enterovirus 70 in

Singapore. J. Med. Virol. 31:245-247.

Götz, P. 1986. Encapsulation in arthropods, pp. 153-170 in M. Brehélin, ed. Immunity in

invertebrates: cells, molecules, and defense reactions. Springer-Verlag, Berlin, Germany.

Gupta, A.P. 1986a. Arthropod immunocytes: identification, structure, functions, and analogies

to the functions of vertebrate B- and T-lymphocytes. pp. 3-59 in A.P. Gupta, ed.

Hemocytic and humoral immunity in arthropods. John Wiley & Sons, New York, NY,

USA.

Gupta, A.P., ed, 1986b. Hemocytic and humoral immunity in arthropods. John Wiley & Sons,

New York, NY, USA.

Harder, L.D. 1986. Influences on the density and dispersion of bumble bee nests

(Hymenoptera: Apidae). Holarctic Ecol. 9:99-103.

Hoffmann, J.A., Kafatos, F.C., Janeway, CA., Jr. & Ezekowitz, R.A.B. 1999. Phylogenetic

perspectives in innate immunity. Science 284:1313-1318.

Hoffmann, J.A. & Reichhart, J.M. 1997. Drosophila immunity. Trends Cell Biol. 7:309-316.

Holsinger, K.E. 1988. Inbreeding depression doesn't matter: the genetic basis of mating-

system evolution. Evolution 42:1235-1244.

—. 1991. Inbreeding depression and the evolution of plant mating systems. Trends Ecol.

Evol. 6:307-308.

Imhoof, B. & Schmid-Hempel, P. 1999. Colony success of the bumble bee, Bombus

terrestris, in relation to infections by two protozoan parasites, Crithidia bombi and

Nosema bombi. Insect. Soc. 46:233-238.

Jarne, P., Perdieu, M.A., Pernot, A.F., Delay, B. & David, P. 2000. The influence of self-

fertilization and grouping on fitness attributes in the freshwater snail Physa acuta:

population and individual inbreeding depression. J. Evol. Biol. 13:645-655.

106

Johnsen, A., Anderson, V., Sunding, C. & Lifjeld, J.T. 2000. Female bluethroats enhance

offspring immunocompetence through extra-pair copulations. Nature 406:296-299.

Johnston, M.O. & Schoen, D.J. 1994. On the measurement of inbreeding depression.

Evolution 48:1735-1741.

Kainer, K.A., de Matos Malavasi, M., Duryea, M.L. & Rodrigues da Silva, E. 1999. Brazil nut

(Bertholletia excelsd) seed characteristics, preimbibition and germination. Seed Sei.

Technol. 27:731-745.

Karp, R.D. 1990. Cell-mediated immunity in invertebrates. Bioscience 40:732-737.

Kittelson, P.M. & Maron, J.L. 2000. Outcrossing rate and inbreeding depression in the

perennial yellow bush lupine, Lupinus arboreus (Fabaceae). Am. J. Bot. 87:652-660.

Klein, S.L. & Nelson, R.J. 1999. Social interactions unmask sex differences in humoral

immunity in voles. Anim. Behav. 57:603-610.

Koelewijn, H.P. 1998. Effects of different levels of inbreeding on progeny fitness in Plantago

coronopus. Evolution 52:692-702.

König, C. & Schmid-Hempel, P. 1995. Foraging activity and immunocompetence in workers

of the bumble bee, Bombus terrestris L. Proc. R. Soc. Lond. B 260:225-227.

Kukovetz, E.M., Bratschitsch, G., Hofer, H.P., Egger, G. & Schaur, R.J. 1997. Influence of

age on the release of reactive oxygen species by phagocytes as measured by a whole

blood chemiluminescence assay. Free Rad. Biol. Med. 22:433-438.

Lackie, A.M. 1988a. Haemocyte behaviour. Adv. Insect Physiol. 21:85-178.

—. 1988b. Immune mechanisms in insects. Parasitai. Today 4:98-105.

Lacy, R.C. 1997. Importance of genetic variation to the viability of mammalian populations. J.

Mammal. 78:320-335.

Lehmann, T. 1993. Ectoparasites: direct impact on host fitness. Parasitai. Today 9:8-13.

Lively, CM., Craddock, C. & Viijenhoek, R.C. 1990. Red queen hypothesis supported by

parasitism in sexual and clonal fish. Nature 344:864-866.

107

Macfarlane, R.P. & Gurr, L. 1995. Distribution of bumble bees in New Zealand. N. Z.

Entomol. 18:29-36.

McCurdy, D.G., Shutler, D., Mullie, A. & Forbes, M.R. 1998. Sex-biased parasitism of avian

hosts: relations to blood parasite taxon and mating system. Oikos 82:303-312.

Mitton, J.B. 1995. Enzyme heterozygosity and developmental stability. Acta Theriol. 40:33-

54.

M0ller, A.P. 1996. Parasitism and developmental instability of hosts: a review. Oikos 77:189-

196.

M0ller, A.P., Sorci, G. & Erritz0e, J. 1998. Sexual dimorphism in immune defense. Am. Nat.

152:605-619.

Moritz, R.F.A. 1985. Inbreeding effects in drones (Apis mellifera L.). Insect. Soc. 32:104-

105.

Müller, C.B. & Schmid-Hempel, P. 1992a. Correlates of reproductive success among field

colonies of Bombus lucorum: the importance of growth and parasites. Ecol. Entomol.

17:343-353.

—. 1992b. Variation in life-history pattern in relation to worker mortality in the bumble-bee,

Bombus lucorum. Funct. Ecol. 6:48-56.

Mutikainen, P. & Delph, L.F. 1998. Inbreeding depression in gynodioecious Lobelia

siphilitica: among-family differences override between-morph differences. Evolution

52:1572-1582.

Norusis, M.J. 1994. SPSS Advanced statistics 6.1. SPSS, Chicago, IL, USA.

O'Brien, S.J. & Evermann, J.F. 1988. Interactive influence of infectious disease and genetic

diversity in natural populations. Trends Ecol. Evol. 3:254-259.

Ochoa, G. & Jaffe, K. 1999. On sex, mate selection and the Red Queen. J. Theor. Biol.

199:1-9.

108

Ouborg, N.J., Bière, A. & Mudde, CL. 2000. Inbreeding effects on resistance and

transmission-related traits in the Silène-Microbotryurn pathosystem. Ecology 81:520-

531.

Owen, R.E. 1988. Body size variation and optimal body size of bumble bee queens

(Hymenoptera: Apidae). Can. Entomol. 120:19-28.

—. 1989. Differential size variation of male and female bumblebees. J. Heredity 80:39-43.

Pamilo, P. & Crozier, R.H. 1997. Population biology of social insect conservation. Mem.

Museum Victoria 56:411-419.

Pathak, J.P.N. 1993. Cell-mediated defence reactions in insects, pp. 47-58 in J.P.N. Pathak,

ed. Insect immunity, Vol. 48 Kluwer Academic Publishers, Dordrecht, The Netherlands.

Penn, D.J. & Potts, W.K. 1999. The evolution of mating preferences and major

histocompatibility complex genes. Am. Nat. 153:145-164.

Plass, F. 1953. Inzuchtwirkung und Heterosiseffekt bei der Honigbiene. Landwirtschaftl.

Informationsdienst 66:49-68.

Plowright, R.C. & Pallett, M.J. 1979. Worker-male conflict and inbreeding in bumble bees

(Hymenoptera: Apidae). Can. Entomol. 111:289-294.

Pomeroy, N. & Plowright, R.C. 1980. Maintenance of bumble bee colonies in observation

hives (Hymenoptera: Apidae). Can. Entomol. 112:321-326.

Ratcliffe, N.A., Rowley, A.F., Fitzgerald, S.W. & Rhodes, C.P. 1985. Invertebrate immunity:

basic concepts and recent advances. Int. Rev. Cytol. 97:183-350.

Ribeiro, M.F. 1994. Growth in bumble bee larvae: relation between development time, mass,

and amount of pollen ingested. Can. J. Zool. 72:1978-1985.

Ribeiro, M.F., Velthuis, H.H.W., Duchateau, M.J. & van der Tweel, I. 1999. Feeding

frequency and caste differentiation in Bombus terrestris larvae. Insect. Soc. 46:306-

314.

109

Roberts, C.W., Cruickshank, S.M. & Alexander, J. 1995. Sex-determined resistance to

Toxoplasma gondii is associated with temporal differences in cytokine production.

Infect. Immun. 63:2549-2555.

Rodrigânez, J., Tora, M.A., Rodriguez, M.C. & Silio, L. 1998. Effect of founder allele survival

and inbreeding depression on litter size in a closed line of Large White pigs. Anim. Sei.

67:573-582.

Roff, D.A. 1997. Evolutionary quantitative genetics. Chapman & Hall, New York, NY, USA.

Saino, N., Canova, L., Fasola, M. & Martinelli, R. 2000. Reproduction and population density

affect humoral immunity in bank voles under field experimental conditions. Oecologia

124:358-366.

Sankilampi, U., Isoaho, R., Blouigu, A., Kivela, S.L. & Leinonen, M. 1997. Effect of age, sex

and smoking habits on pneumococcal antibodies in an elderly population. Int. J.

Epidemiol. 26:420-427.

SAS Institute Inc. 1989. SAS/STAT® user's guide, version 6, 4th edn. SAS Institute Inc.,

Cary, NC, USA.

Schmid-Hempel, P. 1998. Parasites in social insects. Princeton University Press, Princeton,

NJ, USA.

Schmid-Hempel, P. & Koella, J.C. 1995. Ecology and genetics of virulence and resistance.

pp. 589-600 in W.A. Nierenberg, ed. Encyclopedia of environmental biology, Vol. 1.

Academic Press, San Diego, CA, USA.

Schmid-Hempel, R. & Schmid-Hempel, P. 1996. Larval development of two parasitic flies

(Conopidae) in the common host Bombus pascuorum. Ecol. Entomol. 21:63-70.

—. 1998. Colony performance and immunocompetence of a social insect, Bombus terrestris,

in poor and variable environments. Funct. Ecol. 12:22-30.

—. 2000. Female mating frequencies in Bombus spp. from Central Europe. Insect. Soc.

47:36-41.

110

Sladen, F.W.L. 1989. The humble-bee. Its life history and how to domesticate it. Logaston

Press, Little Logaston, UK.

Sommeijer, M.J. & de Ruijter, A., eds. 2000. Insect pollination in greenhouses: proceedings

of the specialists' meeting held in Soesterberg, The Netherlands, 30 September to 2

October 1999. University of Utrecht, The Netherlands.

Stevens, L., Yan, G. & Pray, L.A. 1997. Consequences of inbreeding on invertebrate host

susceptibility to parasitic infection. Evolution 51:2032-2039.

Strauss, S.Y. & Karban, R. 1994. The significance of outcrossing in an intimate plant-

herbivore relationship: II. Does outcrossing pose a problem for thrips adapted to the

host-plant clone? Evolution 48:465-476.

Sutcliffe, G.H. & Plowright, R.C. 1988. The effects of food supply on adult size in the

bumble bee Bombus terricola Kirby (Hymenoptera: Apidae). Can. Entomol. 120:1051-

1058.

Theis, J.H. & Schwab, R.G. 1992. Seasonal prevalence of Taenia taeniaeformis: relationship

to age, sex, reproduction and abundance of an intermediate host {Peromyscus

maniculatus). J. Wildlife Dis. 28:42-50.

Thompson, J.N. & Burdon, J.J. 1992. Gene-for-gene coevolution between plants and

parasites. Nature 360:121-125.

Thornhill, N.W. 1993. The natural history of inbreeding and outbreeding: theoretical and

empirical perspectives. The University of Chicago Press, Chicago, IL, USA.

Uyenoyama, M.K., Holsinger, K.E. & Waller, D.M. 1993. Ecological and genetic factors

directing the evolution of self-fertilization. Oxf. Surv. Evol. Biol. 9:327-381.

Wakabayashi, A., Utsuyama, M., Hosoda, T., Sato, K. & Hirokawa, K. 1999. Differential age

effect of oral administration of an antigen on antibody response: an induction of

tolerance in young mice but enhancement of immune response in old mice. Mech.

Ageing Dev. 109:191-201.

Ill

Wakelin, D. & Apanius, V. 1997. Immune defence: genetic control, pp. 30-58 in D.H.

Clayton & J. Moore, eds. Host-parasite evolution: general principles and avian models.

Oxford University Press, Oxford, UK.

Weeks, S.C., Crosser, B.R., Bennett, R., Gray, M. & Zucker, N. 2000. Maintenance of

androdioecy in the freshwater shrimp, Eulimnadia texana: estimates of inbreeding

depression in two populations. Evolution 54:878-887.

Whiting, P.W. 1940. Multiple alleles in sex determination of Habrobracon. J. Morph.

66:323-355.

Widmer, A., Schmid-Hempel, P., Estoup, A. & Scholl, A. 1998. Population genetic structure

and colonization history of Bombus terrestris s.l. (Hymenoptera: Apidae) from the

Canary Islands and Madeira. Heredity 81:563-572.

Yan, G. 1997. Consequence of larval tapeworm infection for the fitness of the intermediate

hosts, flour beetles {Tribolium spp.). Can. J. Zool. 75:271-279.

112

Seite LeerBlank

113

5 A linkage analysis of sex determination in Bombus

terrestris (L.) (Hymenoptera: Apidea)

(GadauJ., GerloffC.U., Krüger N., ChanH., Schmid-Hempel P., Wille A. & Page

R.E., Jr., in revision for Heredity)

Abstract

We constructed a linkage map of Bombus terrestris (L.) (Hymenoptera, Apidae)

phase unknown. The map contains 79 markers (six microsatellite and 73 RAPD

markers) in 21 linkage groups and spans over 953.1 cM. The minimal recombinational

size of the B. terrestris genome was estimated to be 1073 cM. Using flow cytometry,

the physical size of the haploid genome of B. terrestris was calculated to be 274 Mb.

This is the second linkage map for a social insect species. B. terrestris has on average

five times less recombinational events per kb than the honey bee, Apis mellifera. We

now can exclude several structural explanations like haplo-diploidy or small

chromosomes for the high recombination frequency in A. mellifera. Additionally, we

could also falsify that eusociality per se favors an increase in recombination frequency

because B. terrestris is eusocial as well. Most likely the high recombination frequency

of A. mellifera is associated with the more complex and elaborated system of division of

labor in A. mellifera or with high parasite loads in this species.

Finally, we confirmed a single locus complementary sex determination

mechanism in B. terrestris. Bulk segregation analysis of diploid workers and diploid

114

males was a quick and efficient way to find molecular markers linked with the sex locus

in B. terrestris.

Keywords

Bombus terrestris; complementary sex determination; diploid males; linkage map;

mapping phase unknown.

Introduction

Since Dzierzon's (1845) pioneering work on honey bees it has long been known

that male Hymenoptera are haploid and develop from unfertilized eggs. This so called

haplo-diploid sex determination is used by approximately 20 % of all animals (Bull

1983). Under haplo-diploidy males develop from unfertilized eggs and are haploid

whereas females are diploid and arise from fertilized eggs. The most widespread

proximate mechanism of sex determination in Hymenoptera seems to be the

complementary sex determination system (CSD, Whiting 1943; for reviews see Crozier

1977 & Cook 1993). The idea behind CSD is that individuals which are hemi- (=

haploid) or homozygous at the sex locus/i develop into phenotypic males whereas

heterozygous individuals are females (Whiting 1943). A particular case of CSD is the

single locus CSD (e.g., Butcher et al. 2000), i.e., there is only a single sex determination

locus. If species with a single locus CSD are inbred (e.g., a sister-brother mating) we

expect that half of the diploid brood will be phenotypic males usually called diploid

males. Those diploid males are either inviable, sterile or produce sterile (triploid)

daughters. Therefore, the effects of inbreeding in Hymenoptera with a single locus CSD

are severe. Evidence for single locus CSD is mostly provided by controlled breeding

experiments and the occurrence of diploid males at the expected frequency (Cook

115

1993). However, the molecular basis of the haplo-diploid sex determination system in

Hymenoptera remains unknown.

Apis mellifera L. (the honey bee) has a single locus CSD and is the only social

hymenopteran species for which the location of the sex determination locus is currently

mapped both in a linkage and a physical map (Hunt & Page 1994; Beye et al. 1996).

Here, we extent this analysis by mapping the sex locus in the bumblebee, Bombus

terrestris (L.) in order to provide a second system for the investigation of the molecular

mechanisms of sex determination under haplo-diploidy. Bumble bees belong to the

same family as honeybees (Apidae). As a study system, bumble bees have the

advantage that diploid males are raised to adulthood, in contrast to honey bees where

they are normally removed as larvae (Woyke 1963).

B. terrestris is a well-studied and widespread species in Europe and breeding

experiments suggested a single locus CSD (Duchateau et al. 1994). Another breeding

study with a neotropical bumble bee, B. atratus, proposed a two loci CSD system

(Garöfalo 1973). We used molecular markers and linkage mapping to map the sex

determination locus of B. terrestris. The research aims at confirming the single locus

CSD for B. terrestris. It is also a first step to ultimately isolate sex-determining genes

and understand the molecular mechanisms that determine sex in the haplo-diploid

system of Hymenoptera. Diploid males and females (workers or gynes in the case of

social Hymenoptera) derived from one queen differ under a single locus CSD system

only in the allelic composition at the sex locus; diploid males have to be homozygous

whereas workers have to be heterozygous. This can be used to map the sex

determination locus because molecular markers that are linked with the sex locus should

show the same pattern.

One recurrent problem for the mapping of social Hymenoptera is that only few

species can be bred in the laboratory where control over ancestry and descending lines

is possible. For example, for social insect colonies collected in the field, the pedigree is

116

generally unknown. Therefore, only two generation families are available for the

mapping and information about the linkage phase is missing. Methods for mapping

phase unknown have been developed and implemented in a couple of mapping

programs (e.g., CRIMAP, Lander & Green 1987). Since most of these programs are

written for small family sizes and diploid organisms like humans, we have developed a

simple method to construct linkage maps phase unknown using Mapmaker version 2.0

(Lander et al. 1987). MAPMAKER has the advantage that it can handle haploid data

and family size is unlimited, features of a mapping population typical for social

Hymenoptera. With our new approach it is possible to generate linkage maps for almost

every hymenopteran species as long as enough offspring are produced by a single

female which is normally no problem for eusocial Hymenoptera. Additionally, it makes

the screening of the grandparents unnecessary and avoids long lasting and difficult

breeding experiments, thus saving time and resources.

In this study we generated a linkage map for B. terrestris based on RAPD and

microsatellite markers. We used males derived from a queen for which we had no

marker information of her parents (= phase unknown). We tested our results derived

from our method of mapping phase unknown with a second mapping program

(CRIMAP, Lander & Green 1987) that had a mapping phase unknown routine

implemented. We could place the sex determining locus on this linkage map and

confirmed a single locus complementary sex determination system for B. terrestris.

Finally, we determined the physical genome size of B. terrestris and could therefore

calculate the average amount of recombination frequency per physical unit.

117

Materials and Methods

Origin of bumble bees

All tested individuals (workers, haploid and diploid males) were offspring from one

female mated to one of her brothers. Both were derived from the brood of a queen

caught in spring 1997 around Zürich, Switzerland. The mated female hibernated at 5°C

in the laboratory for four months. Then she was put into a rearing box (acrylic glass,

12.5 x 7.5 x 5.5 cm) in a climate chamber set at 29°C and 60 % relative humidity with a

callow worker of another colony to stimulate egg laying. This worker was removed as

soon as the young queen had produced five individuals, and the entire incipient colony

was transferred into a perlite nest (Pomeroy & Plowright 1980). The colony eventually

produced 57 workers, 409 males (haploid and diploid) and 18 young queens. As

expected for diploid male producing colonies, males were produced concurrently with

workers. The ploidy status of haploid and diploid males was confirmed with

microsatellites (B124-B126, Estoup et al. 1995) and four codominant RAPD markers

(fragment length polymorphism). The mapping population consisted of 116 of the

haploid males.

DNA extraction and RAPD-PCR reactions

DNA from half the thorax of individual males and females was isolated with a

standard CTAB-Phenol extraction method (Hunt & Page 1995). The RAPD-PCR

reactions (Williams et al. 1990) were carried out in 12.5 /xl reaction volumes using 5 ng

of genomic DNA, 0.6 pM primer, 100 juM each of dATP, dCTP, dGTP and dTTP

(Pharmacia), 10 mM Tris-HCL (pH 8.3), 50 mM KCl, 2 mM MgCl2, and 0.75 U Taq.

The ten nucleotide RAPD primers were obtained from Operon Technologies (Alameda,

CA, USA) and the University of British Columbia Biotechnology Center (Vancouver,

Canada). Amplification was performed with the following parameters: 5 cycles of 94°C

118

/ 1 min, 35°C / 1 min and 72°C / 2 min, another 32 cycles at 94°C / 10 s, 35°C / 30 s

and 72°C / 30 s.

Gel electrophoresis and scoring

The amplification products were resolved in 20 x 25 cm horizontal gels using 1 %

Synergel (Diversified Biotech, Newton Center, MA) and 0.6 % Agarose in a 0.5 x TBE

buffer. Gels were run for 500-600 Vh, stained in ethidium bromide for 25 min,

destained in distilled water for another 40 min, and recorded on an UV transilluminator

with Polaroid 667 films. After the map and the linkage groups were established all

markers were ordered according to their position in the linkage groups, and all gels were

scored a second time. This allowed to control for unlikely double crossovers due to

scoring errors.

Microsatellite PCR and scoring

The mother (queen) of the mapping population was heterozygous for 8 of 9

microsatellite loci tested. Six of the heterozygous loci (BIO, Bll, B116, B118, B124,

and B126) were used for further linkage analysis. PCR detection of the microsatellites

was done according to Schmid-Hempel & Schmid-Hempel (2000). The reaction

mixtures contained each 1-10 ng of total DNA, and multiplex PCRs were run for the

loci B10-B11, B116-B118, and B124-B126 (Estoup et al. 1995).

Phase unknown linkage analyses

For mapping it is normally necessary to assign a linkage phase to every marker,

i.e., the alleles coming from the grandmother are coded differently than the alleles

coming from the grandfather. As the grandparents of the mother (queen of our colony)

of our mapping population were not available to determine the linkage phase, we

119

designed the following simple procedure to map phase unknown in MAPMAKER

(Lander et al. 1987), that has no phase unknown mapping procedure implemented.

1. First we assigned a phase arbitrarily to every allele of our markers (our

convention was to assign 1 to present alleles or the longer allele of a fragment length

polymorphism, and 0 to absent or the shorter alleles, respectively).

2. Then the complete dataset (each genotype of every marker from our mapping

population) was doubled and the phase of the doubled markers was flipped i.e., 1 was

changed to 0 and 0 to 1, respectively. Since we had haploid individuals as mapping

population we now had a data set with each possible linkage phase.

3. MAPMAKER (Lander et al 1987, version 2.0 for Macintosh, data type was

coded as "haploid") was used to do a complete two-point analysis of this doubled

dataset. This procedure allowed us to determine the phase and calculated simultaneously

the recombination frequencies between linked markers. Once we knew the phase of the

markers we discarded all markers with the wrong phase for the next step, the multi¬

point analysis. Note, only the high progeny number allowed us to determine the phase

of linked markers in our approach.

4. For the following multi-point analysis phase was known and we could proceed

with a normal linkage analysis.

The mapping procedure with Mapmaker followed a standard protocol described

below:

1. A two-point linkage analysis of the doubled dataset (194 markers, see results)

was performed with the "GROUP" command (setting: LOD = 3.0; theta = 0.4) to find a

preliminary set of linkage groups. Note, that this was the doubled data set and that every

linkage group had therefore to be present twice. All of the linkage groups we found

followed this prediction. Using this procedure we determined the linkage phases of the

markers. Once phase information is established, multi-point analysis could be carried

out with a phase known data-set.

120

2. Multi-point analysis within all putative linkage groups generated in step 1 were

done with the "FIRSTORDER" command (LOD = 3.0, theta = 0.4). This analysis gave

the most likely order of the markers in each linkage group.

3. Using the "RIPPLE" command, the order found in step 2 was tested within

each linkage group for all possible three-point orders of consecutive markers. The most

likely order for every marker is shown. Markers that were linked at 2 cM or less could

not be ordered at a LOD 2 threshold because with the given size of the mapping

population there are too few informative méioses.

All map distances (cM) were calculated from recombination fractions (%)

according to Kosambi's mapping function (Kosambi 1944) because Kosambi's function

resulted in less map expansion than Haldane's function when the "DROP MARKER"

command was used.

Test for the phase unknown mapping procedure -1

We used CRIMAP - a program in which a phase unknown mapping procedure is

implemented - to see whether we would get the same result as with our method. To

generate a phase unknown data set for CRIMAP we introduced a third non-informative

allele for each marker. This simulated a diploid mapping population with an

uninformative father which is necessary to run CRIMAP. Then we did a two-point

analysis to generate linkage groups.

Test for the phase unknown mapping procedure - II

In order to test our phase unknown mapping procedure empirically with a known

linkage map we performed two tests with the Apis mellifera map data set which has 365

mapped markers (Hunt & Page 1995). In the first test we took 100 linked markers

(randomly chosen), doubled the data, and flipped phase for the doubled data. Then we

121

ran MAPMAKER (Lander et al. 1987) to see whether we got exactly the same linkage

groups as in the original map. This test was performed twice.

In the second test we took the markers of the three biggest linkage groups to

check whether we would get the same ordering of markers within the linkage groups.

Although this was not a rigorous statistical test it demonstrated the power of our

mapping phase unknown approach.

Comparing linkage maps and relative map sizes between species

Most of the published maps for non-model organisms are unsaturated (average

marker density > 10 cM). Therefore, estimates of real map sizes for these species are

lacking. However, relative map sizes may be obtained and compared by controlling for

marker numbers as a covariate; we can construct maps with equal numbers of markers

and compare their sizes. A larger genome would yield a larger map size with the same

number of markers. For example, to compare the map sizes of A. mellifera (Hunt &

Page 1995), Nasonia spp. (Gadau et al. 1999), and B. terrestris (this study) the linked

markers of all data sets (A. mellifera = 365 markers, Nasonia = 91 markers, and B.

terrestris = 79 markers) were randomized and linkage of 20, 40, 60, and 80 (79 in the

case of B. terrestris) markers were calculated for each species using MAPMAKER with

the default settings. For all species, we then calculated the total map size as the size of

each linkage group plus 40 cM to account for all unlinked markers. Our justification for

adding 40 cM is that all markers must be linked somewhere, but were left unlinked,

during the mapping procedure, because they had no companion markers within 40 cM

of them (given by the setting of theta = 0.40). Our expectation is that genomes with

overall higher rates of recombination will be relatively much larger. This is because

fewer markers are likely to be linked together into a single linkage group if the linkage

map of a species is large. This effect is expected to be particularly evident when few

122

markers are used. However, as marker numbers increase, linkage groups should

coallesce, resulting in a slower increase in map size as the map approaches saturation.

Physical genome size of B. terrestris

To determine the physical genome size of B. terrestris we conducted standard

flow cytometry of thoracic muscle cells (for detailed methodology see Otto 1994). As

size standard we used thoracic muscle cells of A. mellifera since the physical size of the

haploid genome of A. mellifera is known to be 178 Mb (Jordan & Brosemer 1974; Hunt

& Page 1995).

Mapping of the sex locus

To find markers segregating for the sex determination locus we combined the

DNA of six diploid males and six workers (derived from the same queen as the haploid

males of the mapping population) into separate samples (= bulk). We screened for any

marker present in the diploid males but absent in the worker and vice versa ("bulk

segregation analysis", e.g., de Tomaso et al. 1998). Markers segregating between

diploid males and workers were independently tested with 6 different individual diploid

males and workers, respectively. Markers which again segregated consistently between

diploid males and workers were tested on an additional 48 diploid males and 48 workers

to confirm linkage with the sex locus and derive an estimate of the distance between

these markers and the sex locus. Every marker which shows variability between the

diploid males and workers must also be variable in the haploid males of the same

colony, because the queen must be heterozygous for this marker. Therefore, having

identified markers linked to the sex locus they immediately could be integrated into the

linkage map derived from the 116 males of the mapping population.

123

Results

Linkage map

A total of 1119 RAPD primers (Operon primer sets A-Z and UBC primers 1-599)

were pre-screened on four haploid males of the mapping population. The best 65 RAPD

primers generated 91 segregating nuclear markers; that is an average of 1.4 markers per

RAPD primer. Additionally, six microsatellite markers segregated in the mapping

population. Of the RAPD markers 71 (78%) showed present/absent polymorphism

whereas 20 markers (22%) displayed fragment length polymorphisms. The recovery

rate of all alleles was normally distributed (0.50 ± 0.07 = mean ± SD; Fig. 5.1).

56

52

48

44

w 40

O

IS 36

£

S 32

x>O

28

o

24

20

16

12

8

4

0

CD

E

28,4%

0,5%

0,3-0,35 0,4-0,45 0,5-0,55 0,6-0,65 0,7-0,75

0,25-0,3 0,35-0,4 0,45-0,5 0,55-0,6 0,65-0,7 > 0,75

Recoveryrate of alleles

Fig. 5.1. Plot of the recovery rates (number of individuals with allele 1/number of all

individuals tested) for all marker. The recovery rate is normally distributed (mean ± SD

= 0.50 ± 0.07).

124

MAPMAKER (LOD 3.0; theta 0.40; Lander et al. 1987) mapped 79 of 97 markers

(82 %) into 21 linkage groups (Fig. 5.2). The map spanned 953.1 cM with an average

marker spacing of 12.1 cM. Since B. terrestris has 18 chromosomes (Hoshiba et al.

1995) three of the 21 linkage groups must be linked with other linkage groups. Thus, the

total genome size was estimated to be 1073 cM, assuming that three of the linkage

groups have to be linked and that gaps represent at least 40 cM due to our setting of

MAPMAKER (theta = 0.40).

Test for the phase unknown mapping procedure -1

The two-point linkage analyses with CRIMAP produced the exact same linkage

groups as we did with our approach with MAPMAKER (doubling the dataset and flip-

phase). However, there were three markers, V17-0.87 (LG I), 570- 0.55 (LG II) and

465-0.54 (LG IV) which were linked with LOD scores smaller than 3.0 (2.75, 2.84, and

2.93, respectively) but had higher LOD scores when we used our approach with

MAPMAKER (3.05, 3.14, and 3.23, respectively). In general all LOD scores in

CRIMAP were 0.3 units smaller than the LOD scores in our MAPMAKER approach

(results not shown). This was due to the following differences in calculating the LOD

scores in CRIMAP and MAPMAKER:

MAPMAKER calculated the LOD scores simply by using the likelihood ratio:

L(theta)/L(0.5) = [theta"(rec) * (1 -theta)n(non-rec)]/(0.5)n

Where, theta = recombination fraction; n = number of individuals genotyped; n(rec) =

number of putative recombinants; n(non-rec) = number of putative non-recombinants),

because although we generated two additional markers (by doubling and phase-

switching) the two-point LOD score calculation only uses the information of two loci at

a time.

In contrast, CRIMAP uses the more appropriate formula for the likelihood ratio:

125

L(theta)/L(0.5) = [(0.5 * theta"(rec) * (1 - theta)"(non-rec)) + 0.5 * (theta"(nonrec) * (1 -

theta)n(rec))]/(0.5)"

because it combines the probabilities for both possible phases in the calculation of the

L(theta).

If n is large the likelihood ratio calculated by MAPMAKER is twice the

likelihood ratio computed by CRIMAP, which leads to a MAPMAKER LOD score that

is inflated by log10 2 = 0.3.

To account for the overestimation of the LOD scores in our approach we could

simply increase our LOD threshold by 0.3 units. However, since the size of the linkage

map of B. terrestris is only around 1000 cM an appropriate LOD threshold for the

linkage of two markers in B. terrestris would be 2.5 (see p.68 in Ott (1999) for the

argumentation of LOD thresholds). Therefore, using a LOD score threshold of 3.0 with

our approach already compensates for this deviation.

Once we have established the phase of a marker we can simply discard the

markers with the "wrong" phase and proceed with a multi-point analysis in the "usual

way" (see Material and Methods).

Testing the phase unknown mapping procedure - II

Using our approach with the 395 RAPD marker set of Apis mellifera (Hunt &

Page 1995) we generated the same linkage groups and the same order of markers within

the linkage groups as in the original map of Hunt & Page (1995) where phase was

known.

Mapping of the sex determination locus

We screened 499 primers (UBC primers 100-599) in the bulk segregation analysis

of two DNA template mixtures of six diploid males and six workers each. Two markers

(237-0.30 and 389-0.36) differed between bulks and also segregated in the screen of six

126

different individual workers and six diploid males. A consecutive run on 48 diploid

males and 48 workers each revealed that both markers were linked to the sex locus

(237-0.30 = 16.5 cM and 389-0.36. = 7.5 cM). Since the markers also segregated in the

haploid males of the mapping population the sex locus could be incorporated into the

linkage map. Only one sex determination locus was found and it mapped into a single

linkage group IX (Fig. 5.2), together with five other markers (three additional Operon

markers which were previously scored for the haploid males only, see Fig. 5.2). Only

one of these three Operon markers (Z7-4.0) could also be mapped in the diploid workers

and males. The other two Operon markers could not be scored in diploid individuals

because the mate of the queen had the present allele. Note, that RAPD markers are

generally dominant markers so that in diploid individuals heterozygotes of the allelic

type present/absent cannot be distinguished from homozygous of the type

present/present. The most likely position of the sex locus is given in Fig. 5.2. The

distances between the flanking markers (389-0.36 and 237-0.30) of the sex locus in

linkage group IX, derived from the diploid individuals and the haploid mapping

population, were almost identical (24.1 cM versus 24 cM, results not shown). However,

the distances between the marker 237-0.30 and Z7-0.49 were quite different in the

haploid and diploid map (10.8 versus 1.1 cM).

209

170

195

254

226

•574-0.59f

.17-134f

.N13-0

85f

X20-03H

.V17-087

.125-058

526-78f

II

129

V

276

216

120

388

--

169-068

VF8-1

41

ms-B124

359-075

M3-064

222-057

570-055

204

314

165

198

.A7-1

11

525-061

f

IV

172

53

349

ms-B126

00

44

18

11

X20-1

11f

J%X2

0-055

31-076

Q17-054

101-050

VI

27

316

145

.B20-089f

143

S19-036

ms-B11

VII

305

E11-089

288

.237-047

-465-054

__

-132-110

/rff\

Z7-060f

J\359-098f

242

V:

VIII

82

.490-050f

.H8-044

.504-065

.B20-071

.262-023f

.521-055

.F13-042

.A4-096

511-087

U11-063

.434-103

DC

48

215

24

1

108

XI

XIII

F8-053

E11-053

222

389-036

126

sex-locus

237-030

Z7-049

.J7-1

11

XIV

.177-135f

.A19-085

,207

.A19-1.22

XVI

XII

-4=

297

425-059

Y13-0

501

ms-B10

.157-105

108-086f

.531-1.25f

151-16

XVII

204

XVIII

[]128-1

"U"^.

A13-0

35f

48

XIX

192

105

164-108f

179-080f

277-050

125-053

.J7-087

XX

232

49

S12-096f

115-062f

U16-095

XXI

N11-034

11-(h:-;00

7f

I-073

Fig.

5.2.Linkagemap

ofB.

terrestrisbasedonRAPD

markers.Markers

arenamedby

theirprimersdesignation.

Theprimerdesignation

isa

letterand

anumber

forpr

imer

sofOperon

Technologies,Alameda,CA,USA,

or

anumber

forprimersfrom

theUniversity

of

British

Columbia,Vancouver,Canada.Theprimername

isfollowedby

adashand

theapproximate

sizeoftheamplifiedfr

agme

ntinkilobases.

Markersshowingfr

agme

nt-l

engt

hpolymorphisms

areindicatedwithan

"f

'

foll

owin

gthemarkername.The

position

ofthesex

locuswas

determinedbycomparing

thedistancebetween

thetwomarkers389-0.36

(7.5cM)and237-0.30(1

6.5cM)andtheph

enot

ypic

sex

indiploid

malesandworkers,and

thedistanceofthesemarkers

inha

ploi

dmales

(24.1cM).The

sexlocusmustbe

locatedbetween

thesemarkers

because

16.5cMand7.5cM

(the

distancebetweenthesexdeterminationlocusandthemarkers)sumup

to24.0cM.

128

Physical genome size of B. terrestris

Thoracic muscle cells of B. terrestris contained 1.54 times more DNA than the

cells of A. mellifera (N = 50,000 cells, coefficients of variance (%) for A. mellifera and

B. terrestris were 2.53 and 2.52, respectively). The physical size of the haploid genome

of B. terrestris was thus calculated to be 274 Mb (= 1.54 * 178 Mb).

Discussion

Genome map

Bombus terrestris is the second social insect species for which a linkage map is

now available. The map contains 79 markers and spans over 953.1 cM. The minimal

recombinational size of the B. terrestris genome was estimated to be 1073 cM. The

linkage map of B. terrestris is thus significantly smaller than the linkage map of the

closely related honey bee (Apis mellifera) but it is similar to other hymenopteran species

(Fig. 5.3). On the other hand, the physical genome of B. terrestris was estimated to be

274 Mb and is therefore 1.54 times larger than that of A. mellifera (178 Mb). Therefore,

1 cM in the linkage map of B. terrestris equals on average 255 kb in physical units. In

contrast one cM in the honey bee map equals 50 kb (Hunt & Page 1995). This means

that on average A. mellifera has an astonishing five times more recombinational events

per kb than B. terrestris (see also Fig. 5.3). This is especially interesting because A.

mellifera belongs to the same family, Apidae, as B. terrestris but has one of the highest

recombination frequencies ever reported for higher eukaryotes (Hunt & Page 1995). We

now can exclude several structural explanations like haplo-diploidy or small

chromosomes for this high recombination frequency in A. mellifera as originally

proposed by Hunt & Page (1995) because both A. mellifera and B. terrestris have a

single locus CSD system and both have small chromosomes (Hoshiba et al. 1995). We

129

can also exclude that the evolution of sociality favors an increase in recombination

frequency per se because B. terrestris is eusocial as well. Rather, it appears more likely

that "external", ecological reasons have driven the evolution of high recombination

frequencies in A. mellifera. For example, the high recombination frequency of A.

mellifera could be associated with the more complex and elaborated system of division

of labor in A. mellifera (Page & Robinson 1991) or with high parasite loads in this

species (Otto & Michalakis 1998; Schmid-Hempel 1998). Division of labor is thought

to profit from an increased recombination because it increases the genotypic and

presumably phenotypic diversity among workers within a colony for multigenic traits

assuming that these variable genes are linked together on chromosomes. This could lead

to a more stable division of labor. If division of labor is a determinant of the high

recombination frequency in A. mellifera, we would expect that closely related species

like A. dorsata and very distantly related eusocial hymenopteran species with a very

similar type of division of labor like the leaf cutter ants (e.g., species from the genus

Acromyrmex) should have equally high recombination frequencies. To evaluate this

hypothesis we are currently working on linkage maps for Acromyrmex echinatior.

Alternatively, parasites and pathogens can theoretically produce an increase in

recombination (Peters & Lively 1999). An increase of recombination would produce a

genetically more diverse worker force and might so hamper the colonization of a whole

colony by a single pathogen or parasite genotype.

130

0 20 40 60 80 100

NUMBER OF MARKERS

Fig. 5.3. Comparison of relative map sizes between Apis mellifera, Bombus terrestris,and three parasitic Hymenoptera (Bracon hebetor, Antolin et al. 1996; Trichogrammabrassicae, Laurent et al. 1998; Nasonia spp., Gadau et al. 1999). Dashed lines are the 95

% confidence limits of the regression line. It is very obvious that A. mellifera has a

significantly higher recombination frequency than all other hymenopteran species for

which linkage maps are published.

Mapping of the sex-locus

The bulk segregation analysis of diploid workers and diploid males was a quick

and efficient way to find molecular markers linked with the sex locus in B. terrestris.

Since the diploid individuals were produced by the same female (queen) as the haploid

mapping population we could directly incorporate the molecular markers linked with

the sex locus into the linkage map of B. terrestris.

Our results (Fig. 5.2) confirmed a single locus sex determination mechanism in B.

terrestris (Duchateau et al. 1994), and makes a two locus sex determination system as

proposed for B. atratus, a neotropical bumblebee (Garöfalo 1973), unlikely.

Alternatively, it is possible that in B. atratus different selection pressures like high

131

inbreeding frequencies might have favored the evolution of a multilocus CSD system

but to our knowledge nothing is known about inbreeding frequencies in this species.

However, we cannot exclude the possibility of additional QTL (quantitative trait

loci) influencing the sex determination in B. terrestris as shown by Holloway et al.

(2000) for Bracon sp. near hebetor because we have not done a QTL analysis of the sex

determination of the diploid individuals - workers and diploid males - for all markers.

However, both breeding and the molecular data now support a single locus CSD

mechanism for B. terrestris.

The distance between the sex locus and the closest marker of our linkage map is

7.5 cM, i.e., approximately 1.9 Mb in physical units. This is prohibitively large to

identify the gene directly by genome walking but it will be the starting point in future

studies to search for markers associated more closely with the sex locus in. Clearly,

knowledge of the genomic map, including the sex locus, in B. terrestris has a variety of

exciting applications, from the possibility to characterize sex genes to studies of

selection and maintenance of allelic diversity at the population level.

A linkage map of B. terrestris containing 79 markers and which is produced by a

newly designed phase unknown mapping procedure is sufficient to map a single locus

trait - here the sex determination locus - by using a bulk segregation approach. This

result is promising for future projects in B. terrestris including mapping of quantitative

traits like disease resistance or behavior. In general this new mapping approach

combined with highly variable anonymous molecular markers (AFLP, RAPD, etc.)

opens a new avenue towards a fast and efficient mapping of non-model organisms

which cannot be bred in the laboratory as is the case for most social Hymenoptera.

132

Acknowledgements

This work was supported by a Feodor-Lynen Forschungsstipendium to JG, a grant

by the Swiss NSF to PSH (nr. 3100-049040.95), and a grant to REP (PHS-MH5311).

We thank C. Steinlein and M. Schmid for their help with the flow cytometry.

References

Antolin, M.F., Bosio, CF., Cotton, J., Sweeney, W., Strand, M.R. et al. 1996. Intensive

linkage mapping in a wasp (Bracon hebetor) and a mosquito {Aedes aegypti) with

Single-Strand Conformation Polymorphism analysis of randomly amplified

Beye, M., Moritz, R.F.A., Crozier, R.H. & Crozier, Y.C. 1996. Mapping the sex locus

of the honey bee (Apis mellifera). Naturwissenschaften 83:424-426.

Bull, J.J. 1983. Evolution of sex determining mechanisms. Benjamin/Cummings, Menlo

Park, CA, USA.

Butcher R.D.J., Whitfield, W.G.F. & Hubbard, S.F. 2000. Single locus complemenary

sex determination in Diademan chrysotictes (Gmelin) (Hymenoptera:

Ichneumonidae). J. Hered. 91:104-111.

Cook, J.M. 1993. Sex determination in the Hymenoptera: a review of models and

evidence. Heredity 71:421-435.

Crozier, R.H. 1977. Evolutionary genetics of the Hymenoptera. Annu. Rev. Entomol.

22, 263-288.

Duchateau, M.J., Hoshiba, H. & Velthuis, H.H.W. 1994. Diploid males in the bumble

bee Bombus terrestris: sex determination, sex alleles and viability. Entomol. Exp.

Appl. 71:263-269.

133

Dzierzon, J. 1845. Gutachten über die von Herrn Direktor Stöhr im ersten und zweiten

Kapitel des General-Gutachtens aufgestellten Fragen. Eichstädter Bienenzeit.

1:109-113,1:119-121.

de Tomaso A.W., Saito, Y., Ishizuka, K.J., Palmeri, K.J. & Weissman, I.L. 1998.

Mapping the genome of a model protochordate. I. A low resolution genetic map

encompassing the fusion/histocompatibility (Fu/HC) locus of Botryllus schlössen.

Genetics 149:277-287.

Estoup, A., Scholl, A., Pouvreau, A. & Solignac, M. 1995. Monandry and polyandry in

bumble bees (Hymenoptera; Bombinae) as evidenced by highly variable

microsatellites. Mol. Ecol. 4:89-93.

Gadau J., Page, R.E., Jr. & Werren, J.H. 1999. Mapping of hybrid incompatibility loci

in Nasonia. Genetics 153:1731-1741.

Garöfalo, C.A. 1973. Occurrence of diploid drones in a neotropical bumblebee.

Experentia 29:726-727.

Holloway, A.K., Strand, M.R., Black, W.C, IV, & Antolin, M.F. 2000. Linkage

Analysis of sex determination in Braeon sp. near hebetor (Hymenoptera:

Braconidae). Genetics 154:205-212.

Hoshiba, H., Duchateau, M.J. & Velthuis, H.H.W. 1995. Diploid males in the bumble

bee Bombus terrestris (Hymenoptera): karyotype analyses of diploid females,

diploid males and haploid males. Jpn. J. Ent. 63:203-207.

Hunt, G.J. & Page, R.E., Jr. 1994. Linkage analysis of sex determination in the honey

bee (Apis mellifera). Mol. Gen. Genet. 244:512-518.

—. 1995. Linkage map of the honey bee, Apis mellifera, based on RAPD markers.

Genetics 139:1371-1382.

Jordan, J.R. & Brosemer, R.W. 1974. Characterization of DNA from three different bee

species. J. Insect. Physiol. 20:2513-2520.

134

Kosambi D., 1944. The estimation of map distances from recombination values. Ann.

Eugen. 12:172-175.

Lander E.S. & Green, J. 1987. Construction of multilocus genetic linkage maps in

human. Proc. Natl. Acad. Sei. USA. 84:2363-2367.

Lander E.S., Green, P., Abrahamson, J., Barlow, A., Daly, M.J. et al. 1987.

MAPMAKER: an interactive computer package for constructing primary genetic

linkage maps of experimental and natural populations. Genomics 1:174-181.

Laurent, V., Wajnberg, E., Mangin, B., Schiex, T., Gaspin, C. et al. 1998. A composite

genetic map of the parasitoid wasp Trichogramma brassicae based on RAPD

markers. Genetics 150: 275-282.

Ott, J. 1999. Analysis of human genetic linkage. 3rd edn. The Johns Hopkins University

Press, Baltimore, MD, USA.

Otto, F.J. 1994. High resolution analysis of nuclear DNA employing the fluorochrome

DAPI. pp. 211-217 in Z. Darzynkiewicz, J. P. Robinson & H.A. Crissman, eds.

Methods in cell biology. Academic Press, San Diego, CA, USA.

Otto, S.P. & Michalakis, Y. 1998. The evolution of recombination in changing

environments. Trends Ecol. Evol. 13:145-151.

Page, R.E., Jr. & Robinson, G.E. 1991. The genetics of division of labor in honey bee

colonies. Adv. Insect. Physiol. 23:117-169.

Peters, A.D. & Lively, CM. 1999. The Red Queen and fluctuating epistasis: a

population genetic analysis of antagonistic coevolution. Am. Nat. 154:393-405.

Pomeroy, N. & Plowright, R.C 1980. Maintenance of bumble bee colonies in

observation hives (Hymenoptera: Apidae). Can. Entomol. 112:321-326.

Schmid-Hempel, P. 1998. Parasites in social insects. Princeton University Press,

Princeton, NJ, USA.

Schmid-Hempel, R. & Schmid-Hempel, P. 2000. Female mating frequencies in social

insects: Bombus spp. from Central Europe. Insect. Soc. 47:36-41.

135

Whiting, P.W. 1943. Multiple alleles in complementary sex determination of

Habrobracon. Genetics 28:365-382.

Williams, J.G.K., Kubelik, A.R., Livak, K.K., Rafalski, J.A. & Tingey, S.V. 1990. DNA

polymorphisms amplified by arbitrary primers are useful as genetic markers.

Nucleic Ac. Res. 18:6531-6335.

Woyke, J. 1963. What happens to diploid drones in a honey bee colony? J. Apic. Res.

2:73-76.

136

Seite Leer /

Blank leaf

137

6 Bibliography

Adamson, M.L. 1989. Evolutionary biology of the Oxyurida (Nematoda): Biofacies of a

haplodiploid taxon. Adv. Parasitol. 28:175-228.

Àgren, J. & Schemske, D.W. 1993. Outcrossing rate and inbreeding depression in two

annual monoecious herbs, Begonia hirsuta and B. semiovata. Evolution 47:125-135.

Alford, D.V. 1975. Bumblebees. Davis-Poynter, London, UK.

Allander, K. & Schmid-Hempel, P. 2000. Immune defence reaction in bumble-bee workers

after a previous challenge and parasitic coinfection. Funct. Ecol. 14:711-717.

Allendorf, F.W. & Leary, R.F. 1986. Heterozygosity and fitness in natural populations of

animals, pp. 57-76 in M.E. Soulé, ed. Conservation biology: the science of scarcity

and diversity .Sinauer Associates Inc., Sunderland, MA, USA.

Amiet, F. 1996. Hymenoptera Apidae, 1. Teil. Schweizerische Entomologische

Gesellschaft, Neuchâtel, Switzerland.

Anaila, M., Montilla, J.M. & Serradilla, J.M. 1999. Study of the variability of the response

to inbreeding for meat production in Merino sheep. J. Anim. Breeding Genet.

116:481-488.

Antolin, M.F., Bosio, CF., Cotton, J., Sweeney, W., Strand, M.R. et al. 1996. Intensive

linkage mapping in a wasp (Bracon hebetor) and a mosquito {Aedes aegypti) with

Single-Strand Conformation Polymorphism analysis of randomly amplified

polymorphic DNA markers. Genetics 142:1727-1738.

Bailleul, S. 1993. Observations on the earwigs from the Cher region (Dermaptera).

Entomologiste 49:268.

138

Barranco, P., Cabrero, J., Camacho, J.P.M. & Pascual, F. 1995. Chromosomal basis for a

bilateral gynandromorph in Pycnogaster inermis (Rambur, 1838) (Orthoptera,

Tettigoniidae). Contrib. Zool. 65:123-127.

Baur, B. & Baur, A. 2000. Social facilitation affects longevity and lifetime reproductive

success in a self-fertilizing land snail. Oikos 88:612-620.

Beekman, M. & van Stratum, P. 1998. Bumblebee sex ratios: why do bumblebees produce

so many males? Proc. R. Soc. Lond. B 265:1535-1543.

Beekman, M., van Stratum, P. & Lingeman, R. 1998. Diapause survival and post-diapause

performance in bumblebee queens (Bombus terrestris). Entomol. Exp. Appl. 89:207-

214.

Beekman, M., van Stratum, P. & Veerman, A. 1999. Selection for non-diapause in the

bumblebee Bombus terrestris, with notes on the effect of inbreeding. Entomol. Exp.

Appl. 93:69-75.

Beekman, M. & van Stratum, P. 2000. Does the diapause experience of bumblebee queens

Bombus terrestris affect colony characteristics? Ecol. Entomol. 25:1-6.

Benelli, E.F. 1998. Ecological and adaptive aspects of immunocompetence in a social

insect. Ph.D. thesis, Swiss Federal Institute of Technology (ETH), Zürich,

Switzerland.

Berg, H. & Redbo-Torstensson, P. 1999. Offspring performance in three cleistogamous

Viola species. Plant Ecol. 145:49-58.

Berkelhamer, R.C. 1983. Intraspecific genetic variation and haplodiploidy, eusociality, and

polygyny in the Hymenoptera. Evolution 37:540-545.

Bertsch, A. 1984. Foraging in male bumblebees {Bombus lucorum L.): maximizing energy

or minimizing water load? Oecologia 62:325-336.

139

Beye, M., Moritz, R.F.A., Crozier, R.H. & Crozier, Y.C. 1996. Mapping the sex locus of

the honey bee (Apis mellifera). Naturwissenschaften 83:424-426.

Blumberg, D. & DeBach, P. 1981. Effects of temperature and host age upon the

encapsulation of Metaphycus stanleyi and Metaphycus helvolus eggs by brown soft

scale Coccus hesperidum. J. Invertebr. Pathol. 37:73-79.

Bowers, M.A. 1985. Bumble bee colonization, extinction, and reproduction in subalpine

meadows in northeastern Utah. Ecology 66:914-927.

Bradshaw, J.E., Todd, D. & Wilson, R.N. 2000. Use of tuber progeny tests for genetical

studies as part of a potato (Solanum tuberosum subsp. tuberosum) breeding

programme. Theor. Appl. Genet. 100:772-781.

Breed, M.D., Page, R.E., Jr., Hibbard, B.E. & Bjostad, L.B. 1995. Interfamily variation in

comb wax hydrocarbons produced by honey bees. J. Chem. Ecol. 21:1329-1338.

Brehélin, M., ed. 1986. Immunity in invertebrates: cells, molecules, and defense reactions.

Springer-Verlag, Berlin, Germany.

Brian, A.D. 1952. Division of labor and foraging in Bombus agrorum Fabricius. J. Anim.

Ecol. 2:223-240.

Brown, J.H. 1984. On the relationship between abundance and distribution of species. Am.

Nat. 124:255-279.

Brown, M.J.F., Loosli, R. & Schmid-Hempel, P. 2000. Condition-dependent expression of

virulence in a trypanosome infecting bumblebees. Oikos 91:421-427.

Buchmann, S.L. & Nabhan, G.P. 1996. The forgotten pollinators. Island Press, Washington

D.C., USA.

Bull, J.J. 1983. Evolution of sex determining mechanisms. Benjamin/Cummings, Menlo

Park, CA, USA.

140

Burt, A. & Bell, G. 1992. Tests of sib diversification theories of outcrossing in Impatiens

capensis: effects of inbreeding and neighbour relatedness on production and

infestation. J. Evol. Biol. 5:575-588.

Butcher R.D.J., Whitfield, W.G.F. & Hubbard, S.F. 2000. Single locus complemenary sex

determination in Diademan chrysotictes (Gmelin) (Hymenoptera: Ichneumonidae). J.

Hered. 91:104-111.

Buttermore, R.E. 1997. Observations of successful Bombus terrestris (L.) (Hymenoptera:

Apidae) colonies in Southern Tasmania. Aust. J. Entomol. 36:251-254.

Buttermore, R.E., Pomeroy, N., Hobson, W., Semmens, T. & Hart, R. 1998. Assessment of

the genetic base of Tasmanian bumble bees {Bombus terrestris) for development as

pollination agents. J. Apic. Res. 37:23-25.

Byers, D.L. & Waller, D.M. 1999. Do plant populations purge their genetic load? Effects of

population size and mating history on inbreeding depression. Annu. Rev. Ecol. Syst.

30:479-513.

Calow, P. 1998. The encyclopedia of ecology & environmental management. Blackwell

Science, Oxford, UK.

Cameron, S.A. 1985. Brood care by male bumble bees. Proc. Natl. Acad. Sei. USA

82:6371-6373.

Carton, Y. & Boulétreau, M. 1985. Encapsulation ability of Drosophila melanogaster. a

genetic analysis. Dev. Comp. Immunol. 9:211-220.

Carton, Y., Frey, F. & Nappi, A. 1992. Genetic determinism of the cellular immune

reaction in Drosophila melanogaster. Heredity 69:393-399.

Carton, Y. & Nappi, A. 1993. Methods for genetic investigation of cellular immune

reaction in insects, with the parasitic wasp-Drosophila system as a model, pp 91-101

141

in J.P.N. Pathak, ed. Insect immunity. Kluwer Academic Publishers, Dordrecht, The

Netherlands.

Chapman, B.R. & George, J.E. 1991. The effects of ectoparasites on cliff swallow growth

and survival, pp. 69-92 in J.E. Loye & M. Zuk, eds. Bird-parasite interactions:

ecology, evolution, and behaviour, Vol. 2. Oxford University Press, Oxford, UK.

Charlesworth, D. & Charlesworth, B. 1987. Inbreeding depression and ist evolutionary

consequences. Annu. Rev. Ecol. Syst. 18:237-268.

Charnov, E.L. 1982. The theory of sex allocation. Princeton University Press, Princeton,

NJ, USA.

Christe, P., M0ller, A.P. & de Lope, F. 1998. Immunocompetence and nestling survival in

the house martin: the tasty chick hypothesis. Oikos 83:175-179.

Clayton, D.H. & Moore, J., eds. 1997. Host-parasite evolution: general principles and avian

models. Oxford University Press, Oxford, UK.

Coltman, D.W., Pilkington, J.G., Smith, J.A. & Pemberton, J.M. 1999. Parasite-mediated

selection against inbred Soay sheep in a free-living, island population. Evolution

53:1259-1267.

Cook, J.M. 1993. Sex determination in the Hymenoptera: a review of models and evidence.

Heredity 71:421-435.

Cook, J.M. & Crozier, R.H. 1995. Sex determination and population biology in the

Hymenoptera. Trends Ecol. Evol. 10:281-286.

Cornuet, J.-M. 1986. Population genetics, pp. 235-262 in T.E. Rinderer, ed. Bee genetics

and breeding. Academic Press, Orlando, FL, USA.

Crozier, R.H. 1975. Insecta 7. Hymenoptera. Gebrüder Borntraeger, Berlin, Germany.

—. 1977. Evolutionary genetics of the Hymenoptera. Annu. Rev. Entomol. 22, 263-288.

142

—. 1985. Adaptive consequences of male-haploidy. pp. 201-222 in W. Helle & M.W.

Sabelis, eds. Spider mites. Their biology, natural enemies and control. Elsevier

Science Publishers B. V., Amsterdam, The Netherlands.

Crozier, R.H. & Pamilo, P. 1996. Evolution of social insect colonies: sex allocation and kin

selection. Oxford University Press, Oxford, UK.

Cushman, J.H., Boggs, C.L., Weiss, S.B., Murphy, D.D., Harvey, A.W. & R., E.P. 1994.

Estimating female reproductive success of a threatened butterfly: influence of

emergence time and hostplant phenology. Oecologia 99:194-200.

da Silva, E.R. & Pereira, S.M. 1993. Mayflies of Serra dos Orgaos, Rio de Janeiro State,

Brazil: I. Description of a gynandromorph of the genus Baetis Leach, 1815

(Ephemeroptera: Baetidae). Anais Acad. Brasil. Ciencias 65:77-79.

de Lope, F., M0ller, A.P. & de la Cruz, C. 1998. Parasitism, immune response and

reproductive success in the house martin Delichon urbica. Oecologia 114:188-193.

de Tomaso A.W., Saito, Y., Ishizuka, K.J., Palmeri, K.J. & Weissman, I.L. 1998. Mapping

the genome of a model protochordate. I. A low resolution genetic map encompassing

the fusion/histocompatibility (Fu/HC) locus of Botryllus schlössen. Genetics

149:277-287.

del Castillo, R.F. 1998. Fitness consequences of maternal and nonmatemal components of

inbreeding in the gynodioecious Phacelia dubia. Evolution 52:44-60.

Delaplane, K.S. & Mayer, D.F. 2000. Crop pollination by bees. CABI Publishing, Oxon,

UK.

Deuve, T. 1992. Segmentary origin of male and female ectodermic genitalia in insects:

New data from a gynandromorph of Coleoptera. Compt. Rend. Acad. Sei. Ser. Ill Sei.

Vie 314:305-308.

143

Donovan, B.J. & Weir, S.S. 1978. Development of hives for field population increase, and

studies on the life cycle of four species of introduced bumble bees in New Zealand.

N. Z. J. Agric. Res. 21:733-756.

Doums, C. & Schmid-Hempel, P. 2000. Immunocompetence in workers of a social insect,

Bombus terrestris L., in relation to foraging activity and parasitic infection. Can. J.

Zool. 78:1060-1066.

Doums, C, Moret, Y., Benelli, E.F. & Schmid-Hempel, P. Senescence of immune defence

in Bombus workers. Ecol. Entomol. In press.

Duchateau, MJ. & Velthuis, H.H.W. 1988. Development and reproductive strategies in

Bombus terrestris colonies. Behaviour 107:186-207.

Duchateau, M.J., Hoshiba, H. & Velthuis, H.H.W. 1994. Diploid males in the bumble bee

Bombus terrestris: sex determination, sex alleles and viability. Entomol. Exp. Appl.

71:263-269.

Duchateau, MJ. & Marien, J. 1995. Sexual biology of haploid and diploid males in the

bumble bee Bombus terrestris. Insect. Soc. 42:255-266.

Duchateau, MJ. 2000. Biological aspects of rearing bumble bees for pollination, pp. 25-29

in MJ. Sommeijer & A. de Ruijter, eds. Insect pollination in greenhouses:

proceedings of the specialists' meeting held in Soesterberg, The Netherlands, 30

September to 2 October 1999. University of Utrecht, The Netherlands.

Dudash, M.R. 1990. Relative fitness of selfed and outcrossed progeny in a self-compatible,

protandrous species, Sabatia angularis L. (Gentianaceae): a comparison in three

environments. Evolution 44:1129-1139.

Dürrer, S. & Schmid-Hempel, P. 1995. Parasites and the regional distribution of bumble

bee species. Ecography 18:114-122.

144

Duvoisin, N. 1998. Mating behaviour in the bumblebee Bombus terrestris L.

(Hymenoptera: Apidae): Mate choice, sperm transfer and plugging. Diploma thesis.

Swiss Federal Institute of Technology (ETH) Zürich, Switzerland.

Duvoisin, N., Baer, B. & Schmid-Hempel, P. 1999. Sperm transfer and male competition in

a bumblebee. Anim. Behav. 58:743-749.

Dzierzon, J. 1845. Gutachten über die von Herrn Direktor Stöhr im ersten und zweiten

Kapitel des General-Gutachtens aufgestellten Fragen. Eichstädter Bienenzeit. 1:109-

113,1:119-121.

Ellstrand, N.C. & Elam, D.R. 1993. Population genetic consequences of small population

size: implications for plant conservation. Annu. Rev. Ecol. Syst. 24:217-242.

Erhardt, G.M. 1996. Blütenökologische Untersuchungen an Aconitum lycoctonum

(Ranunculaceae) und Bombus gerstaeckeri (Hymenoptera, Apidae). Diploma thesis.

University of Zürich, Switzerland.

Estoup, A., Scholl, A., Pouvreau, A. & Solignac, M. 1995. Monandry and polyandry in

bumble bees (Hymenoptera; Bombinae) as evidenced by highly variable

microsatellites. Mol. Ecol. 4:89-93.

Estoup, A., Solignac, M., Cornuet, J.M., Goudet, J. & Scholl, A. 1996. Genetic

differentiation of continental and island populations of Bombus terrestris

(Hymenoptera: Apidae) in Europe. Mol. Ecol. 5:19-31.

Falconer, D.S. 1989. Introduction to quantative genetics, 3rd edn. Longman Scientific &

Technical, Harlow, UK.

Fair, S.C. 1889. Introduction of the humble bee into New Zealand. N. Z. Country J. 13:284-

287.

Foster, R.L. 1992. Nestmate recognition as an inbreeding avoidance mechanism in bumble

bees (Hymenoptera: Apidae). J. Kansas Entomol. Soc. 65:238-243.

145

Frank, S.A. 2000. Specific and non-specific defense against parasitic attack. J. Theor. Biol.

202:283-304.

Frankham, R. 1995. Conservation genetics. Annu. Rev. Genet. 29:305-27.

Frankham, R. & Ralls, K. 1998. Inbreeding leads to extinction. Nature 392:441-442.

Free, J.B. & Butler, C.G. 1959. Bumblebees. Collins, London, UK.

Gadau J., Page, R.E., Jr. & Werren, J.H. 1999. Mapping of hybrid incompatibility loci in

Nasonia. Genetics 153:1731-1741.

Garöfalo, C.A. 1973. Occurence of diploid drones in a neotropical bumblebee. Experientia

29:726-727.

Gaston, K.J. & Lawton, J.H. 1989. Insect herbivores on bracken do not support the core-

satellite hypothesis. Am. Nat. 134:761-777.

Gilpin, M.E. & Soulé, M.E. 1986. Minimum viable populations: processes of species

extinction, pp. 19-34 in M.E. Soulé, ed. Conservation biology: the science of scarcity

and diversity. Sinauer Associates Inc., Sunderland, MA, USA.

Goh, K.T., Ooi, P.L., Miyamura, K., Ogino, T. & Yamazaki, S. 1990. Acute hemorrhagic

conjunctivitis: seroepidemiology of coxsackievirus A24 variant and enterovirus 70 in

Singapore. J. Med. Virol. 31:245-247.

Gordon, D.M. 1991. Behavioral flexibility and the foraging ecology of seed-eating ants.

Am. Nat. 138:379-411.

Götz, P. 1986. Encapsulation in arthropods, pp. 153-170 in M. Brehélin, ed. Immunity in

invertebrates: cells, molecules, and defense reactions. Springer-Verlag, Berlin,

Germany.

Graur, D. 1985. Gene diversity in Hymenoptera. Evolution 39:190-199.

Gupta, A.P. 1986. Arthropod immunocytes: identification, structure, functions, and

analogies to the functions of vertebrate B- and T-lymphocytes. pp. 3-59 in A.P.

146

Gupta, ed. Hemocytic and humoral immunity in arthropods. John Wiley & Sons, New

York, NY, USA.

Gupta, A.P., ed, 1986. Hemocytic and humoral immunity in arthropods. John Wiley &

Sons, New York, NY, USA.

Hadi, U.K., Aoki, C, Saito, K. & Kanayama, A. 1994. Sexual mosaics in two blackfly

species (Diptera: Simuliidae) collected from Shizuoka Prefecture, Japan. Jpn. J. Sanit.

Zool. 45:297-299.

Hanski, I. 1982. Communities of bumblebees: testing the core-satellite species hypothesis.

Ann. Zool. Fennici 19:65-73.

—. 1982. Dynamics of regional distribution: the core and satellite species hypothesis. Oikos

38:210-221.

—. 1982. Structure in bumblebee communities. Ann. Zool. Fennici 19: 319-326.

Hanski, I., Kouki, J. & Halkka, A. 1993. Three explanations of the positive relationship

between distribution and abundance of species, pp. 108-116 in R.E. Ricklefs & D.

Schlüter, eds. Species diversity in ecological communities: historical and

geographical perspectives. University of Chicago Press, Chicago, IL, USA.

Harder, L.D. 1986. Influences on the density and dispersion of bumble bee nests

(Hymenoptera: Apidae). Holarctic Ecol. 9:99-103.

Hedrick, P.W. & Kalinowski, S.T. 2000. Inbreeding depression in conservation biology.

Annu. Rev. Ecol. Syst. 31:139-162.

Hochberg, Y. 1988. A sharper Bonferroni procedure for multiple tests of significance.

Biometrika 75:800-802.

Hoffmann, J.A. & Reichhart, J.M. 1997. Drosophila immunity. Trends Cell Biol. 7:309-

316.

147

Hoffmann, J.A., Kafatos, F.C., Janeway, CA., Jr. & Ezekowitz, R.A.B. 1999. Phylogenetic

perspectives in innate immunity. Science 284:1313-1318.

Holloway, A.K., Strand, M.R., Black, W.C, IV, & Antolin, M.F. 2000. Linkage Analysis

of sex determination in Bracon sp. near hebetor (Hymenoptera: Braconidae).

Genetics 154:205-212.

Holm, S.N. 1972. Weight and life length of hibernating bumble bee queens (Hymenoptera:

Bombidae) under controlled conditions. Entomol. Scand. 3:313-320.

Holsinger, K.E. 1988. Inbreeding depression doesn't matter: the genetic basis of mating-

system evolution. Evolution 42:1235-1244.

—. 1991. Inbreeding depression and the evolution of plant mating systems. Trends Ecol.

Evol. 6:307-308.

Hoop, M. 1964. Ein Gynander von Ancistrocerus ichneumonideus Ratzeburg. Sehr.

Naturw. Ver. Schlesw.-Holst. 35:28-32.

Horber, E. 1961. Beitrag zur Domestikation der Hummeln. Vierteljahresschr. Naturf. Ges.

Zürich 106:424-447.

Hoshiba, H., Duchateau, M.J. & Velthuis, H.H.W. 1995. Diploid males in the bumble bee

Bombus terrestris (Hymenoptera): karyotype analyses of diploid females, diploid

males and haploid males. Jpn. J. Ent. 63:203-207.

Hovorka, O., Urbanovâ, K. & Valterovâ, I. 1998. Premating behavior of Bombus confusus

males and analysis of their labial gland secretion. J. Chem. Ecol. 24:183-193.

Hunt, G.J. & Page, R.E., Jr. 1994. Linkage analysis of sex determination in the honey bee

(Apis mellifera). Mol. Gen. Genet. 244:512-518.

—. 1995. Linkage map of the honey bee, Apis mellifera, based on RAPD markers. Genetics

139:1371-1382.

148

Imhoof, B. & Schmid-Hempel, P. 1999. Colony success of the bumble bee, Bombus

terrestris, in relation to infections by two protozoan parasites, Crithidia bombi and

Nosema bombi. Insect. Soc. 46:233-238.

Inouye, D.W. 1977. Species structure of bumblebee communities in North America and

Europe. Pp. 35-40 in W.J. Mattson, ed. The Role of Arthropods in Forest Ecosystems.

Springer, New York, NY, USA.

Jarne, P., Perdieu, M.A., Pernot, A.F., Delay, B. & David, P. 2000. The influence of self-

fertilization and grouping on fitness attributes in the freshwater snail Physa acuta:

population and individual inbreeding depression. J. Evol. Biol. 13:645-655.

Johnsen, A., Anderson, V., Sunding, C. & Lifjeld, J.T. 2000. Female bluethroats enhance

offspring immunocompetence through extra-pair copulations. Nature 406:296-299.

Johnston, A.B. & Wilson, E.O. 1985. Correlates of variation in the major/minor ratio of the

ant Pheidole dentata (Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 78:8-11.

Johnston, M.O. & Schoen, D.J. 1994. On the measurement of inbreeding depression.

Evolution 48:1735-1741.

Jordan, J.R. & Brosemer, R.W. 1974. Characterization of DNA from three different bee

species. J. Insect. Physiol. 20:2513-2520.

Josephrajkumar, A., Subrahmanyam, B. & Ramamurthy, V.V. 1998. Gynandromorph of

Helicoverpa armigera (Lepidoptera: Noctuidae). Entomol. News 109:288-292.

Kainer, K.A., de Matos Malavasi, M., Duryea, M.L. & Rodrigues da Silva, E. 1999. Brazil

nut {Bertholletia excelsa) seed characteristics, preimbibition and germination. Seed

Sei. Technol. 27:731-745.

Karp, R.D. 1990. Cell-mediated immunity in invertebrates. Bioscience 40:732-737.

Kearns, C.A., Inouye, D.W. & Waser, N.M. 1998. Endangered mutualisms: the

conservation of plant-pollinator interactions. Annu. Rev. Ecol. Syst. 29:83-112.

149

Kinomura, K. & Yamauchi, K. 1994. Frequent occurrence of gynandromorphs in the

natural population of the ant Vollenhovia emeryi (Hymenoptera: Formicidae). Insect.

Soc. 41:273-278.

Kittelson, P.M. & Maron, J.L. 2000. Outcrossing rate and inbreeding depression in the

perennial yellow bush lupine, Lupinus arboreus (Fabaceae). Am. J. Bot. 87:652-660.

Klein, S.L. & Nelson, R.J. 1999. Social interactions unmask sex differences in humoral

immunity in voles. Anim. Behav. 57:603-610.

Koelewijn, H.P. 1998. Effects of different levels of inbreeding on progeny fitness in

Plantago coronopus. Evolution 52:692-702.

König, C. & Schmid-Hempel, P. 1995. Foraging activity and immunocompetence in

workers of the bumble bee, Bombus terrestris L. Proc. R. Soc. Lond. B 260:225-227.

Korpimäki, E. & Wiehn, J. 1998. Clutch size of kestrels: Seasonal decline and experimental

evidence for food limitation under fluctuating food conditions. Oikos 83:259-272.

Kosambi D., 1944. The estimation of map distances from recombination values. Ann.

Eugen. 12:172-175.

Kukovetz, E.M., Bratschitsch, G., Hofer, H.P., Egger, G. & Schaur, R.J. 1997. Influence of

age on the release of reactive oxygen species by phagocytes as measured by a whole

blood chemiluminescence assay. Free Rad. Biol. Med. 22:433-438.

Kulincevic, J.M. 1986. Breeding accomplishment with honey bees. pp. 391-414 in T.E.

Rinderer, ed. Bee genetics and breeding. Academic Press, Orlando, FL, USA.

Lackie, A.M. 1988. Haemocyte behaviour. Adv. Insect Physiol. 21:85-178.

—. 1988. Immune mechanisms in insects. Parasitol. Today 4:98-105.

Lacy, R.C. 1997. Importance of genetic variation to the viability of mammalian

populations. J. Mammal. 78:320-335.

150

Lande, R. & Schemske, D.W. 1985. The evolution of self-fertilization and inbreeding

depression in plants: I. Genetic models. Evolution 39:24-40.

Lander E.S. & Green, J. 1987. Construction of multilocus genetic linkage maps in human.

Proc. Natl. Acad. Sei. USA. 84:2363-2367.

Lander E.S., Green, P., Abrahamson, J., Barlow, A., Daly, M.J. et al. 1987. MAPMAKER:

an interactive computer package for constructing primary genetic linkage maps of

experimental and natural populations. Genomics 1:174-181.

Landolt, E. 1984. Unsere Alpenflora. 5th edn.Verlag Schweizer Alpen-Club, Wallisellen,

Switzerland.

Laurent, V., Wajnberg, E., Mangin, B., Schiex, T., Gaspin, C et al. 1998. A composite

genetic map of the parasitoid wasp Trichogramma brassicae based on RAPD

markers. Genetics 150: 275-282.

Lehmann, T. 1993. Ectoparasites: direct impact on host fitness. Parasitai. Today 9:8-13.

Lively, CM., Craddock, C. & Vrijenhoek, R.C. 1990. Red queen hypothesis supported by

parasitism in sexual and clonal fish. Nature 344:864-866.

Macfarlane, R.P. & Gurr, L. 1995. Distribution of bumble bees in New Zealand. N. Z.

Entomol. 18:29-36.

Martin, T.E. 1987. Food as a limit on breeding birds: a life-history perspective. Annu. Rev.

Ecol. Syst. 18:453-487.

McCurdy, D.G., Shutler, D., Mullie, A. & Forbes, M.R. 1998. Sex-biased parasitism of

avian hosts: relations to blood parasite taxon and mating system. Oikos 82:303-312.

Michener, CD. 1964. Reproductive efficiency in relation to colony size in hymenopterous

societies. Insect. Soc. 11:317-342.

151

Miller, D.R. & Williams, D.J. 1995. Systematic revision of the family Micrococcidae

(Homoptera: Coccoidea), with a discussion of its relationships, and a description of a

gynandromorph. Boll. Laborat. Entomol. Agrar. Filippo Silvestri 50:199-247.

Milne, C.P., Jr. 1985. Estimating primordial cell numbers giving rise to honeybee adult

structures. J. Apic. Res. 24:7-12.

—. 1986. Cytology and Cytogenetics, pp. 205-233 in T.E. Rinderer, ed. Bee genetics and

breeding. Academic Press, Orlando, FL, USA.

Mitton, J.B. 1995. Enzyme heterozygosity and developmental stability. Acta Theriol.

40:33-54.

M0ller, A.P. 1996. Parasitism and developmental instability of hosts: a review. Oikos

77:189-196.

M0ller, A.P., Sorci, G. & Erritz0e, J. 1998. Sexual dimorphism in immune defense. Am.

Nat. 152:605-619.

Moritz, R.F.A. 1985. Inbreeding effects in drones (Apis mellifera L.). Insect. Soc. 32:104-

105.

Morse, D.H. 1982. Behavior and ecology of bumble bees. pp. 245-322 in H.R. Hermann,

ed. Social Insects. Academic Press, New York, NY, USA.

Motro, U. 1991. Avoiding inbreeding and sibling competition: The evolution of sexual

dimorphism for dispersal. Am. Nat. 137:108-115.

Müller, C.B. & Schmid-Hempel, P. 1992. Correlates of reproductive success among field

colonies of Bombus lucorum: the importance of growth and parasites. Ecol. Entomol.

17:343-353.

—. 1992. Variation in life-history pattern in relation to worker mortality in the bumble-bee,

Bombus lucorum. Funct. Ecol. 6:48-56.

152

Mutikainen, P. & Delph, L.F. 1998. Inbreeding depression in gynodioecious Lobelia

siphilitica: among-family differences override between-morph differences. Evolution

52:1572-1582.

Nilsson, G.E. 1987. A gynandromorphic specimen of Evylaeus albipes (Fabricius)

(Hymenoptera, Halictidae) and a discussion of possible causes of gynandromorphism

in haplodiploid insects. Notulae Entomol. 67:157-162.

Norusis, M.J. 1994. SPSS Advanced statistics 6.1. SPSS, Chicago, IL, USA.

Obeso, J.R. 1992. Geographic distribution and community structure of bumblebees in the

northern Iberian Peninsula. Oecologia 89:244-252.

O'Brien, S.J. & Evermann, J.F. 1988. Interactive influence of infectious disease and genetic

diversity in natural populations. Trends Ecol. Evol. 3:254-259.

Ochoa, G. & Jaffe, K. 1999. On sex, mate selection and the Red Queen. J. Theor. Biol.

199:1-9.

Ohgushi, T. 1991. Lifetime fitness and evolution of reproductive pattern in the herbivorous

lady beetle. Ecology 72:2110-2122.

Oldroyd, B., Rinderer, T. & Buco, S. 1991. Heritability of morphological characteristics

used to distinguish European and Africanized honeybees. Theor. Appl. Genet.

82:499-504.

Ott, J. 1999. Analysis of human genetic linkage. 3rd edn. The Johns Hopkins University

Press, Baltimore, MD, USA.

Otto, F.J. 1994. High resolution analysis of nuclear DNA employing the fluorochrome

DAPI. pp. 211-217 in Z. Darzynkiewicz, J. P. Robinson & H.A. Crissman, eds.

Methods in cell biology. Academic Press, San Diego, CA, USA.

Otto, S.P. & Michalakis, Y. 1998. The evolution of recombination in changing

environments. Trends Ecol. Evol. 13:145-151.

153

Ouborg, N.J., Bière, A. & Mudde, CL. 2000. Inbreeding effects on resistance and

transmission-related traits in the Silene-Microbotryum pathosystem. Ecology 81:520-

531.

Owen, R.E., Rodd, F.H. & Plowright, R.C. 1980. Sex ratios in bumble bee colonies:

complications due to orphaning? Behav. Ecol. Sociobiol. 7:287-291.

Owen, R.E. & Plowright, R.C. 1982. Worker-queen conflict and male parentage in bumble

bees. Behav. Ecol. Sociobiol. 11:91-99.

Owen, R.E. 1988. Body size variation and optimal body size of bumble bee queens

(Hymenoptera: Apidae). Can. Entomol. 120:19-28.

—. 1989. Differential size variation of male and female bumblebees. J. Heredity 80:39-43.

Owen, R.E. & Harder, L.D. 1995. Heritable allometric variation in bumble bees:

Opportunities for colony-level selection of foraging ability. J. Evol. Biol. 8:725-738.

Page, R.E., Jr. 1980. The evolution of multiple mating behavior by honey bee queens (Apis

mellifera L.). Genetics 96:263-274.

Page, R.E., Jr., Robinson, G.E. & Fondrk, M.K. 1989. Genetic specialists, kin recognition

and nepotism in honey-bee colonies. Nature 338:576-579.

Page, R.E., Jr., Metcalf, R.A., Metcalf, R.L., Erickson, E.H., Jr. & Lampman, R.L. 1991.

Extractable hydrocarbons and kin recognition in honeybee {Apis mellifera L.). J.

Chem. Ecol. 17:745-756.

Page, R.E., Jr. & Robinson, G.E. 1991. The genetics of division of labor in honey bee

colonies. Adv. Insect. Physiol. 23:117-169.

Pamilo, P. & Crozier, R.H. 1981. Genie variation in male haploids under deterministic

selection. Genetics 98:199-214.

Pamilo, P. & Rosengren, R. 1984. Evolution of nesting strategies of ants: Genetic evidence

from different population types of Formica ants. Biol. J. Linn. Soc. 21:331-348.

154

Pamilo, P. & Crozier, R.H. 1997. Population biology of social insect conservation. Mem.

Museum Victoria 56:411-419.

Papazian, M. 1997. A morphological anomaly with gynandromorphic aspect in Gynacantha

kirbyi Kruger, 1898 (Odonata, Aeshnidae). Bull. Soc. Entomol. France 102:103-109.

Pathak, J.P.N. 1993. Cell-mediated defence reactions in insects, pp. 47-58 in J.P.N. Pathak,

ed. Insect immunity, Vol. 48 Kluwer Academic Publishers, Dordrecht, The

Netherlands.

Pekkarinen, A. & Teräs, I. 1993. Zoogeography of Bombus and Psithyrus in northwestern

Europe (Hymenoptera, Apidae). Ann. Zool. Fennici 30:187-208.

Penn, D.J. & Potts, W.K. 1999. The evolution of mating preferences and major

histocompatibility complex genes. Am. Nat. 153:145-164.

Peters, A.D. & Lively, CM. 1999. The Red Queen and fluctuating epistasis: a population

genetic analysis of antagonistic coevolution. Am. Nat. 154:393-405.

Plass, F. 1953. Inzuchtwirkung und Heterosiseffekt bei der Honigbiene. Landwirtschaft!.

Informationsdienst 66:49-68.

Plowright, R.C. & Pallett, M.J. 1979. Worker-male conflict and inbreeding in bumble bees

(Hymenoptera: Apidae). Can. Entomol. 111:289-294.

Pomeroy, N. & Plowright, R.C. 1980. Maintenance of bumble bee colonies in observation

hives (Hymenoptera: Apidae). Can. Entomol. 112:321-326.

Porter, S.D. & Tschinkel, W.R. 1985. Fire ant polymorphism: the ergonomics of brood

production. Behav. Ecol. Sociobiol. 16:323-336.

Queller, D.C., Negron Sotomayor, J.A., Strassmann, J.E. & Hughes, CR. 1993. Queen

number and genetic relatedness in a neotropical wasp, Polybia occidentalis. Behav.

Ecol. 4:7-13.

155

Ranta, E. & Vepsäläinen, K. 1981. Why are there so many species? Spatio-temporal

heterogeneity and northern bumblebee communities. Oikos 36:28-34.

Rasmont, P., Durieux, E.-A., Iserbyt, S. & Baracetti, M. 2000. Why are there so many

bumblebee species in Eyne (France, Pyrénées-Orientales, Cerdagne)? pp. 83-92 in

M.J. Sommeijer & A. de Ruijter, eds. Insect pollination in greenhouses. University of

Utrecht, The Netherlands.

Rasmont, P., Verhaeghe, J.C., Rasmont, R. & Terzo, M. 2000. West-Palaearctic

bumblebees, pp. 93-99 in M.J. Sommeijer & A. de Ruijter, eds. Insect pollination in

greenhouses: proceedings of the specialists' meeting held in Soesterberg, The

Netherlands, 30 September to 2 October 1999. University of Utrecht, The

Netherlands.

Ratnieks, F.L.W. 1990. The evolution of polyandry by queens in social Hymenoptera: the

significance of the timing of removal of diploid males. Behav. Ecol. Sociobiol.

26:343-348.

Reinig, W.F. 1981. Synopsis der in Europa nachgewiesenen Hummel- und

Schmarotzerhummelarten (Hymenoptera, Bombidae). Spixiana 4:159-164.

Ribeiro, M.F. 1994. Growth in bumble bee larvae: relation between development time,

mass, and amount of pollen ingested. Can. J. Zool. 72:1978-1985.

Ribeiro, M.F., Velthuis, H.H.W., Duchateau, M.J. & van der Tweel, I. 1999. Feeding

frequency and caste differentiation in Bombus terrestris larvae. Insect. Soc. 46:306-

314.

Ritsema, C. 1881. Verslag van de veertiende Wintervergadering der Nederlandsche

Entomologische Vereeniging, gehouden te Leiden op 19 December 1880. Tijdschr.

Entomol. 24:CXI.

156

Roach, D.A. & Wulff, R.D. 1987. Maternal effects in plants. Annu. Rev. Ecol. Syst.

18:209-235.

Roberts, C.W., Cruickshank, S.M. & Alexander, J. 1995. Sex-determined resistance to

Toxoplasma gondii is associated with temporal differences in cytokine production.

Infect. Immun. 63:2549-2555.

Rodd, F.H., Plowright, R.C. & Owen, R.E. 1980. Mortality rates of adult bumble bee

workers (Hymenoptera, Apidae). Can. J. Zool. 58:1718-1721.

Rodriganez, J., Toro, M.A., Rodriguez, M.C. & Silio, L. 1998. Effect of founder allele

survival and inbreeding depression on litter size in a closed line of Large White pigs.

Anim. Sei. 67:573-582.

Roff, D.A. 1997. Evolutionary quantitative genetics. Chapman & Hall, New York, NY,

USA.

Röseler, P.-F. 1962. Über einen Fall von Gynandromorphismus bei der Hummel Bombus

agrorum Fabr. Mitt. bad. Landesver. Naturkunde u. Naturschutz. N. F. 8:289-303.

Ross, K.G. & Fletcher, DJ.C. 1986. Diploid male production: A significant colony

mortality factor in the fire ant Solenopsis invicta (Hymenoptera: Formicidae). Behav.

Ecol. Sociobiol. 19:283-292.

Rothenbuhler, W.C. 1958. Progress and problems in the analyses of gynandromorphic

honey bees. Proc. Tenth Int. Congr. Entomol. 2:867-873.

Rothenbuhler, W.C, Kulincevic, J.M. & Kerr, W.E. 1968. Bee genetics. Annu. Rev. Genet.

2:413-438.

Saccheri, I., Kuussaari, M., Kankare, M., Vikman, P., Fortelius, W. & Hanski, I. 1998.

Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494.

157

Saino, N., Canova, L., Fasola, M. & Martinelli, R. 2000. Reproduction and population

density affect humoral immunity in bank voles under field experimental conditions.

Oecologia 124:358-366.

Saito, Y., Sahara, K. & Mori, K. 2000. Inbreeding depression by recessive deleterious

genes affecting female fecundity of a haplo-diploid mite. J. Evol. Biol. 13:668-678.

Sankilampi, U., Isoaho, R., Blouigu, A., Kivela, S.L. & Leinonen, M. 1997. Effect of age,

sex and smoking habits on pneumococcal antibodies in an elderly population. Int. J.

Epidemiol. 26:420-427.

SAS Institute Inc. 1989. SAS/STAT® user's guide, version 6, 4th edn. SAS Institute Inc.,

Cary, NC, USA.

Sauter, A., Brown, M.J.F., Baer, B. & Schmid-Hempel, P. 2001. Males may prevent the

evolution of super-maters in social insects. Proc. R. Soc. Lond. B, in press.

Schaal, B.A. 1984. Life-history variation, natural selection, and maternal effects in plant

populations, pp. 188-206 in R. Dirzo & J. Sarukhan, eds. Perspectives on plant

population ecology. Sinauer Associates, Sunderland, MA, USA.

Schmid-Hempel, P. & Heeb, D. 1991. Worker mortality and colony development in

bumblebees, Bombus lucorum (L.) (Hymenoptera, Apidae). Mitt. Schweiz. Entomol.

Ges. 64:93-108.

Schmid-Hempel, P. & Koella, J.C. 1995. Ecology and genetics of virulence and resistance.

pp. 589-600 in W.A. Nierenberg, ed. Encyclopedia of environmental biology, Vol. 1.

Academic Press, San Diego, CA, USA.

Schmid-Hempel, P. 1998. Parasites in social insects. Princeton University Press, Princeton,

NJ, USA.

Schmid-Hempel, P. & Loosli, R. 1998. A contribution to the knowledge of Nosema

infections in bumble bees, Bombus spp. Apidologie 29:525-535.

158

Schmid-Hempel, P., Puhr, K., Krüger, N., Reber, C. & Schmid-Hempel, R. 1999. Dynamic

and genetic consequences of variation in horizontal transmission for a microparasitic

infection. Evolution 53:426-434.

Schmid-Hempel, R. & Schmid-Hempel, P. 1996. Larval development of two parasitic flies

(Conopidae) in the common host Bombuspascuorum. Ecol. Entomol. 21:63-70.

—. 1998. Colony performance and immunocompetence of a social insect, Bombus

terrestris, in poor and variable environments. Funct. Ecol. 12:22-30.

—. 2000. Female mating frequencies in social insects: Bombus spp. from Central Europe.

Insect. Soc. 47:36-41.

Schultz, E.T. 1993. The effect of birth date on fitness of female dwarf perch, Micrometrus

minimus (Perciformes: Embiotocidae). Evolution 47:520-539.

Schultz, S.T. & Willis, J.H. 1995. Individual variation in inbreeding depression: the roles of

inbreeding history and mutation. Genetics 141:1209-1223.

Semmens, T.D., Turner, E. & Buttermore, R. 1993. Bombus terrestris (L.) (Hymenoptera:

Apidae) now established in Tasmania. J. Aust. Entomol. Soc. 32:346.

Shykoff, J.A. & Müller, C.B. 1995. Reproductive decisions in bumble-bee colonies: The

influence of worker mortality in Bombus terrestris (Hymenoptera, Apidae). Funct.

Ecol. 9:106-112.

Sichel, J. 1858. Note sur un cas d'hermaphrodisme observé chez un Bombus lapidarius.

Ann. Soc. Entomol. France 6:CCXLVII-CCXLIX.

Sladen, F.W.L. 1989. The humble-bee. Its life history and how to domesticate it. Logaston

Press, Little Logaston, UK.

Smithers, C.N. 1996. A gynandromorph of Ectopsocus australis Schmidt and Thornton

(Psocoptera: Ectopsocidae) from Australia. Aust. Entomol. 23:93-95.

159

Sommeijer, M.J. & de Ruijter, A. 2000. Insect pollination in greenhouses: proceedings of

the specialists' meeting held in Soesterberg, The Netherlands, 30 September to 2

October 1999. University of Utrecht, The Netherlands.

Soulé, M.E. 1986. Conservation biology: the science of scarcity and diversity. Sinauer

Associates Inc., Sunderland, MA, USA.

Stevens, L., Yan, G. & Pray, L.A. 1997. Consequences of inbreeding on invertebrate host

susceptibility to parasitic infection. Evolution 51:2032-2039.

Stöckhert, F.K. 1921. Über einen Fall von frontaler Gynandromorphie bei Bombus

lapidarius L. (Hym.). Z. wiss. Ins.-biol. 16:132-135.

Stouthamer, R., Luck, R.F. & Werren, J.H. 1992. Genetics of sex determination and the

improvement of biological control using parasitoids. Environ. Entomol. 21:427-435.

Strauss, S.Y. & Karban, R. 1994. The significance of outcrossing in an intimate plant-

herbivore relationship: II. Does outcrossing pose a problem for thrips adapted to the

host-plant clone? Evolution 48:465-476.

Stupf, R. 1992. Vorkommen und Verbreitung von Bombus-Arten (Apidae, Hymenoptera)

und ihrer Parasiten in einer Alpenregion. Diploma thesis. Swiss Federal Institute of

Technology (ETH), Zürich, Switzerland.

Sutcliffe, G.H. & Plowright, R.C. 1988. The effects of food supply on adult size in the

bumble bee Bombus terricola Kirby (Hymenoptera: Apidae). Can. Entomol.

120:1051-1058.

Svensson, B.G. 1979. Patrolling behaviour of bumble bee males (Hymenoptera, Apidae) in

a supalpine/alpine area, Swedish Lapland. Zoon 7:67-94.

Theis, J.H. & Schwab, R.G. 1992. Seasonal prevalence of Taenia taeniaeformis:

relationship to age, sex, reproduction and abundance of an intermediate host

(Peromyscus maniculatus). J. Wildlife Dis. 28:42-50.

160

Thompson, J.N. & Burdon, J.J. 1992. Gene-for-gene coevolution between plants and

parasites. Nature 360:121-125.

Thornhill, N.W. 1993. The natural history of inbreeding and outbreeding: Theoretical and

empirical perspectives. The University of Chicago Press, Chicago, IL, USA.

Utelli, A.-B. & Roy, B.A. 2000. Pollinator abundance and behavior on Aconitum

lycoctonum (Ranunculaceae): An analysis of the quantity and quality components of

pollination. Oikos 89:461-470.

Uyenoyama, M.K., Holsinger, K.E. & Waller, D.M. 1993. Ecological and genetic factors

directing the evolution of self-fertilization. Oxf. Surv. Evol. Biol. 9:327-381.

von Hagen, E. 1994. Hummeln: bestimmen, ansiedeln, vermehren, schützen. Naturbuch

Verlag, Augsburg, Germany.

Wakabayashi, A., Utsuyama, M., Hosoda, T., Sato, K. & Hirokawa, K. 1999. Differential

age effect of oral administration of an antigen on antibody response: an induction of

tolerance in young mice but enhancement of immune response in old mice. Mech.

Ageing Dev. 109:191-201.

Wakelin, D. & Apanius, V. 1997. Immune defence: genetic control, pp. 30-58 in D.H.

Clayton & J. Moore, eds. Host-parasite evolution: general principles and avian

models. Oxford University Press, Oxford, UK.

Webb, M.C. 1961. The biology of the bumblebee of a limited area in eastern Nebraska.

Ph.D. thesis. University of Nebraska, Lincoln, NE, USA.

Weeks, S.C, Crosser, B.R., Bennett, R., Gray, M. & Zucker, N. 2000. Maintenance of

androdioecy in the freshwater shrimp, Eulimnadia texana: estimates of inbreeding

depression in two populations. Evolution 54:878-887.

Whiting, P.W. 1940. Multiple alleles in sex determination of Habrobracon. J. Morph.

66:323-355.

161

—. 1943. Multiple alleles in complementary sex determination of Habrobracon. Genetics

28:365-382.

Widmer, A., Schmid-Hempel, P., Estoup, A. & Scholl, A. 1998. Population genetic

structure and colonization history of Bombus terrestris s.l. (Hymenoptera: Apidae)

from the Canary Islands and Madeira. Heredity 81:563-572.

Widmer, A. & Schmid-Hempel, P. 1999. The population genetic structure of a large

temperate pollinator species, Bombus pascuorum (Scopoli) (Hymenoptera: Apidae).

Mol. Ecol. 8:387-398.

Williams, J.G.K., Kubelik, A.R., Livak, K.K., Rafalski, J.A. & Tingey, S.V. 1990. DNA

polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic

Ac. Res. 18:6531-6335.

Williams, P.H. 1982. The distribution and decline of British bumble bees {Bombus Latr.). J.

Apic. Res. 21:236-245.

—. 1988. Habitat use by bumble bees {Bombus spp.). Ecol. Entomol. 13:223-237.

—. 1989. Why are there so many species of bumble bees at Dungeness? Biol. J. Linn. Soc.

101:31-44.

—. 1998. An annotated checklist of bumble bees with an analysis of patterns of description

(Hymenoptera: Apidae, Bombini). Bull. Nat. Hist. Mus. Lond. (Ent.) 67:79-152.

Wolfner, M.F. 1997. Tokens of love: functions and regulation of Drosophila male

accessory gland products. Insect Biochem. 27:179-192.

Woyke, J. 1963. Drone larvae from fertilized eggs of the honeybee. J. Apic. Res. 2:19-24.

—. 1963. What happens to diploid drones in a honey bee colony? J. Apic. Res. 2:73-76.

Yan, G. 1997. Consequence of larval tapeworm infection for the fitness of the intermediate

hosts, flour beetles {Tribolium spp.). Can. J. Zool. 75:271-279

162

Zar, J.H. 1996. Biostatistical Analysis. 3rd edn. Prentice-Hall, Upper Saddle River, NJ,

USA.

163

Curriculum Vitae

Christine Ute Gerloff-Gasser

born December 23, 1969 in Cheverly, Maryland, USA

Nationalities: American / German

Current Position:

High school teacher for biology in English, Gymnasium am Münsterplatz, Basel,

Switzerland (since 1999)

Education

1988

1988/89

1989-1994

1995

1995-2001

Graduation from high school, Kantonsschule Zug, Switzerland,

specialization in natural sciences (Matura Typus C)

Exchange Student at Chetek High School, Wisconsin, USA

Studies in biology at the Swiss Federal Institute of Technology (ETH)

in Zürich, Switzerland, specialization in ecology and systematics

Diploma thesis: "Der Einfluss von Entbuschungsmassnahmen auf die

Zusammensetzung der Heuschreckenfauna (Orthoptera) im Schaffhauser

Randen bei Merishausen SH" (supervised by Dr. S. Ingrisch)

Diploma in teaching biology at high school level (Didaktischer

Ausweis/Höheres Lehramt), ETH Zürich, Switzerland

Ph.D. thesis in biology, ETH Zürich, Switzerland:

"Distributional Ecology and Inbreeding in Bumble Bees"

(supervised by Prof. Dr. P. Schmid-Hempel)

164

Professional Experience

1989 Research Assistant with the Group of Phytopathology, Swiss Federal

Research Station for Fruit-Growing, Viticulture and Horticulture,

Wädenswil, Switzerland

1992 Research Assistant with "Biotech", Moscow, Russia (Food Sciences)

1994 Research Assistant for invertebrates with the Management Group of the

Grande Cariçaie, Champ-Pittet, Yverdon, Switzerland

1995 Biology Teacher at upper secondary school, Lehrerinnenseminar

Heiligkreuz, Cham, Switzerland

Publications

Gerloff C. & Ingrisch S. 1994: Der Einfluss von Entbuschungsmassnahmen auf die

Zusammensetzung der Heuschreckenfauna (Orthoptera) im Schaffhauser Randen. Mitt.

Schweiz. Entomol. Ges. 67:437-452.

Co-author of: Survival Guide. A Handbook for PhD Students at ETH Zurich. 1998. Prod.

AVETH, The Assistant's Association of ETH, Zürich, Switzerland. 147 pp.

Gadau J., Gerloff C.U., Krüger N., Chan H., Schmid-Hempel P., Wille A. & Page R.E. Jr.

A linkage analysis of sex determination in Bombus terrestris (L.) (Hymenoptera: Apidae).

Heredity (in revision).

Gerloff C.U., Ottmer B. & Schmid-Hempel P. Effects of inbreeding, sex, and ploidy on

offspring quality in a social insect, Bombus terrestris (L.) (Hymenoptera: Apidae). J. Evol.

Biol, (in revision).

Gerloff C.U. & Schmid-Hempel P. Inbreeding depression and family variation in a social

insect, Bombus terrestris (L.) (Hymenoptera: Apidae). Evolution (in revision).

Appendix

(Chapters 2,3, and 4)

II

Table A2.1. Site and sampling characteristics of the first (top line) and the second (bottom

line) visit of 18 study sites in southwestern Switzerland.

Site no. Locality* Mean Altitudinal Collecting Number Duration of

exposition range [m] datesf of

collectors

collection

[min]$

1 Grimentz E 1700-2130 134 2 480

1700-2130 197 2 540

2 Gletsch S 1760-2120 154 2 420

1750-2130 220 2 240

4 Rothwald N 1770-1880 135 2 300

1850-2000 199 1 215

5 Täsch W 1700-2260 136 2 500

1750-2220 198 2 220

6 Mayens de My E 1720-1890 137 2 300

1750-1930 201 1 175

7 Ulrichen S 1700-2100 139 2 540

1700-2020 203 2 400

8 La Forclaz w 1720-2160 141 1 465

1740-2020 207 2 260

9 Beialp s 1730-2100 140 1 300

1770-2200 200 1 575

10 Verbier s 1700-2220 142 1 405

1675-2350 202 2 450

11 Visperterminen w 1700-2210 143 1 310

1870-2190 205 2 330

12 Bourg-St.-Pierre w 1750-1950 145 2 630

1720-1890 209 1 215

13 Saas-Almagell w 1870-1970 146 2 240

1770-1950 206 2 240

14 Bellwald s 1720-2020 147 1 235

1700-2470 212 1 185

15 Ovronnaz E 1700-1800 148 1 190

1780-2000 210 1 210

16 Leukerbad W 1760-2000 149 2 270

1750-2310 211 1 220

17 Mauvoisin N 1700-1900 151 1 255

1720-1860 218 2 160

18 Thyon E 1750-2100 156 2 460

1980-2180 219 2 200

20 Vercorin N 1720-2080 155 2 260

1720-2330 219 2 270

* for site location, see Fig. 2.1, Chapter 2

t days of the year 1997 (Julian days)

$ time spent collecting was doubled if two people collected

Table

A2.2.Abundance(number

ofcollected

individuals;

queens,

workers,andmalessampled

duringtwo

visitspooled)

ofthesubalpineand

alpi

nebumblebee

species

at18

sites

insouthwestern

Switzerland.

Species

Site

no.*

12

45

67

89

10

11

12

13

14

15

16

17

18

20

TOTAL

Bombus

argi

llaceus

11

gerstaeckeri

12

12

6

hortorum

315

82

67

15

451

humilis

312

117

11

510

41

55

hypnorum

12

69

jone

llus

13

12

23

11

11

16

lapida

rius

21

13

32

16

120

lucorum

71

134

11

27

24

16

227

25

25

34

35

56

719

7322

mendax

15

31

20

12

21

118

21

168

mesomelas

10

61

77

227

62

36

87

10

23

1108

monticola

75

10

21

24

719

35

16

17

18

818

27

187

mucidus

14

24

82

21

pascuorum

12

41

13

315

pratorum

12

130

18

66

24

21

329

14

22

17

33

25

12

32

74

379

pyrenaeus

732

12

52

21

25

46

210

43

619

15

5160

ruderarius

88

113

22

59

20

11

11

21

71

16

42

3153

sichelii

88

25

11

15

11

211

41

14

23

70

soroeensis

229

22

71

163

12

17

13

815

933

25

619

13

16

320

subterraneus

11

2

sylvarum

11

2

wurflenii

828

12

30

18

23

20

10

117

23

918

14

26

25

5287

Total

78

129

127

123

127

128

131

123

129

130

120

130

126

126

128

133

130

133

2252

Subgenus

Psithyrust121243

210

43213

38

*for

site

location,see

Fig.

2.1,

Chapter2

fnot

identified

tothespecies

level

Table

A2.3.Abundance(number

ofqueens/workers/males

coll

ected)

ofthesubalpineand

alpine

bumble

bee

species

atthe

first(top

line

)and

thesecond

(bottom

line

)visitto18

sites

insouthwestern

Switzerland

The

dates

ofthe

visitsaregiven

inTableA2

5Forthesubgenus

Psithyrus,

theabundance

isgivenas

thenumber

ofqueens/males

collected

Siteno

•

Species

(Bombus)

argillaceus

t

gars

taac

kenf

hortorum

1/-I-

humilis

hypnorum

lonellus

lapidanus

lucorum

mendax

mesomglas

monticola

mucHfusjkV

pascuorum

pralorum

pyrenaeus

rudemmis

sicheln

M-l-

-/1/1

-/3/-

21-1-

3141-

1/-/-,

•IW-

1/4/-

-121-

-1M-

71-/-

-/5/-

M-l-

-161-

"t4itin

131-

-U"'"%éî

*"-t~

subterraneus

t

wurflenu

3H-

-noi-

-im-

5/-/-

-/1/-

M-l-

-141-

-i-n

61-1-

-/23/3

m

M-/-

26/1/-

-17/-

\t-l-

M-l-

8/-/-

-/20/2

M-l-

-/10/1

A-ft/-

31-1

/I4/1

-/5/

1/-/-

-/11/-

-/1/-

IM-

1/-/-

-/1Q/-

21-1

131-

21-/-

,.

-151-

41-1-

-/15/2

-121-

1/-/-

-/15/lf

-151-

21/2/-

-141-

21-1

-IM-

41-/-

1/2/-

1/1/-

141-

3/-I-

-13/-

'<"A

"VmW-'*,

-/8/-

-I2I-

19/-/-

?/-/-

-/to/-

vi^/'V

14/-/-

71-1-

-/14/-

-/5/-

,/7/

l

•IM-

Sl-t-

-nit

11/-/-

1/18/-

10

-161-

11

-I2I-

5/-/-

1/-/-

,

-,/1

i1/1

^,|<>/1

1/-/-

14/2/-

-/8/2

1/1/-*

1/1/-

-1-12

12/2/-

121-

31-1-

-/i/i

-IM-

Pi*4^ -IM-

-1214

WW

51-1-

-/4/2

-/19/-

-/2/-

-/1/-

-/!/

-'£'

-/3/1

-/8/1

-/1/-

-/1/-

6/-/-

2/-/-

1/53/3»V^iQ/W,

-/3/-

18/-/-

1/7/1

71-1-

-/13/-

1/-/-

/il

l

^J/S/-»

4/6/-

-191-

1>-/4/-

41-1-

1/13/-

-n/-

51-1-

-1U31-

13/-/-

-nn

M-l-

-1241-

ngf

tf

151-

91-/-

Itfi

l-f

6/-/-

-/14/-

2/-/-

-/1/-

-/!/-

1/-/-

-/10/-

6/-/-

-/?/

-

-/1/-

71-1-

-131-

M-l—

-IM-

1/-/-

-/1/-

10/7/-

-1312

-/12/-

-121-

-/15/-

-/20/-

»/7/1

31-1-

131-

-IM-

-121-

61-1-

•121-

M-l-

-/M-

3/1/-

-/19/2

-121-

-m-

1/7/-

-16/2

-12/-

$/-{-

-/8/1

-121-

41-1-

-/4/3

M-l-

-/10/-

1/-/-

-/13/1

5/-/-

-/12/-

32/-/-

-121-

M-l-

151-

-1M-

W-t-

1/9/3

-/10/-

12/-/-

-/a/-

141-

31-1-

2/41-

61-1-

-/15/-

1/-/-

-/2/-

25/3/-

-/4/3

-IM-

1/-/-

»t//?/--

1/2/-

-/3/1

3/1/-

•1914

M-l-

-/2/1

"

1/-/-

-/5/1

1/-/-

17/-/-

-/16/-

6/-/-

-/2/1

15

2/-/-

3/f/

-

2/-/-

-/1/-

2/1/-

-/1/1

M-l-

2/-/-

'p/5/-

3/15/-

11/-

/-',

-I2<31

2;

M-l-

-121-

-i-n

16/-/-

-mi-

9/-/-

-/9/-

16

-/-/1

1/-/-

2/-/-

-/5/-

1/-/-

-/3/-

,\/¥

•

-13/2

71-1-

-nM-

6Î-Â

-/4f-

-/7/1

1/-/-

-13/-

M-l-

•l"\5t\

161-

14/-/-

-12/-

-IM-

31-1-

-/31-

41-1-

-/9/-

21-t-

M-l-

1/-/-

1/3/2

M-l-

1/-/-

-/3/3

-/2t-

-131-

-1614

5/-/-

-/3/-

M-l-

1/-/-

-/fi/5

8/-/-

-/10/1

3J-f-4

-IM-

21-1-

121-

17/-/-

1/1/-

18/-/-

-161-

M-l-

-IM3

11-/-

161-

5/10/-

-1212

IM-

3/-/-

-IM-

-IM-

-131-

%1-t-

-/19/11

1/1/-

-/10/3

tl'-

l-

21-1-

3f-t-

IM-

71-1-

-/13/5

M-l-

-131- *",St

1/-/-

-/

/1

1/2/-

-12/2

-IM-

-IM\

-/4/1

f-/1/1

1/

/-

-121-

38/-/-

-/2S/11

1/-/-

-/3/1

3/-/-

M-l-

-121-

10/-/-,

^-10

A

41-1-

-i-n

Total

-/-/1

©/-/-

31-1-

1/37/5

11/-/-

1/39/4,

I

3/-/-

1/3/2

i«'

-171-

71-1-

-/12/1

175/35/-

1/89/22

19/

/-

-/49/-

33/-/-

2/73/-

22/67/1

-/80/17

6/-/-

-/14/1

41-1-

-/11/-

fcas/t/-

pf/f92/48t

24/1/-

-/125/10

47/1/-

-/94/11

71-1-

1/62/

139/-/-

4/173/4

-121-

>,-m-

115/-/-

2/162/8

1 6

51

I55 9

16

20

322

68

T08

187

21

15

379

160

4*'HS3

70

320 2 2

Total

16/8/-

51/-/-

58/5/-

11/-/-

59/3/-

25/41-

32121-

32/1/-

46/6/-

421221-

23/8/-

65/1/-

57/6/-

50/16/-

491-1-

57/8/-

32/12/-

62/3/1

767/105/1

873

-/52/2

-/73/5

-/61/3

-/106/5t

2/62/1

3/86/10

-/95/2

1/83/6

1/74/2

-/61/5

-/80/9

3/58/3

1/51/11

2/54/4

1/71/7

6/47/15

-/62/24

-/49/18

20/1226/132*

1379

78

129

127

123

127

128

131

123

129

130

120

130

126

126

128

133

130

133

2252

Subgenus

1/-

21-

1/-

21-

3/-

4/-

1/-

1/-

1/-

16/-

38

Psithyrus§

1/1

12

31-

-12

2/5

-13

-/1

12

6/16

*for

site

location,see

Fig

21,

Chapter2

tspecimensfound

only

duri

ng2nd

visit

%plus

onegynandromorph,seeChapter2

§not

identified

tospecies

level

V

Table A2.4. Relative abundance (mean, minimum, and maximum) of the bumble

bee species found in the subalpine and alpine zones in southwestern Switzerland.

Species Relative abundance

mean minimum maximum

Bombus

argillaceus 0.78 0.78 0.78

gerstaeckeri 1.17 0.78 1.59

hortorum 4.63 0.75 12.20

humilis 4.46 0.75 12.98

hypnorum 2.30 0.79 4.51

jonellus 1.26 0.75 2.38

lapidarius 1.84 0.75 4.62

lucorum 14.21 1.63 27.78

mendax 4.08 0.75 15.50

mesomelas 5.24 0.75 20.61

monticola 8.84 0.77 26.92

mucidus 2.78 1.28 6.02

pascuorum 1.66 0.75 3.15

pratorum 16.71 0.78 55.64

pyrenaeus 7.13 0.76 24.81

ruderarius 6.92 0.77 17.32

sichelii 3.48 0.76 10.26

soroeensis 13.98 2.56 49.22

subterraneus 0.77 0.77 0.78

sylvarum 0.79 0.76 0.81

wurflenii 13.42 0.77 23.62

VI

Table A2.5. Species richness S (number of species present) and diversity D

(Simpson's index) of bumble bee communities of 18 sites in southwestern

Switzerland.

Site no.* Species Diversity D t

richness S

1 13 9.97

2 10 5.62

4 12 5.49

5 14 9.44

6 14 6.65

7 10 3.30

8 10 6.58

9 13 6.75

10 15 8.30

1 1 13 5.67

12 12 7.75

13 9 5.94

14 12 5.52

15 13 6.21

16 17 9.46

17 17 8.57

18 15 6.77

20 15 2.98

* for site location, see Fig. 2.1, Chapter 2

t for calculation of the diversity, see Chapter 2, Material and Methods

Table

A3.1.

Experimental

char

acte

rist

ics,

colony

life

-histo

rydata,andbody

size

ofthefoundingqueen,

itsmateand

thesexualsproduced

oftheexperimental

colonies(Bombus

terrestris

).

Colony

Treatment

Family

ine

Date

of

Workers

Haploid

Young

Occur¬

Body

sizeH

(inbred/

maternal

paternal

matingt

malest

queens

rence

founding

founding

haploid

young

outbred)*

dipl

oid

males§

queen

queen's

mate

males

queens

mean±SE

(n)

C116

11

1193

33

01

4.22

C124

01

18

195

82

04.22

C22

12

2170

50

01

4.36

C269

12

2170

65

258

00

C298

02

4170

111

388

33.72

3.22±0.04

(20)

4.18±0.01

(2)

C323

02

6170

18

00

4.28

3.46

C79

02

7170

299

282

94

4.22

3.64

3.41±0.06

(20)

3.75±0.03

(15)

C19

44

169

193

497

10

4.32

(1)

C78

44

169

106

209

01

C134

44

169

09

00

4.06

C169

44

169

332

391

20

C227

44

169

50

259

01

4.16

C290

44

169

65

421

31

C327

44

169

48

312

00

4.32

C251

44

174

75

86

00

C183

04

2174

263

62

13

4.24

3.64

3.30±0.04

(20)

C291

04

17

174

143

506

03.72

3.41±0.03

(20)

C37

04

30

174

90

219

03.56

3.53±0.02

(20)

C364

15

5170

77

90

00

4.20

C83

17

7169

22

97

00

4.36

C188

17

7169

157

336

86

04.36

4.34±0.02

(20)

C268

17

7169

98

417

24

04.24±0.03

(20)

C241

07

2169

20

04.44

3.66

C245

07

4169

10

04.28

3.66

C25

07

5169

20

04.24

3.56

C181

07

12

169

29

26

04.36

3.76

3.59±0.02

(16)

C66**

07

16

169

53

100

1

C370

07

17

169

44

89

04.36

3.56

3.60±0.04

(17)

Table

A3.1.Continued

Colony

Treatment

Family

line

Date

of

Workers

Haploid

Young

Occur¬

Body

sizen

(inbred/

maternal

paternal

matingt

malest

queens

rence

founding

founding

haploid

young

outbred)*

diploid

queen

queen's

males

queens

males§

mate

mean±SE

(n)

C283

18

8183

96

214

73

14.36

4.04±0.04

(20)

C351

08

2183

297

521

316

4.36

3.56

3.53±0.02

(20)

3.88±0.04

(15)

C46

08

5195

816

03.60

3.57±0.06

(16)

C105

08

6190

242

357

180

3.72

3.61±0.02

(20)

3.99±0.02

(12)

C293

08

16

186

117

51

13.76

C140

08

24

186

185

00

4.28

C271

09

5197

120

265

24.32

3.24

3.47±0.04

(20)

3.98±0.19

(2)

C91

09

16

174

20

04.32

3.68

C63

11

183

44

303

00

4.22

C205

11

183

29

405

11

04.24

C216

11

183

65

573

01

4.24

C346

11

183

66

520

01

4.18

C359

11

183

71

304

70

4.24

3.95±0.05

(6)

C373

11

183

51

01

4.16

C375

11

183

57

179

11

04.34±0.02

(11)

C144

04

183

33

234

04.18

3.68

3.27±0.03

(20)

C354

08

183

68

574

14.32

3.68

3.33±0.04

(20)

C3

09

186

10

33

04.12

3.46±0.04

(20)

C60

012

190

132

551

14.24

3.72

3.04±0.07

(20)

C191

014

183

41

49

64.24

3.52

3.53±0.04

(19)

4.29±0.06

(6)

C210

017

184

24

248

04.36

3.64

3.55±0.04

(20)

C53

030

184

62

188

04.20

3.68

3.33±0.04

(20)

C267

012

30

171

20

04.40

3.72

C117

114

14

183

10

0?

4.08

C157

014

8183

136

634

03.56

3.23±0.03

(18)

C189

115

15

193

30

177

01

4.32

C236

115

15

193

70

73

36

03.86±0.05

(16)

C113

015

20

195

85

53

03.48

3.45±0.02

(20)

C223

015

30

193

70

04.32

3.32

Table

A3.1.

Continued

Colony

Treatment

Family

line

Date

of

Workers

Haploid

Young

Occur-

Body

sizeTl

(inbred/

maternal

paternal

matingf

malest

queens

rence

founding

founding

haploid

young

outbred)*

dipl

oid

queen

queen's

males

queens

males§

mate

mean±SE

(n)

C178

116

16

186

14

18

00

4.04

C153

016

8181

260

74

113

3.88

3.60±0.03

(15)

3.90+0.04

(20)

C141

016

11

183

02

04.20

C88

117

17

169

59

51

4.24

4.31±0.02

(5)

C164

117

17

169

12

15

01

C186

117

17

169

32

33

13

14.36

4.10±0.05

(13)

C337

017

2169

145

439

10

4.28

3.76

3.37±0.04

(20)

4.22±0.07

(6)

C302

017

5169

78

88

34.24

3.52

3.35±0.06

(20)

4.16±0.05

(3)

C175

017

9169

77

265

34.40

3.84

3.64±0.03

(20)

4.42±0.08

(2)

C8

017

10

169

101

409

24.32

3.44

3.42±0.04

(20)

4.04±0.08

(2)

C237

017

16

169

85

438

03.76

3.39±0.03

(20)

C138

017

30

169

175

590

83.60

3.32±0.04

(20)

4.28±0.02

(3)

C185

018

25

190

71

106

04.36

C231

119

19

195

93

342

61

4.22±0.05

(6)

C235

119

19

195

169

88

20

C340

119

19

195

75

62

25

14.40

4.16+0.01

(20)

C2

019

12

195

20

36

24.42

3.68

3.58±0.03

(20)

4.30+0.02

(2)

C355

019

13

195

20

04.32

C350

019

15

195

37

81

6

C193

019

30

195

23

04.44

3.36

*

0=

outbred,

1=

inbred

tdays

oftheyear1997

(Jul

iandays)

tforthe

dipl

oidmale

colonies

calculated

as:#males

-

(#workers+#youngqueens)

§0=

no,

1=

yes,

?=

toofew

offs

prin

gtobeassigned

Hle

ngth

of

radial

cell

[mm]

**

dupl

icate

ofamating

combination,

excludedfrom

the

analysis

Table

A3.2.Average

colony

life

-his

tory

data

forBombus

terrestris

ofseven

maternal

fami

lylinesaccording

totreatment

(inb

red/

outb

red)

.

Maternal

Colony

size(SE)*

Caloricinvestment

(SE)

[kJ]f

inbred

outbred

Shaw-Mohler

in

inbred

dex(SE)+

family

line

inbred

outbred

outbred

235.0±30.0

142.7±82.7

303.15±303.15

778.00+411.36

0.011±0.011

0.102±0.083

4108.6±37.6

165.3±51.2

647.42±141.35

650.41±280.17

0.024±0.006

0.033±0.007

792.3±39.1

15.6±8.9

952.93±372.37

54.05±40.45

0.119±0.071

0.003±0.002

11

48.1+9.1

52.9+15.3

799.55±173.50

639.08±192.00

0.037±0.007

0.025±0.007

15

50.0±20.0

46.0±39.0

434.69±18.74

62.28±62.28

0.057±0.043

0.002+0.002

17

16.3±8.1

110.2±16.6

91.63±44.45

906.96±171.35

0.017±0.010

0.042±0.008

19

112.3±28.8

15.3+8.4

471.53±192.66

86.16±55.33

0.042±0.017

0.008±0.005

Mean±SE§

66.1±14.4

78.3±23.1

528.70+111.42

453.85±140.72

0.044±0.014

0.031+0.013

*number

ofworkers

tcalculated

as:(#youngqueens

x7.83

kJ)+

(#haploidmales

x2.35

kJ)

+.forcalculation

oftheShaw-Mohler

fitnessindex,

seeChapter

3,MaterialsandMethods

§basedon

family

averages

Table

A4.1.Averageimmune

responseandbody

size,andsample

sizeofbumblebeeworkersand

hapl

oidmales

of

sixmaternal

fami

lylines

according

to

treatment

(inb

red/

outb

red)

.

Maternal

Immuneresponse(SE)*

Workers

Haploidmales

inbred

outbred

inbred

outbred

Body

size(SE)t

Sample

sizet

family

Workers

Haploidmales

Workers

Haploidmales

line

inbred

outbred

inbred

outbred

inbred

outbred

inbred

outbred

261.79±10.84

51.38±13.96

54.56±4.11

40.09±3.51

2.95±0.02

2.92±0.05

3.41±0.03

3.31±0.06

10

20

918

456.96±6.47

60.71+11.09

46.82±2.44

44.42±4.28

2.83±0.05

2.94±0.05

3.47±0.05

3.59±0.03

20

26

28

26

759.74±9.73

48.57±10.50

41.92±2.21

39.74±2.95

2.92±0.03

2.89±0.05

3.44±0.02

3.39±0.03

60

29

58

29

850.82±14.71

50.40±8.71

43.98±3.34

40.01±2.75

3.00±0.09

2.92±0.04

3.68±0.04

3.57±0.04

10

40

628

11

58.25±8.56

62.32±6.59

46.69±2.55

50.93±2.03

2.69±0.04

2.76±0.04

3.39±0.03

3.36±0.03

47

48

50

58

17

49.39±18.03

57.59±8.65

56.69±4.13

51.39±1.68

2.66±0.15

2.73±0.04

3.63±0.02

3.43±0.03

10

57

14

58

Mean±SE§

56.16±2.03

55.16±2.37

48.44±2.41

44.43±2.24

2.84±0.06

2.86±0.04

3.50±0.05

3.44±0.05

26.2±9.0

36.7±5.8

27.5±9.0

36.2±7.1

Total

157

220

165

217

*

stre

ngth

ofencapsulation,

give

nin

arbi

trar

yunits,

seeChapter

4,MaterialsandMethods

tlength

of

radial

cell[mm]

tindividuals

tested

§basedon

fami

lyaverages

Table

A4.2.Averageimmune

responseandbody

size,andsample

size

ofworkers,

hapl

oidand

diploidmales

oftheexperimentalbumblebee

colonies

(Bombus

terrestris

).

Colony

Treatment

(inbred/

Maternal

family

Immuneresponse(SE)t

workers

haploid

diploid

Body

size(SE)+

Sample

size§

workers

haploid

diploid

workers

haploid

dipl

oid

outbred)*

line

males

males

males

males

males

males

C269

12

61.79±3.43

54.56±4.11

2.95±0.02

3.41±0.03

10

9

C298

02

47.11±5.84

34.47±5.47

2.95+0.07

3.11±0.07

10

9

C79

02

55.64±6.64

47.79±5.22

2.89±0.07

3.50±0.05

10

9

C19

14

50.62±4.91

37.66±6.66

3.04±0.03

3.53±0.06

10

10

C78

14

56.14±5.21

31.80+6.04

36.46±4.61

2.80±0.07

3.53±0.05

3.06±0.07

10

714

C169

14

55.39±5.48

44.08±5.74

2.98±0.07

3.23±0.04

10

10

C290

14

50.39±5.71

47.96±4.47

34.14±5.00

2.93±0.06

3.53±0.03

3.16±0.05

10

12

10

C327

14

66.93±3.20

45.64±4.85

2.84±0.05

3.45±0.04

10

9

C251

14

54.37±3.64

40.52±4.49

2.92±0.06

3.37±0.03

10

10

C183

04

67.53±5.60

39.00±7.55

2.93±0.14

3.33±0.04

99

C291

04

64.67±6.92

46.09±3.80

2.83±0.09

3.46±0.07

10

10

C37

04

50.62±4.91

34.06±3.21

2.92+0.03

3.39+0.06

10

10

C83

17

38.49±4.12

3.77±0.04

9

C188

17

60.71±3.48

51.83±4.22

2.84±0.09

3.43±0.04

10

9

C268

17

58.77±5.26

49.82±3.46

2.82±0.06

3.25±0.05

10

10

C181

07

59.33+6.90

48.95±5.70

2.97±0.19

3.62±0.04

610

C66

07

50.79±2.86

36.96±3.45

3.01±0.04

3.43±0.04

10

7

C370

07

39.90+7.15

45.19±10.45

2.85±0.04

3.68±0.03

10

9

C283

18

50.82±4.65

43.98±3.34

35.88±3.21

3.00±0.09

3.68±0.04

3.07±0.04

10

617

C351

08

58.50±4.72

34.83±6.62

2.85±0.05

3.32±0.05

10

9

C105

08

41.65±6.80

42.30±4.96

2.74:t0.07

3.71±0.03

10

9

C293

08

43.63±5.15

43.41±6.30

2.95:t0.07

3.67+0.02

10

10

C140

08

57.83±3.09

3.13:t0.04

10

Table

A4.2.Continued

Colony

Treatment

(inbred/

Maternal

family

Immuneresponse(SE)t

workers

haploid

diploid

Body

size(SE)+

Sample

size§

workers

haploid

diploid

workers

haploid

dipl

oid

outbred)*

line

males

males

males

males

males

males

C63

11

150.82±6.08

35.06±3.43

2.57±0.08

3.30±0.04

10

20

C216

11

148.00±7.72

41.76±5.01

27.65±4.59

2.37±0.08

3.36+0.06

2.73±0.09

910

11/1OH

C346

11

166.58±3.49

53.66±5.12

41.86±4.67

2.87±0.08

3.23±0.04

3.11+0.10

913

7/5H

C359

111

66.75±4.40

43.40±5.88

2.71±0.06

3.32±0.07

10

10

C375

111

58.98±7.22

53.28±5.23

2.94±0.06

3.67±0.03

910

C144

011

69.54±4.54

50.69+5.54

2.64±0.05

3.23±0.07

10

9

C354

01

158.19±4.50

46.81±4.72

2.72±0.09

3.33±0.05

10

10

C60

01

167.57±2.66

51.57±5.24

2.97±0.07

3.38±0.05

10

10

C191

011

66.51±4.29

61.27±4.29

2.76±0.09

3.43±0.08

810

C210

01

141.14±3.69

3.41±0.04

10

C53

01

150.61±5.15

54.41±4.89

2.73±0.09

3.40±0.06

10

9

C186

117

49.39±5.70

56.69±4.13

35.77±3.80

2.66±0.15

3.63±0.02

3.08-t0.06

10

14

10

C337

017

78.00+2.76

53.69+2.15

2.79±0.10

3.26+0.03

99

C302

017

42.73±7.27

53.53±4.21

2.43±0.13

3.41±0.06

10

10

C175

017

65.76±3.93

60.40±4.59

2.87±0.14

3.66±0.05

910

C8

017

50.07±5.64

41.69±4.27

2.72±0.08

3.52±0.03

10

9

C237

017

44.96±7.86

50.88±4.48

2.78±0.05

3.43±0.05

910

C138

017

65.60±4.96

47.16±2.17

2.83±0.07

3.30±0.06

10

10

inbred

mean±SE**

56.65±1.63

45.31+1.77

35.29±1.87

2.83±0.04

3.45±0.04

3.03:t0.06

157

178

69/66H

n(colonies

tested)

16

17

6

outbred

mean±SE**

56.38±2.20

46.36±1.61

2.84±0.03

3.43±0.03

220

217

n(colonies

tested)

23

23

*

0=

outbred,

1=

inbred

tstrength

ofenca

psul

atio

n,give

nin

arbitrary

unit

s,seeChapter

4,MaterialandMethods

tleng

thof

radial

cell

[mm]

§individuals

tested

Himmune

response/body

size

**

basedon

colonyaverages