NATURAL HISTORY AND SOCIOGENETIC ORGANIZATION OF THE BEE HALICTUS FARINOSUS (HYMENOPTERA: HALICTIDAE) IN NORTHERN UTAH

JENNIFER ALBERT

A THESIS SUBMITTED TO THE FACULTY OF GRADUATE STUDIES IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

GRADUATE PROGRAM IN BIOLOGY YORK UNIVERSITY TORONTO, ONTARIO

AUGUST 2012

©JENNIFER ALBERT, 2012 Library and Archives Bibliotheque et Canada Archives Canada

Published Heritage Direction du Branch Patrimoine de I'edition

395 Wellington Street 395, rue Wellington Ottawa ON K1A0N4 Ottawa ON K1A 0N4 Canada Canada Your file Votre reference ISBN: 978-0-494-90067-3

Our file Notre reference ISBN: 978-0-494-90067-3

NOTICE: AVIS: The author has granted a non­ L'auteur a accorde une licence non exclusive exclusive license allowing Library and permettant a la Bibliotheque et Archives Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par I'lnternet, preter, telecommunication or on the Internet, distribuer et vendre des theses partout dans le loan, distrbute and sell theses monde, a des fins commerciales ou autres, sur worldwide, for commercial or non­ support microforme, papier, electronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in this et des droits moraux qui protege cette these. Ni thesis. Neither the thesis nor la these ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent etre imprimes ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author's permission.

In compliance with the Canadian Conformement a la loi canadienne sur la Privacy Act some supporting forms protection de la vie privee, quelques may have been removed from this formulaires secondaires ont ete enleves de thesis. cette these.

While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n'y aura aucun contenu removal does not represent any loss manquant. of content from the thesis. Canada Abstract

Nesting biology, phenology and sociobiology were studied in an aggregation of the primitively eusocial ground-nesting bee Halictus farinosus in Northern Utah. Nest architecture was typical of the genus but key phenological events were delayed up to two weeks when compared to earlier studies of the same population. Bees were genotyped at six variable microsatellite loci to reveal kin structure within each nest.

Polyandry was uncommon in H. farinosus queens whose population wide effective mating frequency was 1.07. The queen produced the vast majority of the brood (98%) while she was present and workers took over reproduction upon being orphaned. There were significant differences in sex ratios between female-biased queenright and male- biased queenless nests (t = -3.72, p = 0.003). Together these results generally agree with the predictions of inclusive fitness theory and support the view that haplodiploidy is important in the evolution of eusociality.

ii Acknowledgements

This work was funded by an NSERC discovery grant awarded to Dr. Laurence Packer. I would like to sincerely thank a number of people for their contribution to the project. I am grateful to Terry Griswold for sharing his expertise and helping me initiate my field study. He and his wife Rhonda also graciously welcomed me into their home during my field season. I would like to thank David Trew for his help with nest excavations and for answering my many questions over the last few years. I am grateful to Dr. Scott Tarof,

Dr. Amro Zayed, Dr. Robert Paxton and Pat Kramer for their lab related advice and assistance and to Dr. Cory Sheffield for his assistance with the analysis. A special thanks to Stefanie Cargnelii for all of her hard work assisting me in the lab. My thanks also go to

Jessica Albert, whose skillful editing has vastly improved the work. Finally, I am ever grateful to my supervisor Dr. Laurence Packer for his help with my field study and for his comments on the many versions of my thesis. His advice, encouragement and guidance over the years have been invaluable to me.

To Mom, Dad and Tom who supported me emotionally and financially during this process, I dedicate this thesis to you.

iii Table of Contents

Abstract ii Acknowledgements iii Table of Contents Iv List of Tables v List of Figures vi

Chapter One: Introduction 1

Chapter Two: Nesting biology and phenology of Halictus farinosus (Hymenoptera: Halictidae) in Northern Utah 12 Abstract 13 Introduction 13 Methods 16 Results 20 Discussion 26 Tables and Figures 31

Chapter Three: Sociogenetic organization in the primitively eusocial bee Halictus farinosus (Hymenoptera: Halictidae) 50 Abstract 51 Introduction 51 Methods 55 Results 60 Discussion 66 Conclusion 73 Tables and Figures 75

References 82

iv List of Tables

Table 2-1: List of all excavated nests 31

Table 2-2: Comparison of phenological events between years of study 33

Table 2-3: Comparison of H.farinosus nest sizes between 2002 and 2012 34

Table 2-4: Ovary development and matedness in queens and workers 35

Table 2-5: Comparison of social and life history traits between social Halictus 36

Table 3-1: Microsatellite primers and variability in H.farinosus 75

Table 3-2: Estimates of relatedness in worker and reproductive brood 76

Table 3-3: Summary of kin structure in reproductive brood nests 77

Table 3-4: Sociobiological data in all excavated nests 78

v List of Figures

Figure 2-1: Percentage of brood in different developmental stages 37

Figure 2-2: Halictus farinosus nests (external) 38

Figure 2-3: Internal structure of a typical H. farinosus nest (Or4b) 39

Figure 2-4: Proportion of nests where the queen was present at excavation 40

Figure 2-5: Head widths of H. farinosus workers and queens 41

Figure 2-6: Wing lengths of H. farinosus workers and queens 42

Figure 2-7: Head width vs. wing length in adult females 43

Figure 2-8: Difference in mandible and wing wear between each queen and her workers 44

Figure 2-9: Mandible wear in queens and workers 45

Figure 2-10: Wing wear in queens and workers 46

Figure 2-11: Average daily maximum and minimum temperatures in the summer months for the years of study in Logan, Utah 47

Figure 2-12: Total rainfall per summer month for the years of study in Logan Utah 48

Figure 2-13: Cumulative degree days between March and August in years of study with key phenological markers 49

Figure 3-1: Genetic relatedness among females in all genotyped nests 80

Figure 3-2: Size difference between replacement queen and workers 81

vi Chapter One: Introduction

1 In nature there are numerous examples of cooperation among individuals, including altruistic behaviour. An action is considered altruistic when it comes at the expense of the actor's personal fitness in terms of breeding potential but enhances the fitness of another individual. Eusociality is essentially an extreme example of altruism. A eusocial society is defined by three characteristics: a reproductive division of labour between members, overlapping generations, and intergenerational help with rearing the brood (Wilson, 1971). In eusocial societies the helpers often form a sterile worker caste that exist to help the queen produce the brood by provisioning resources and helping to construct and defend the nest. Within the context of Darwin's theory of evolution by natural selection the origin of a sterile caste that does not contribute directly to the next generation becomes all the more difficult to explain: how do some individuals forgo reproduction if it is through reproducing that the genes for such behaviour would be passed on? Darwin himself recognized this problem, admitting that it initially seems "insuperable and actually fatal to the whole theory" (Darwin, 1859).

There are two essential difficulties with the evolution of sterility in workers. First is the paradox of the inheritance of sterility when the trait is necessarily carried by individuals who do not produce progeny. Secondly, it seems that sterility would never be selectively advantageous (Crozier, 2008). Darwin attempts to explain inheritance of the trait, saying that "selection may be applied to the family as well as the individual" (Darwin,

1859) indicating that selection can occur on multiple levels, not just at the level of the individual but at the colony level as well. Modern group selection emphasizes the

2 competition between groups as a more significant driving force in selection than the competition between members of the group (Wilson and Holldobler, 2005; Nowak at al.,

2010). Group selection theory stands in opposition to kin selection theory, which is discussed further in this review, although it has been suggested that the two do not differ empirically in their expectations or mathematical framework (Queller, 1992;

Gardiner et al., 2007; West et al., 2007a).

In 1964 William D. Hamilton published two papers that contained the first mathematical proof that altruism can be explained by evolutionary theory without invoking selection operating at a level "above" the individual (Hamilton, 1964 a, b).

Hamilton's rule states that a particular behaviour is selectively favoured when it leads to a net increase in the inclusive fitness of the actor. Mathematically it is represented as follows: br — c > 0. Where c is the lifetime cost to the productivity of the actor behaving altruistically, b is the lifetime benefit of the behaviour to the recipient and r is the coefficient of relatedness between the two individuals. This statistic describes in genetic terms how closely related the actor is to the beneficiary relative to the genetic similarity in the general population (Frank, 1998; West et al., 2007b). The expression br is the indirect fitness benefit on the actor calculated as the product of the benefit to the recipient and its relatedness to the actor. The cost must be less than the indirect fitness in order for the altruistic behaviour to be selectively advantageous (Hamilton, 1964a).

Hamilton's rule of inclusive fitness emphasizes the importance of close relatedness

3 between individuals in the evolution of altruistic behaviour. Essentially an individual's genes can be passed onto the next generation in an indirect way, through a family member who carries genes that are identical by descent. Behaviour that promotes productivity of those family members can be favoured, even when the cost of performing the act is quite high. The selective force that acts to increase inclusive fitness through direct and indirect fitness effects is known as kin selection (Maynard Smith,

1964).

Ensuring maintenance of an altruistic trait would require that organisms are able to direct their altruistic actions towards other altruists either through kin recognition or limited dispersal (Hamilton 1964 a, b). The ability of an altruist to recognize kin or other carriers of an altruistic trait should be selected for since it allows them to direct altruistic behaviour towards those individuals who are likely to share the trait (Hamilton, 1964b).

Kin discrimination has been demonstrated in a variety of cooperative animal groups including humans (Jacob et al., 2002), ground squirrels (Mateo, 2002), mice (Manning et al., 1992), long tailed tits (Russell and Hatchwell, 2001; Sharp et al., 2005), carpenter ants (Ozaki, 2005), and even in a species of unicellular slime mould (Mehdiabadi et al.,

2006). However it is thought that discrimination of this nature is most often based on environmental cues such as prior association rather than genetic ones (West et al.,

2007b). This is partly because genetic cues are predicted to be evolutionarily unstable, often inspiring cheating (acquisition of the discriminatory cue but not the altruistic behaviours) and becoming fixed as a result (Keller, 1997; Downs and Ratnieks, 1999;

4 Ratnieks et al., 2006; West et al., 2007b). If dispersal in the species is limited and the social group is mainly composed of close relatives who share genes by descent, altruistic actions may be performed indiscriminately towards the members of the group who share the same propensity to perform the altruistic action (Hamilton; 1964ab, Hamilton,

1972). While high population viscosity (i.e. low levels of gene flow) may cause increased competition between relatives (Taylor, 1992; Wilson et al., 1992; West et al., 2002) there are a number of environmental and behavioural factors that promote selection of kin cooperation when dispersal is limited (Queller, 1994; West et al. 2007b). These forces can work together in some instances. For example, in some bird species kin recognition is stronger when dispersal is more common (Cornwallis et al., 2010).

Eusociality is expressed in highly diverse ways in different taxonomic groups that converge on its defining characteristics. Because of the specificity of these characteristics it is presumed that there is a general evolutionary pathway leading from solitary or primitive social behaviour toward obligate eusociality. However, the specific steps in this pathway and the biological and environmental factors that promote its evolution are still under debate (Wilson, 2005; Nowak et al., 2010; Abbot et al., 2010).

The conventional model for the evolution of eusociality relies on inclusive fitness theory: close relatedness between nest mates means that the benefit of helping need not be as great in comparison to its cost. The close relatedness between siblings which results from monogamy facilitates the evolution of sib-social care likely by manipulation of the expression of maternal care genes in insect workers (Linksvayer and Wade, 2005). It is generally accepted that relatedness between siblings must be high in order to support the evolution and early maintenance of eusociality (Boomsma, 2009). In fact, it is thought that all current eusocial species come from monogamous ancestors and it is only after passing through this "monogamy window" and through fixation of obligate castes in highly eusocial organisms that reduced relatedness due to polygamy becomes possible (Hughes et al., 2008).

Eusociality is thought to have evolved at least nine times in the Hymenoptera: six times in bees (Cameron and Mardulyn, 2001; Coelho, 2002; Danforth 2002; Brady et al.,

2006; Schwarz et al., 2007; Cardinal and Danforth, 2011), twice in vespid wasps (Schmitz and Moritz 1998; Hines et al., 2007) and once in ants (Moreau et al., 2006). It has been suggested that the propensity for this trait to evolve in Hymenoptera might be due to their haplodiploid genetic system (Hamilton 1964 b). This hypothesis is supported by studies conducted on gall thrips (order Thysanoptera), which too are haplodiploid and also have an evolutionary origin of eusocial behaviour (Crespi, 1992; McLeish et al.,

2006; McLeish and Chapman, 2007). In haplodiploid organisms females are diploid and have two copies of each gene while healthy males develop from unfertilized eggs and, having only one copy, are thus haploid. The relatedness values between family members are different in haplodiploid organisms when compared with diploid ones. Because, barring mutation, each sperm a haploid male produces is genetically identical to all others, females on average share V* of their genetic material with their full-sibling sisters as opposed to in diploid organisms. This means that worker females are more closely

6 related to their sisters than they are to their own daughters. On the other hand, daughters will share only % of their genetic material with their brothers which develop from unfertilized eggs laid by their mother. The average relatedness between siblings in haplodiploid organisms is therefore Vi, the same as it is in diploid organisms. This dynamic has led theory to predict conflict between queens and workers over sex allocation in the nest (Trivers and Hare, 1976). As workers are three times more closely related to sisters than to brothers, those who remain to help in a nest headed by a monogamous queen would be expected to attempt to skew the queen's brood sex ratio in favour of females and/or to lay haploid male-destined eggs themselves (Trivers and

Hare, 1976; Pamilo, 1991). There is evidence of both sex ratio biasing and male production by workers in eusocial Hymenoptera (Trivers and Hare, 1976; Bourke, 1988;

Packer and Owen, 1994; Schwarz et al., 1994; Bourke and Franks, 1995; Sundstrom et al., 1996; for reviews see Queller and Strassmann, 1998; Chapuisat and Keller, 1999;

Ratnieks et al., 2006; Crozier 2008).

Whether male haploidy (haplodiploidy) has a role to play in the evolution of eusociality has become a somewhat controversial topic (Queller and Strassmann, 1998;

Wilson and Holldobler, 2005; Wilson, 2008; Nowak et al., 2010). Critics have pointed to low relatedness values in many eusocial colonies facilitated by multiple mating

(Muralidharan et al., 1986; Gadagkar, 1990), however, eusociality may have become established in these groups before the evolution of polygamy (Hughes et al., 2008;

Boomsma, 2009). The haplodiploid hypothesis is also challenged by the fact that many

7 diploid species, including termites, some beetles, shrimp, and naked mole rats, have been shown to exhibit eusociality (Wilson, 2008; Nowak, 2010), however, diploid organisms too are predicted to evolve eusociality under inclusive fitness theory. Under

Hamilton's rule it is possible for eusociality to evolve even when relatedness is low so long as the benefits are great enough to outweigh the cost of forgoing reproduction

(Queller and Strassmann 1998; Boomsma, 2009). Still, more haplodiploid groups exhibit eusociality that would be expected by chance (Crozier, 2008) and models suggest that it is easier for eusociality to evolve with male haploidy compared to diploidy (Pamilo,

1991; Linksvayer and Wade, 2005). At the very least haplodiploidy is likely to be one of the promoting factors in the evolution of eusociality and the predictions of this hypothesis warrant further investigation.

Empirical evidence does not line up with simplistic predictions based on relatedness and male haploidy (Richards et al., 1995; Hammond and Keller, 2004).

Confounding behavioural and ecological factors related to colony efficiency, conflict resolution and reproductive control in the Hymenoptera must be considered when modelling the predicted evolutionary outcome (Chapuisat and Keller, 1999; Ratnieks et al., 2006). Numerous mathematical models which are based on inclusive fitness and take into account ecological and behavioural interactions have been developed to explain intergenerational cooperation and the tendency for the queen to produce most of the offspring (high reproductive skew). Reeve and Keller (2001) gave an overview of these models, dividing them into several different types and showing how they are

8 described by Hamilton's rule. First, transactional models where either: the dominant individual has control and concedes reproduction to subordinates as staying or peace incentives; or, where the subordinate has control over reproduction and chooses to restrain themselves and allow the dominant to breed (Johnstone and Cant, 1999).

Second, tug of war models assume no complete control over reproduction by either the dominant or the subordinate. This predicts an amount of reproductive skew intermediate to concession and restraint models (Reeve et al., 1998). According to

Reeve and Keller (2001) empirical data suggest that the transactional models are most accurate in their predictions and are therefore the most useful, however, evidence can be presented to support various models and the assumptions of each model should be tested along with its predictions (Johnstone, 2000; Gardiner and Foster, 2008).

If we are to study the origins and maintenance of eusocial behaviour it is important that we look at weak forms of eusociality, such as those displayed in the bee family Halictidae, instead of advanced forms where distinct anatomical differences exist between castes from early development (Packer, 1992; Bourke and Franks, 1995;

Schwarz et al., 2007). In the Halictinae, eusociality has evolved at least three times and these events have taken place relatively recently compared to the origin of this trait in highly eusocial insects such as the corbiculate bees which have been social since the

Cretaceous (Michener and Grimaldi, 1988; Danforth, 2002; Brady et al. 2006; Cardinal and Danforth, 2011). Members of the Halictinae display considerable variability in their

9 social behaviour (Sakagami and Munakata, 1972; Eickwort et al., 1996; Soucy, 2002;

Richards et al, 2003; Boesi, 2009; Soro et al., 2010; and see Schwarz et al., 2007). In addition, social behaviour in many halictines shows a high degree of flexibility; for example, there have been multiple transitions from eusocial to solitary or parasitic behaviour in some clades (Danforth, 2002; Danforth et al., 2003, Schwarz et al., 2007;

Gibbs et al., 2012). Furthermore, some prospective workers may have the ability to reproduce instead of working (Yanega, 1988). The existence of socially polymorphic species in the Halictinae (Eickwort, 1996; Richards et al, 2003) indicates that eusocial behaviour in some members of the group may not be obligatory. Helping in these groups persists presumably because remaining in the natal nest to assist in rearing the brood is in some way beneficial to the helpers themselves. Thus, the halictines are an ideal clade in which to study the origins of eusociality.

The primitively eusocial bee Halictusfarinosus (Hymenoptera: Halictidae) is common in western North America. While it is known to be a eusocial species, social parameters of H. farinosus have never been examined in detail, for example it is not yet known whether workers lay eggs that contribute to the reproductive brood. This species was selected for study because it is common, yet rarely studied, it is large, its nest sites are conspicuous and a large aggregation was available to study. More importantly, it is closely related to Halictus species that are well-known for their interesting social behaviour and which include solitary as well as socially polymorphic species (Yanega,

10 1988; Richards et al., 1995; Eickwort et al., 1996; Packer, 1997; Richards, 2001). In this study I have compared natural history, nesting biology, dissection, measurement and emergence data in H. farinosus with previous studies on this bee in the same area (Nye,

1980; Sellars, 2004). In addition, highly variable microsatellite loci were used to investigate relatedness among nestmates, allowing us to determine parentage in the reproductive brood and compare predictions of models of social evolution to empirical data in a social halictid species.

11 Chapter Two: Nesting biology and phenology of Halictus farinosus Smith

(Hymenoptera: Halictidae) in Northern Utah.

12 Abstract

Nesting biology and phenology in an aggregation of the primitively eusocial ground-nesting bee Halictus farinosus was studied at Green Canyon, Utah from May to

August, 2010. Nest architecture was typical of the genus. Nests were small with an average of 3.5 workers and 13.5 reproductives per colony. Most workers were mated

(77.5%) and had ovary development (71.4%) and the queen-worker size differential was moderate (8.8% for head width and 5.6% for wing length), indicating that sociality in this species is weaker than in some other social Halictus species. The study's findings were compared with those of two previous studies on the same population (from 1977/1978 and 2002). Varying weather patterns in the years of study led to changes in phenological milestones: in the comparatively cold and wet spring season of 2010, nests were delayed by up to two weeks. While nest productivity was comparable between years, in

2010 the size difference between queens and workers was significantly larger than in

2002 indicating that weather conditions may have an effect on social parameters in this species.

Introduction

It is of the upmost importance that we have a detailed understanding of the biology and natural history of our native pollinators (Sheffield et al., 2012). Repeat field studies

13 in the same population can allow us to make comparisons between years to determine how weather patterns might affect phenology, productivity and other biological attributes which will help us to better predict how a group might respond to changing ecological conditions (Richards and Packer, 1995; Bartomeus et al., 2011). investigating the natural history and life cycle of a bee species inevitably means studying their nesting behaviour. Most bees burrow into the ground and construct brood cells off the side of one or many main tunnels (Michener, 2007, pp. 23-29). Most Halictinae (the sweat bees) are ground nesters and there is considerable diversity in the size and structure of these nests (see Sakagami and Michener, 1962 for an early review). These bees are also mass provisioning: enough pollen and nectar are collected before the egg is laid on the pollen mass within the cell and the brood cell entrance is then sealed, usually until the individual is fully developed (Sakagami and Michener, 1962). The size of an adult bee depends on the quality (Roulston and Cane, 2002) and size (Plateaux-Quenu, 1983;

Richards and Packer, 1994) of the pollen mass provided for it. The size of the pollen mass will depend on multiple environmental and biological factors including the size and activity levels (Gathmann and Tscharntke, 2002; Pereboom and Biesmeijer, 2003) of active foragers in the nest and the amount of available resources which, in turn, are highly dependent on weather and other environmental factors (Packer, 1990; Minckley et al., 1994; Richards, 2004).

14 Halictus farinosus Smith is a non-metallic halictine species that is readily identified by its large size (12 -14 mm) and strong apical bands of yellowish pubescence on the metasomal terga. It is found in the North American West ranging from British Colombia to California and has been reported as far east as Nebraska (Ascher and Pickering, 2011).

Halictus farinosus is polylectic: Nye (1980) reports foraging by H. farinosus on 43 plant species from 14 different families. It is thought to be an important pollinator of carrot

(Nye, 1980), onion (Parker, 1982), and sunflower (Parker, 1981). In Utah, H. farinosus nests in dry, sandy soil in areas of sparse vegetation and their simple, usually unbranched, nests extend up to 65 cm into the ground (Nye, 1980; Seliars, 2004).

Halictus farinosus has a two-phase life cycle and exhibits primitive eusociality similar to other primitively eusocial halictines (Michener, 1974, pp. 269-300; Packer, 1986; Packer,

1992; Richards, 2001; Soucy, 2002; Schwarz et al., 2007; Richards et al., 2010). Queens found nests independently in April to May in north Utah (Nye, 1980; Seliars, 2004), and produce a worker brood that help to produce a second, reproductive, brood which begins to emerge in mid August. The behaviour of this bee is similar in California: they nest in sandy soil in open areas of sparse vegetation and have a two-phase nesting cycle but with the warmer climate the nests tend to extend deeper (up to 80 cm) and more workers are produced (Eickwort, 1985). The nesting season is also longer in California, lasting from mid February to October (Eickwort, 1985).

15 In this study we present phenological, sociobiological and nest architecture data from field studies on H.farinosus in North Logan, Utah. As this population has been studied previously (Nye, 1980; Sellars, 2004) we compare phenological and nest productivity data among years of study. We draw similar comparisons between Halictus farinosus in Utah and other closely related Halictus species.

Methods

Study sites and nest excavations

All nests were excavated in July and August 2010 at Green Canyon, North Logan,

Utah (41.769° N, 111.773° W, 1589 m). This site was selected because it was the largest known H.farinosus nesting aggregation (Nye, 1980; Sellars, 2004). The site was a 0.25 km2 field near the mouth of the canyon, just before the trail-head. Six 2 x 2 m study sites, marked B to G, were divided into quadrants of 1 square meter each, nests beginning with N were outside of these areas. Each site was monitored during daylight hours for at least two hours every second day from June 13th until July 30th excepting days with exceptionally bad weather. Nests referred to as first brood nests are those excavated on or before July 17th, the date when the first workers emerged. Second brood nests were any nests excavated after this date and late summer nests were excavated after August 7th. Nests were excavated throughout the season but because

16 we wanted to maximize sociobiologically relevant data the majority of second brood nests were excavated during a short period in early to mid August, just before the reproductive brood started to emerge (see Table 2-1for excavation dates).

Nest entrances were blocked before sunrise on the day of excavation to prevent escape of any inhabitants. Talcum powder was blown down the entrance and a plant stalk or wire was inserted into the tunnel to help follow it during excavation. A 30 cm deep hole was dug in a clear area next to each nest and careful excavation continued from the hole toward the nest until the tunnel was uncovered. All juvenile and adult bees were removed from the nests and stored directly in 95% ethanol; no adults present in the nests during excavations were lost. All components of the nests were recorded during excavation including unfinished pollen balls, mouldy cell contents and empty cells. Nest depth measurements were taken for nests by running a measuring tape from the ground level and following along the main tunnel. Activity in most excavated nests was monitored from June 13th onwards. Foundresses in these nests, with the exception of those nests from areas B and N were marked in the spring with unique combinations of three colours of Testors™ model paint on their thorax. Observations of nests from areas B and N began as the first brood emerged because high mortality at other sites led to a low number of viable nests for excavation. Queens in B and N nests were determined using variable microsatellite loci following methods outlined in Chapter 3.

17 Dissections and measurements

Ovary dissections and measurements were performed on all adult females using a Leica WILD M3B light microscope at 16 x magnification with an ocular micrometer following the general methods of Ordway (1965). Wing and mandible wear were each given a score ranging from 0 to 5 where 0 was no wear and 5 was extremely worn. Wear scores, head width and wing length measurements followed the methods of Richards

(2010). Ovary development was assessed based on the size of developing oocytes: 0 when none were present and 1 for each fully developed oocyte. Intermediate conditions were scored to the nearest quarter based on size relative to a complete egg. These data were summed to give an overall development score (Packer, 1986). Development and wear scores were averaged for each group analysed and for the entire population.

Matedness was determined by inspection of the spermatheca when it was found intact

(it was not found in 9 of the 86 bees dissected). Sexing of juvenile brood was accomplished by genotyping at six microsatellite loci (see chapter 3) and confirmed in pupae by morphological inspection. Gut contents including pollen, nectar and in some cases ingested oocytes were also recorded during dissections.

Weather

18 Weather data were gathered from the Utah State University climate centre website (2011) from a weather station 5.5 km away from the study site for 2010 and all previous years of study in the population (1976,1977,2002).

Calculations and statistical analyses

Size differences between castes were calculated following (Packer 1992) as

D=m-sw1 SQ * '

where S^and Sw are the sizes (head width or wing length in mm) of the queen and worker respectively. First brood productivity was measured as the average number of individuals in first brood nests after the foundress stops foraging, indicating that all brood had been produced. Productivity was calculated as brood per working female in each nest. Because worker mortality was high, minimum productivity was also calculated as the number of brood divided by the average number of spring (worker) brood individuals found in nests multiplied by the proportion of spring brood that was female. The sex ratio presented is the number of males divided by all brood.

Degree days (base 10) were calculated for each day by the basic equation:

DD = ————————— - 10 (2)

19 Where Tmin and Tmax are the minimum and maximum temperatures reached that day, and cumulative degree days are calculated (Zalom et al., 1983). Correlation between wing length and head width were calculated using the standard coefficient of correlation calculation (r). All tests of significance were 2 tailed unpaired t-tests using

Minitab* Statistical Software.

Results

Phenology

All nests studied were singly founded and foraging had commenced in some cases by the initiation of the study on June 13th. Only one H.farinosus nest founding event was witnessed after this point but it was outside of the areas under observation.

Brood production by the foundress alone occurred until early July with the last foraging trip observed on July 12th. In early July, foundresses were frequently seen guarding nest entrances and rarely left the nest. Nest excavations began on June 28th, the first fully developed adult worker found in a cell was excavated on July 15th and the first adult worker brood individual to leave a nest was observed on July 17th. No new nests were founded after their emergence in the area of observation. Worker foraging slowed noticeably after the first week of August. The first fully developed second brood male was found on August 11th, the first adult female in the second brood was excavated on

20 August 12th and second brood individuals continued to emerge throughout August and into September. Figure 2-1 shows nest contents by developmental stage over 6 day periods throughout the season.

Nest surveys conducted in mid-June and repeated in early August indicated a nest mortality of 59.5%, although nests of very low activity may have been missed in either survey.

Nest architecture

Halictus farinosus nests in Green Canyon were easily identified by their large tumuli (Figure 2-2). Bees entered the nest through a tunnel that ran horizontally through the tumulus to the main tunnel which typically extended approximately vertically. Tunnels changed course occasionally to avoid obstacles (i.e. roots or rocks).

Two nests had a single brood-containing branch off the side of the main tunnel and one had two brood-containing branches. During wet weather conditions in the spring, the nest entrances were plugged with soil, but otherwise remained open daily and overnight throughout the foraging season.

21 Nest depth for first brood nests ranged from 11 to 41 cm (x = 22.10 cm, SD =

7.44, N = 21), however the shallowest and deepest brood cells at this time were 7 cm and 20 cm respectively (x = 14.47 cm, SD = 2.64, N = 21). Second brood nests excavated ranged from 12 cm to 44.5 cm in depth (x = 30.3 cm, SD = 6.10, N = 37), were significantly larger (t = 4.54, p < 0.001) than first brood nests and had cells reaching to or very near to the end of the burrow.

All cells were constructed directly off the side of the main tunnel (Figure 2-3) and measured approximately 1.4 cm (x = 1.40, SD = 0.29, N = 7) long and 0.7 cm (x = 0.66, SD

= 0.14, N = 7) wide with a slightly constricted neck.

The average number of cells in first brood nests was 4.41 (SD = 2.04, N = 29).

There were significantly more cells in second brood nests with an average of 20.06 in late summer nests (SD = 10.08, N = 32, t = 8.2, p < 0.0001) including empty cells that were previously occupied by first or second brood individuals and cells with mouldy contents. The percentage of cells containing mouldy contents (mostly pollen) over all nests was 6.69%.

Spring and summer brood

22 Sixty-six nests were excavated in Green Canyon throughout the season, 29 contained the juvenile first brood and 37 contained mostly the second, reproductive brood. The first brood nests contained between 1and 6 individuals with an average of

3.52 (SD = 1.96, N = 29). There were fewer brood than brood cells because in some cases cells were unfinished or contained mouldy pollen instead of live brood. At the time of excavation an average of only 1.74 workers were present within second brood nests. Active summer brood nests produced between 1 and 46 individuals with an average of 13.49 (SD = 9.93, N = 37). In both broods, individuals at later developmental stages tended to be closer to the nest entrance and cells containing only fresh pollen were always among the deepest excavated indicating that shallower cells were completed first.

A breakdown of all nest contents in the first and second brood nests is presented in Table 2-1. A total of 22 of 29 first brood and 12 of 32 second brood nests contained the founding individual at the time of excavation (Figure 2-4). It could not be determined if the foundress was present in the remaining five nests because genotype data were not available. Mean productivity of the queen alone (first brood nests) was

3.52. Mean productivity of worker brood females was calculated using an estimate of the number of first brood females over the lifetime of the nest. As there was an average of 3.52 individual bees produced in first brood nests and the proportion of the first

23 brood that was female was 0.851, this suggests there were 2.99 first brood females per nest on average. As second brood productivity was 13.48 individuals per nest, this gives

4.51 reproductive brood per worker female. Mean maximum second brood productivity using the actual number of workers found in the nest (not including the queen) was

6.99. The sex ratio in the first brood was 0.149 (SD = 0.31, N = 16) weighting nests equally and 0.159 with individuals weighted equally. The sex ratio in the second brood was 0.464 (SD = 0.352, N = 27) averaged over nests and 0.445 in the population as a whole; neither of these estimates were significantly different from a 1:1sex ratio.

Caste dimorphism

Dissections of workers revealed that 71.4% (N = 43) had developed ovaries and

77.5% (N = 40) were mated. All queens dissected were mated (N = 31) and all but one that was found dead in the nest with desiccated ovaries had some level of ovary development (N = 31). The level of ovary development in queens (x = 1.185, SD = 0.668,

N = 31) was significantly greater than in workers (x = 0.648, SD = 0.673, N = 49, t = 3.49, p = 0.0008). Workers in nests where the queen was present had lower but not significantly less egg development (x = 0.519, SD = 0.457, N = 26) than workers in orphaned colonies (x = 0.793, SD = 0.841, N = 23, t = 1.44, p = 0.157).

24 Queens averaged significantly larger than workers in both head width (x = 3.52 mm, SD

= 0.15, N = 33 in queens, x = 3.30 mm, SD = 0.15, N = 55 in workers, t = 6.61, p < 0.0001) and wing length (x = 3.14 mm, SD = 0.15, N = 27 in queens, x = 2.94 mm, SD = 0.13, N =

45 in workers, t = 5.80, p < 0.0001) (Figures 2-5 and 2-6). This gives a population-wide caste size difference of 6.12% for head width and 6.11% for wing length. Queens were also generally larger in both measures than their own workers with a mean percent difference of 8.79% for head width and 5.63% for wing length. There was a strong positive correlation between head width and wing length over the whole population

(Figure 2-7, r = 0.695). Queens were not generally more worn than workers in either mandible (x = 3.75, SD = 1.21, N = 28 for queens and x = 3.57, SD = 1.53, N = 45 for workers, t = 0.51, p = 0.61) or wing wear (x = 3.13, SD = 1.36, N = 30 for queens and x =

2.81, SD = 1.64, N = 43 for workers, t = 0.88, p = 0.38). Queens were generally more worn than their own workers (Figure 2-8). Both wing and mandible wear increased throughout the season in queens, and especially in workers (Figures 2-9 and 2-10).

Temperature and rainfall

Minimum and maximum temperatures were low in the spring of 2010 compared to the 65 year average but were comparable to the average throughout the summer

(Figure 2-11). Spring rainfall was high in 2010 compared to average, but was close to the average throughout the rest of the season (Figure 2-12).

25 Discussion

Comparisons of nesting biology of H. farinosus among years

Halictus farinosus has been studied at Green Canyon in the 1970s by Nye (1980) and in 2001 and 2002 by Sellars (2004). It has also been studied, though for a shorter duration, in Davis, California by Eickwort (1985). In this section we compare the results obtained among years and between localities.

Phenological events in H. farinosus 2002 and 2010 coincide well with the number of degree days accumulated in these years, with the first adult brood emerging at around 500 days (by equation (2), Figure 2-13). Thus, spring temperature likely had an impact on phenological events relating to the life history of this bee species. However, in

1976 and 1977 minimum temperatures were much higher and a higher number of degree days had accumulated before workers emerged (800 and 600 days respectively,

Figure 2-13).

The amount of rainfall may have been responsible for the differences in emergence dates between years and the poor correlation with degree day

26 accumulation. Brood development and emergence dates (Table 2-2) were delayed several weeks in 2010 compared to 2002 (Sellars, 2004) but matched those reported by

Nye (1980) for 1977. Timing in 1976 was intermediate (Table 2-2). The amount of spring rainfall followed the same pattern: high in May of 1977 and 2010 low in 2002 and intermediate in 1976. The drier weather in May of 1976 and 2002 likely allowed queens to initiate foraging earlier, to forage more often than in 1977 and 2010 and to produce brood sooner.

Nests were shallower in both brood producing periods in 2010 (Brood 1: x = 22.1, SD

= 7.4, N = 21, t = 2.30, p = 0.026, Brood 2: x = 30.3, SD = 6.1, N = 37, t = 6.60, p < 0.0001) compared to 2002 (Brood 1: x = 26.7, SD = 6.8, N = 30, Brood 2: x = 42.7, SD = 10.6, N =

86). However, the number of brood produced per nest in both worker and reproductive phases was comparable between years. The number of first brood individuals produced, x = 3.52, (SD = 1.96, N = 29) in 2010 and x = 3.20 (SD = 2.64, N = 30) in 2002, was not significantly different between years (t = 0.52, p = 0.60), although excavations commenced much earlier in 2002. Also, the number of second brood individuals produced per nest was not significantly different between years (2010: x = 13.49, SD =

9.93, N = 37, 2002: x = 13.06, SD = 13.36, N = 81, t = 0.17, p = 0.86) (Table 2-3).

Productivity estimations (brood per working female) were not significantly different

27 between years with an average productivity of 6.99 (SD = 4.21, N = 35) in 2010 and 6.09

(SD = 7.11, N = 70) in 2002 (t = 0.692, p = 0.491).

The percent difference in head width between queens and workers was significantly larger in 2010 (t = 2.01, p = 0.050) at 8.79%, compared to only 5.79% in 2002. Weather was much harsher in the spring of 2010 with more rain and colder temperatures compared to 2002 (Figures 2-11,2-12). Poor spring weather conditions may have led to smaller workers being produced in 2010 compared to 2002. It is generally thought that smaller individuals are less efficient at foraging compared to larger individuals because they cannot forage as far or carry as much pollen (Gathmann and Tscharntke, 2002;

Richards, 2004). Floral resources in Green Canyon are generally very sparse and foraging trips in the H.farinosus population there are typically very long, often well over an hour

(Nye, 1980; Sellars, 2004; personal observations). Thus, foraging in this population is likely energetically taxing. Despite this, reproductive nest productivity between the two years was not significantly different.

The proportion of workers that were mated also differed greatly between years and a greater percentage of workers in 2010 had developed ovaries but the average level of egg development in workers was not significantly different between years (Table 2-4).

This result is unexpected as the workers would be predicted to be less fertile when the

28 queen-worker size differential is larger and queens presumably have more control

(Richards et al., 1995; Richards and Packer, 1996). However, excavations in 2002 were performed at regular intervals throughout the season as opposed to 2010 where many were excavated in a short period of time late in the season. The difference in sampling dates between years could account for the greater number of mated and ovarially developed workers in 2010.

Comparisons to other species

Phylogenetic data, both morphological (Pesenko, 1984) and molecular (Danforth et al., 1999) have been used to understand the systematics of Halictus. Halictus farinosus forms a sibling species pair with H. parallelus which, although stated by Knerer (1980) to be solitary, is probably a social species (Packard, 1868; Taylor and Packer, unpublished observations). These two are then closely related to the socially polymorphic H. rubicundus (Soucy, 2002; Soro et al., 2010) and the solitary, occasionally communal, H. quadricinctus (Knerer, 1980; Sitdikov, 1987). It is more distantly related to the other

Halictus species that are known to be eusocial such as H. ligatus and H. sexcinctus

(Packer, 1986; Richards, 2001), though the latter exhibits some additional social complexities (Richards et al., 2003).

29 Halictus farinosus produces a small number of workers and has low to moderate worker-queen size dimorphism compared to other social Halictus species. It also has a comparatively large number of worker females that are mated and/or have ovarian development, but this varies widely between years. Small size dimorphism, the small number of workers and high degree of mating and ovarian development in workers indicates that the level of sociality in H. farinosus is relatively weak compared to that in other social Halictus species (Packer, 1992). The nesting biology and behaviour of social populations of Halictus rubicundus are similar to those of H. farinosus in brood size, nest architecture and the level of caste size dimorphism (Table 2-5, Soucy, 2002). However, unlike H. farinosus, H. rubicundus is socially polymorphic (Eickwort et al., 1996; Soucy and Danforth, 2002; Soro et al., 2010). While solitary and social populations of H. rubicundus may form phylogenetically separate lineages (Soucy and Danforth, 2002) some halictines exhibit social polymorphism even within the same population (Packer,

1990; Richards et al., 2003; Hirata and Higashi, 2008). While there are no known solitary

H. farinosus populations, seasonal climatic differences might affect its sociality through the queen-worker size differential (Richards and Packer, 1996). The biology of H. farinosus should be studied in locations throughout its geographic distribution, including populations at high altitude and latitude where the species is more likely to be solitary, to fully understand the effect of climate conditions on life history and sociality in this species and determine whether solitary behaviour is within the species' repertoire.

30 Tables and figures

Table 2-1: List of all excavated nests, the date of excavation and a breakdown of their brood contents by developmental stage. Adults listed were found in the tunnel; callow adults found cells are included with pupae.

Nest Excavation Adults Pupae Larvae Eggs Pollen Number Date balls

C301 28-Jun-10 0 2 0 0 0 C302 28-Jun-10 0 0 0 0 1 C307 28-Jun-10 1 0 0 1 2 C308 28-Jun-10 1 1 2 0 1 C311 28-Jun-10 1 2 3 0 0 N0702-1 02-Jul-10 1 3 2 0 3 N0702-2 02-Jul-10 1 0 5 1 0 N0702-3 02-Jul-10 0 1 0 0 N0702-4 02-Jul-10 1 0 1 0 0 N0702-5 02-Jul-10 1 0 2 0 1 N0704-1 04-Jul-10 1 0 3 0 2 N0704-2 04-Jul-10 1 1 1 0 1 N0704-3 04-Jul-10 1 3 1 0 0 N0706-1 06-Jul-10 1 0 3 0 0 N0708-1 08-Jul-10 1 0 4 0 2 N0708-2 08-Jul-10 1 3 2 0 1 N0709-1 09-Jul-10 0 1 0 0 N0709-4 09-Jul-10 1 4 2 0 0 N0709-5 09-Jul-10 1 2 2 0 1 N0712-1 12-Jul-10 1 2 2 0 2 N0712-2 12-Jul-10 1 1 2 0 0 N0712-3 12-Jul-10 1 4 2 0 0 N0712-4 12-Jul-10 2 2 0 0 N0715-1 15-Jul-10 5 1 0 0 N0715-2 15-Jul-10 1 1 4 0 0 N0715-3 15-Jul-10 1 1 4 0 0 N0717-1 17-Jul-10 2 3 0 0 0 N0717-2 17-Jul-10 0 5 1 0 0 N0717-3 17-Jul-10 0 0 0 0 0 N0725-1 25-Jul-lO 3 5 0 0 0 N0728-1 28-Jul-lO 3 1 0 0 1 N0728-2 28-Jul-lO 0 1 0 0 0 N0728-3 28-Jul-lO 2 0 3 0 1 OrlOb 29-Jul-10 2 2 1 0 2 Or4b 31-Jul-lO 2 2 8 1 2 B105 08-Aug-10 3 2 4 0 4 G402 09-Aug-10 1 2 10 0 0 G408 09-Aug-10 2 4 11 1 0 F203 10-Aug-10 3 2 8 0 4 F211 10-Aug-10 1 0 4 0 0 F205 10-Aug-10 1 4 10 0 0 F101 10-Aug-10 1 2 0 0 0 N0811-1 10-Aug-10 3 12 12 0 1 E411 ll-Aug-10 2 6 9 0 1 E407 ll-Aug-10 1 1 11 1 1 E401 ll-Aug-10 2 1 11 0 0 N0811-3 ll-Aug-10 3 2 7 0 0 B101 12-Aug-lO 8 11 35 0 0 B108 12-Aug-lO 5 18 14 0 1 B211 12-Aug-lO 2 11 7 0 2 B102 12-Aug-lO 2 8 7 0 0 B106 12-Aug-lO 2 4 19 0 2 B107 13-Aug-10 5 11 22 0 0 B301 13-Aug-lO 3 6 2 0 0 B302 13-Aug-10 1 1 8 0 0 D303 14-Aug-10 1 9 8 0 0 D401 14-Aug-10 1 9 1 0 0 C208 15-Aug-10 11 12 0 1 C209 15-Aug-10 1 5 3 0 0 B311 16-Aug-10 1 6 2 0 0 B304 16-Aug-10 17 6 0 0 B311-2 18-Aug-10 1 8 7 0 0 B303 18-Aug-10 1 2 6 0 0 B306 18-Aug-10 5 15 11 0 1 B313 18-Aug-10 2 2 11 0 0 F310 18-Aug-10 1 3 0 0 0 Table 2-2: Summary of phenological data for all H. farinosus studies in Green Canyon. N/A indicates that the information in the column was not given in the study. In Sellars (2004) female and male reproductive emergence dates were not differentiated so the same date was given for each.

Queen First First First Source foraging First adult reproductive reproductive reproductive Year slows worker brood cells adult male adult female

1976 Nye N/A Late June N/A Late July Early August (1980) Nye 1977 N/A July 8 N/A Early August Mid August (1980) Sellars 2002 June 28 July 2 July 3 July 30 July 30 (2004) Present 2010 July 12 July 15 July 17 August 11 August 12 study

33 Table 2-3: Comparison of H. farinosus nest size, number of cells and mean offspring number between 2002 and 2010.

First brood

Depth (cm) Cells per Mean offspring nest X SD N X SD N 2002 26.71 6.78 30 1-12 3.20 2.64 30 2010 22.10 7.44 21 1-9 3.52 1.95 29 Compari t = 2.30, p = 0.026 t = 0.52, p = 0.60 son of means

Second brood

Depth (cm) Cells per Mean offspring nest X SD N X SD N 2002 42.67 10.64 86 3-56 13.06 13.36 81 2010 30.32 6.14 38 3-45 13.49 9.93 37 Compari t = 6.60, p < 0.0001 t = 0.17, p = 0.86 son of means Table 2-4: Summary of egg development and matedness in queens and workers in 2002 and 2010. Numbers in brackets are percentages, averages are presented with standard deviations.

2002 2010 Workers: Workers: Queen Queen not Queens Workers Queens Workers present present N 46 91 31 49 26 23 % with 1 egg 50% of full size (80.4) (41.8) (90.3) (44.9) (42.3) (47.8) % with 1 egg 75% of full size (50.0) (28.6) (41.9) (26.5) (26.9) (26.1) % with 1 egg of full size (130) (11.0) (25.8) (10.2) (7.7) (13.0) % with developed ovaries (97.8) (42.10) (96.8) (71.4) (73.1) (69.6) % mated (91.3) (46.10) (100) (77.5) (75.0) (80.0) Average amount of 1.429 +/- 0.913 +/- 1.185 +/- 0.648+/- 0.625 +/- 0.793 +/- ovary development 0.765 1.070 0.668 0.673 0.458 0.842

35 Table 2-5: Comparison of aspects of sociality and natural history in social species of the subgenus Halictus. A species with more advanced sociality would be expected to have higher brood sizes, a lower brood 1 sex ratio, fewer workers that were mated or with developed ovaries and a larger queen-worker size dimorphism (Packer and Knerer, 1985).

% workers First Second Sex with % Queen- brood brood ratio develope workers worker size size size brood 1 d ovaries mated dimorphism Source 3.5- 13.0- 5.0- 42.1- 46.1- Present study; H. farinosus 3.6 13.3 15.0% 77.4% 77.5% 6.4 - 8.6% Sellars (2004) Soucy, 2002; (social H. rubicundus 4.4 10.5 10.3% n/a n/a 9.6% populations) Packer, 1986; Richards and Packer (1995); 5.8-> 5.5%- Richards et al.. H. ligatus 9 10-16 14.6% 57.16% 42% 12.7% 1995 » In H. sexcinctus 0° 4.9 - 5.8 23.0% 73.7% 56% 6.9 -11.8% Richards, 2001

36 100%

Pollen 8 40% • Eggs c § 30% Larvae a 20% Pupae

Jun 28- July 4-9 July 10- July 16- July 22- July 28- Aug 3- Aug 9- July 3 15 21 27 Aug 2 8 14 Date (2010)

Figure 2-1: Percentage of brood in different developmental stages throughout the season, taken every six days. The rate of nest excavation throughout the season was not consistent. Number of nests excavated during the periods: June 28-July 3, N= 10; July 4-9, N=8; July 10-15, N=7; July 16-21, N=3; July 22-27, N=l; July 28-Aug 2, N=5; Aug 3-8, N=l; Aug 9-14, N=23; Aug 15- 20, N=9.

37 Figure 2-2: Four H.farinosus nests. Tumuli are apparent, markers designate nest number. 1 cm

Spr4b.17 dead pupa )0r4b.01 - empty

0r4b.02 emptyZy»-2' Or4b.19 empty 0r4b.03 mouldy pollen

0r4b.04 mouldy pollen

0r4b.05 pupa

0r4b.06 larva

Or4b.24 mouldy pollen 0r4b,07 larva Or4b.21 - mouldy pollen 15 larva Or4b.18 pollen Or4b 23 larva ~iOr4b.16 larva 0r4b.20 larva Or4b.l4 larva 0r4b.08 mouldy pollen

_Or4b.l3 mouldy pollen Or4b.22 pollen Or4b.12 larva Or4b.11 larva 0r4b.09 - adult 0r4b.10 adult

Figure 2-3: A typical second brood nest (Or4b) showing all cells with brood contents. 100%

• Queen not present • Queen present

Figure 2-4: Proportion of nests with queens in the nest at the time of excavation, organized in 6 day intervals. Very few excavations took place between July 16th and August 8th so data are lumped together to show overall trend over this period. Nests in which the queen could not be determined conclusively were omitted. Number of nests: June 28-July 3, N= 10; July 4-9, N=9; July 10-15, N=8; July 16-Aug 8, N=7; July 9-14, N=20; August 15-20, N=6.

40 • Workers

• Queens

Figure 2-5: Histogram of head widths in millimetres for adult H. farinosus females (workers and queens).

41 • Workers

• Queens

wing length (mm)

Figure 2-6: Histogram of wing lengths in millimetres for H. farinosus adult females (workers and queens).

42 y = 0.626x + 0.894

Queens

Workers

Reproductives

- Linear (all)

3.4 Head width (mm)

Figure 2-7: Head width vs. wing length in workers, queens and adult reproductives (brood 2 adults) in the H. farinosus population in Green Canyon showing a strong correlation between the two variables (r = 0.665).

43 3.5

3

2.5

2

I Mandible wear 1.5 c Wing wear 2 1 £ 2 0.5

0 -Q JO o ~rq—"_ -0.5 r- r>» o o o -1

Nest Figure 2-8: Differences between mandible and wing wear scores in each nest between the queen and the average wear in her workers for each nest. Queen mandibles were always more worn than their workers and wing wear in queens was nearly always higher except in two nests. Only nests where the queen and at least one worker were present are presented. Nests are listed in order of excavation date showing decreasing differences between the amount of wear in workers and queens as the season progressed.

44 A A W W

01ft. ••• •• §ft. (9 • wear in workers 01 A i • • • w Ml V • wear in queens JB •o c •• A. A / AA A ra w w w m / Jfc m - Linear (wear in E workers) - Linear (wear in •• 4 / / ^ m queens)

0 •— .—AA AA i— A . 21-Jun-10 ll-Jul-10 31-Jul-10 20-Aug-10 Date

Figure 2-9: Wear on the mandibles of queens (red) and workers (blue) throughout the season showing an increase in the amount of wear over time in both groups.

45 • wear in workers • wear in queens

- Linear (wear in workers) - Linear (wear in queens)

^—i' A A r- 21-Jun-10 06-Jul-10 21-Jul-10 05-Aug-10 20-Aug-10 04-Sep-10 Date Figure 2-10: Wear on the wings of queens (red) and workers (blue) throughout the season showing an increase in the amount of wear over time in both groups.

46 Average daily maximum temperature

• 1976 w 20 • 1977 5b 15 • Average • 2002 • 2010

July August

Average daily minimum temperature

25

20 11976 15 11977 w 10 3 i Average 12002

12010

April May June July August Month

Figure 2-11: Average daily minimum and maximum temperatures for the four years of study on H.farinosus in Green Canyon and the seasonal average over the previous 65 years.

47 • 1977 Average 2002

August

Figure 2-12: Total rainfall per month for the four years of study on H. farinosus in Green Canyon and the monthly average over the previous 65 years.

48 lioe

• Quwwtongnaiaw Mb om • FMopoductovbrooa •saaas- • FjrVKMM*rapTMUOMt

? ? 9 1000 + 9 J J 2°

!§

Dale Date Figure 2-13: Cumulative degree days between March and August in 1977,1978, 2002 and 2010. Phenological markers (queen foraging slows, first adult worker, first reproductive brood cell, first adult male reproductive, first adult female reproductive) have been included as coloured bars to note the dates of the events. Male and female reproductives were not differentiated between in the 2002 study, and some events were not recorded in 1977 and 1978.

49 Chapter 3: Sociogenetic organization in the primitively eusociai bee Halictus farinosus Smith (Hymenoptera: Halictidae)

50 Abstract

Halictus farinosus is a primitively eusocial species of sweat bee that is common in

Western North America. Adult and juvenile bees from thirty-seven nests within an H. farinosus nesting aggregation in Northern Utah were collected and genotyped at six highly variable microsateliite loci to reconstruct kin structure and estimate relatedness.

Polyandry was uncommon in H. farinosus queens whose population wide effective mating frequency was 1.07, with a second mate confirmed in four of the thirty-seven nests. The queen was present in only 52% of the excavated reproductive brood nests but she produced the vast majority of the brood (98%) while living. Workers took over reproduction upon being orphaned. The brood was heavily female biased when produced by a queen but heavily male biased when worker-produced resulting in a significant difference in sex ratio between queenright and queenless nests (t = -3.72, p =

0.003), leading to split sex ratios. Together these results generally agree with the predictions of inclusive fitness theory and support the view that haplodiploidy is important in the evolution of eusociality.

Introduction

Eusocial behaviour, as exhibited in the Hymenoptera, is highly altruistic requiring subordinate worker offspring to reduce or eliminate personal reproductive options to instead help their parent to reproduce (Wilson, 1971). Inclusive fitness (Hamilton,

1964a) is the most widely accepted theory describing how eusociality could have evolved. Hamilton's rule of inclusive fitness allows individuals to maximize their own fitness indirectly by helping their close relatives to survive and reproduce. It states that if the product of the measure of relatedness between two individuals and the beneficial action one performs to help the other is greater than the cost to the acting individual

(rb>c) then costly traits can be selected for. Thus the more closely related individuals are the easier it is for altruistic traits such as eusociality to be selected (Hamilton, 1964a).

Recent challenges to inclusive fitness theory as it applies to the evolution of eusociality have been brought forth in part by the founder of sociobiology (Wilson and Holldobler,

2005; Wilson and Wilson, 2007). Mathematical models have been presented that dispute the assumption that close genetic relatedness is important for the evolution of eusociality favoring instead selection at the group level (Nowak et al., 2010). Despite this, monogamy, which promotes very close genetic relatedness among full sisters, has been shown to be particularly important in the evolution of various levels of cooperation including eusociality (Hughes et al., 2008; Boomsma, 2009; Cornwalis et al.,

2010). The prevalence of eusociality in the Hymenoptera may in part be explained by their genetic system of haplodiploidy (male haploidy) which causes female offspring of a singly mated queen to be more closely related to their sisters than their daughters, but less closely related to their brothers than their sons (Hamilton, 1964b; Bourke and

Franks 1995; Crazier and Pamilo 1996; Boomsma, 2009). If relatedness and haplodiploidy are important factors in maintenance of eusociality, worker females are

52 predicted to lay male destined eggs themselves and to prefer that the queen produce reproductive sisters (Trivers and Hare, 1976; Pamilo, 1991).

It is difficult to understand the origins and early evolution of eusocial behaviour when studying highly derived eusocial forms, such as the corbiculate apids and most ants in which distinct anatomical differences exist between castes from early development (Michener, 1974; Bourke and Franks, 1995). Members of the Halictidae exhibit diverse social behaviours that are much more flexible than those of highly eusocial clades (Schwarz et al., 2007). There have been at least three evolutionary origins of eusociality within the subfamily Halictinae (Danforth, 2002) as well as multiple transitions from eusocial to solitary behaviour (Packer, 1997; Danforth et al., 2003). The existence of socially polymorphic species in the Halictinae (Sakagami and Munakata,

1972; Eickwort et al., 1996; Richards et al, 2003; reviewed by Packer, 1997) and the relatively recent evolutionary origin of eusociality (Brady et al., 2006) compared to highly eusocial bees (Michener and Grimaldi, 1988; Cardinal and Danforth, 2011) further strengthens the case for studying this group as a model for the origin of primitive eusociality.

53 Studies of Halictus farinosus in Utah and California (Nye, 1980; Eickwort, 1985;

Sellars, 2004; Chapter 2) indicate that it is a social ground nesting bee with a colony cycle comprising two distinct seasonal phases; first the overwintered foundress independently produces worker females; and second the workers forage to produce the reproductive brood. Small colony size, little size dimorphism between castes as well as significant ovary development and frequent mating in workers point to a system of weak eusociality in H. farinosus (Sellars, 2004; Chapter 2) making it an ideal system in which to study the origins and early maintenance of eusociality. It is not known however if H. farinosus workers contribute directly to brood production through oviposition that leads to adult reproductives. Studies in other social halictines based on allozymes, DNA fingerprinting and microsatellites generally show that nestmates are closely related

(Crozier et al., 1987; Mueller et al., 1994; Packer and Owen, 1994; Soro et al., 2009) and that reproductive skew, which measures the degree to which reproductive output is dominated by an individual, is typically high and in favour of the queen (Mueller et al.,

1994; Paxton et al., 2002; Soro et al., 2009; Ulrich et al., 2009). However, simple kin selection theory is not sufficient to explain all the complexities of behaviour in these species: multiple paternities (Richards et al., 1995; Paxton et al., 2002), worker production of female brood (Richards et al., 1995; Paxton et al., 2002) and nest switching (Paxton et al., 2002; Soro et al., 2009; Ulrich et al., 2009) are not predicted by inclusive fitness theory and yet they appear to be common. In this study we use

54 microsatellites to determine the breeding system and kin structure in nests of Halictus farinosus and compare this to studies of related species of sweat bee.

Methods

Field work and nest excavations

For a detailed description of field site, field methods and dissection methods see chapter 2.

Genotyping of microsatellite loci

A total of 10 first brood nests and 27 second brood nests containing a total of

572 individuals were genotyped at six variable microsatellite loci developed for study of the closely related species H. rubicundus (Soro and Paxton, 2009, Table 3-1). DNA was extracted using a high salt extraction protocol (Paxton et al., 1996) from half of the thorax of adult bees and a small amount of tissue taken from juvenile stages. A nested three primer PCR adapted from Schuelke (2000) was used to isolate and label the microsatellite loci. The M13 sequence (Table 3-1) was added to the 5' region of the forward primer at each locus that matched a third fluorescently labelled M13 primer

(Table 3-1). Products were labelled with either WELLRED D2 (black), IRDye700 (green) or

TYE665 (blue). The PCR conditions used followed those outlined for the nested PCR by

55 Schuelke (2000) using ideal annealing temperatures found by gradient PCR (Table 3-1).

PCR products were genotyped following the manufacturer's instructions for fragment analysis using a CEQ8000 molecular sequencer from Beckman Coulter Inc.. In cases where the results were ambiguous or unexpected, the PCR and genotyping were repeated.

Analyses

A population sample of 42 first brood females was used to calculate observed heterozygosity, Nei's unbiased (expected) heterozygosity (Nei, 1978) and the inbreeding

coefficient (Fis= 1 - H0/HE) by hand. The same sample of females was used to calculate deviation from Hardy-Weinberg equilibrium and the frequency of alleles using GENEPOP

4.1 (Rousset, 2008). Linkage disequilibrium between loci was calculated with GENEPOP

4.1 using the same subset of females and a similar subset using 28 males, with at most one per nest.

Sexing the brood was done initially by inspection of an individual's alleles at all loci: individuals that were heterozygous at one or more loci were scored as female. It is possible that diploid males could have been mistaken for females (Cook and Crozier,

1995). However, physical sexing of juvenile brood is possible after they are past the larval stages and no males were found among bees determined to be diploid. In

56 addition, inbreeding in this population was relatively low (Table 3-1), thus diploid males are not expected to influence the results presented here. The probability of a female being scored as a male was calculated as the probability that she would be homozygous at all loci using the population allele frequencies:

] (1)

Where p,2 is the probability that the allele (i) is homozygous and j is the locus and the product is taken over all available loci. If a bee was past the larval stages sexing was confirmed by inspection of the specimen. Sex ratio was calculated as the number of male offspring divided by the total number of offspring weighted equally over nests or over individuals. Investment ratios were calculated using wet weight of pupae weighted equally over nests or over individuals. Since fathers could not be captured, their genotypes were deduced using Matesoft 1.2 (Moilanen et al., 2004) from female offspring and a single paternity was assumed when all daughters in a nest carried the same paternal alleles. The chance that multiple mating went undetected was calculated as the probability of two males being identical over all loci:

j (2) (Seppa et al., 2011)

57 Bees identified as queens in nests labelled with a suffix of C, D, E, F or G were marked early in the season so they could be identified during excavation. They were later identified by their markings and confirmed as the queen by their multilocus genotype. In aggregations B and N, some females in nests that were discovered late in the season were determined to be queens by their genotype. Colony 2.0 (Wang, 2004;

Jones and Wang, 2010) was able to reliably assign marked queens to individual offspring within their nests (p > 0.977), thus permitting confidence in queen assignments in unmarked nests. The program was able to do this in all cases for both marked and unmarked queens except for 4 offspring in 3 nests where poor genetic information was available (offspring had missing allele data for at least four loci). In five nests where the queen was not present (N0811-1, B313, C208, D303, G402) her genotype was unambiguously deduced from the nest's eldest offspring with the help of Matesoft 1.2

(Moilanen et al., 2004). In two additional orphaned nests the queen's multilocus genotype was one of two possible genotypes because homozygosity at one locus made it impossible to distinguish between maternal and paternal alleles at that locus in heterozygous daughters. In an additional six nests too few of the queen's offspring were present to reliably deduce her genotype. Unambiguously deduced queens were added to the dataset for relatedness and sibship calculations.

58 Genetic relatedness (r) was estimated using Queller and Goodknight's (1989) regression relatedness algorithm and was calculated using Relatedness 5.0.8. Individuals found to be drifters (alien workers) and one locus in one nest found to have a null paternal allele (Rub73, B304) were excluded from the relatedness analysis. Relatedness estimates were done for the population as a whole and for a subset of nests that were found to be monogamous. Standard errors were obtained by jackknifingover loci.

Multilocus genotypes were manually inspected for all offspring in each nest and individuals were assigned first to matrilines, then patrilines within the matrilines, allowing us to group nestmates into full and half sibships and to deduce parentage and mating frequency (as in Seppa et al., 2011). Sibships were confirmed using Colony 2.0

(Wang, 2004; Jones and Wang, 2010). Relatedness data, sibship data and manual inspection of alleles allowed us to reliably deduce kin structure in most nests.

Reproductive skew was calculated following the general formula of Crozier and

Pamilo, (1996). Where NT is the number of potentially reproductive females (workers and queens) and

Pi is the actual reproductive output of the rth individual. Skew may have been overestimated in some cases as males carrying only maternal alleles were assumed to have been produced by the queen. Effective mating frequency was calculated using the following formula:

me = =r"2 (4) (Starr, 1984; Soro et al., 2009)

Where y. is the proportion of the queen's daughters fathered by male / and the summation is over all fathers. Sex ratios in queenright and queenless nests were analysed by a two sample, two tailed T-Test using Minitab* Statistical Software after arcsin transformation of sex ratio proportions.

Results

Microsatellites

Microsatellite loci were highly variable with many alleles per locus (x = 11.8, range 4

-15) and high observed and expected heterozygosities (Table 3-1). Loci were not in linkage disequilibrium: p > 0.11 for all loci in both male and female datasets. There was no significant deviation from Hardy-Weinberg equilibrium in the population subset (p >

0.05 for all loci and over all loci). The inbreeding coefficient in the population sample was not significantly different from zero (Table 3-1, t = 0.510, p = 0.621). At least one

60 locus (Rub 73) had null alleles that were unreliably amplified; however, this locus was useful for assigning parentage in most cases, particularly with the aid of Colony 2.0 which can account for the presence of null alleles.

The probability that an additional paternity was present when none were detected was very small in both worker and queen-produced females (p < 0.0007). The probability that an individual scored as a male by genotype alone was actually a female was even smaller (p < 0.0002).

Worker brood

On average only 3.4 workers were produced per nest and an average of 1.7 workers were found in those nests when excavated in late summer. Sex ratio in the worker brood was strongly in favour of females with an average ratio of 0.149 (SD =

0.31, N=16) over nests and an overall ratio of 0.159. All but one of the worker brood males came from two nests (five from N0708-2 and four from N0709-4). All individuals in both broods that were found to be males by inspection were haploid. Only one female was produced in either of these two (N0709-4.07), and she was the youngest member of the nest. It could not be determined if the queen in N0708-2 was mated because of damage to her abdomen during nest excavation. The single worker brood male produced outside of these two nests was the oldest of his nestmates. Workers were significantly smaller than their queen (t = 6.49, p < 0.0001) with a size differential of 8.79% for head width and 5.12% for wing length. Dissections revealed that 71.4% of workers had developed ovaries (N = 43) and 77.5% were mated (N = 40) (Chapter 2).

Queen's had significantly more ovary development than workers whether or not a queen was present in the worker's nest at the time of excavation (see Chapter 2).

Overall, workers from queenless colonies did not differ significantly in their level of ovary development from workers in queenright colonies (Chapter 2).

Relatedness

The average genetic relatedness between all nestmates, including the queen, was not significantly different from 0.5 (r = 0.508, SE = 0.024, CI = 0.062), and between all female brood (workers and gynes) in each nest was r = 0.641 (SE = 0.026, CI = 0.066), which is significantly higher than 0.5 and significantly lower than 0.75 (Figure 3-1).

Relatedness between females (excluding the queen) in monogamous nests was not significantly different than 0.75 (r = 0.730, SE = 0.0134, CI = 0.035). Overall, workers were not more closely related to the reproductive brood than queens were: r = 0.646

(SE = 0.029, CI = 0.075), compared to r = 0.534 (SE = 0.024, CI = 0.061) for queens. In addition workers were not more closely related to gynes than queens were overall (r =

0.591, SE = 0.041, CI = 0.107 for workers; r = 0.470 SE = 0.026, CI = 0.066 for queens).

However, in monogamous nests the relatedness value between workers and gynes was the expected 0.75 (r = 0.749, SE = 0.024, CI = 0.062). Worker to worker (0.538) and gyne

62 to gyne (0.632) relatedness values were significantly different from the predicted value of 0.75 in the population overall but not in monogamous nests (see Table 3-2). A list of relatedness values over nests in the population can be found in Table 3-2. There were no differences in relatedness predictions when deduced queen genotypes from absent queens were left out.

Social structure reproductive brood

Four general types of social structure were found among summer nests (Table 3-

3). The queen was generally present in type 1nests; she and one or two male mates could account for all offspring within the nest and reproductive skew was complete (S =

1). In type 2 nests the queen was present and some of the offspring had paternal alleles at one or more loci indicating that they were worker-produced. Reproductive skew in these nests was still very high (S > 0.957, Table 3-4). In type 3 nests the queen was not present at the time of excavation and the nesfs workers were responsible for producing some or all of the offspring. The eldest offspring, if any, were daughters and sons of the queen and siblings to the workers while the youngest were laid by the workers, suggesting that the workers only reproduced successfully after the queen had died.

Reproductive skew in type 3 nests was variable (from 0.02 to 1.00, with an average of

0.77, Table 3-3,3-4) and dependent on when the nest was orphaned. In three of the type 3 nests (C208, B101, N0811-1) it was determined that at least two workers were

63 responsible for producing offspring. In the remaining six type 3 nests the worker-laid offspring, all of which were male, could have all been produced by the same individual but because workers were so closely related this could not be verified reliably with only six loci. Type 4 nests were those in which kin structure could not be determined, either because the relationships between nest members were too complicated or some nestmates appeared to be unrelated to others (see discussion for details).

Type 1nests were the most common (N = 12), closely followed by type 3 (N = 9).

Workers laying eggs while the queen was present was uncommon (type 2, N = 3) as were situations where genealogies could not be determined (type 4, N = 3). Most reproductive brood males were produced in the type 3 nests and the majority of all males (60.8%) were produced by workers. Worker-produced brood was overwhelmingly male (93.8%) while queen-produced brood was predominately female (74.2%). The queen produced 96.6% of the female reproductives in the population.

Polyandry

A second patriline was identified in two spring and two summer nests (see Table

3-2). No more than two patrilines were present in any nest and the second patriline was responsible for less than 33% of the offspring in all cases. No worker polyandry was confirmed but the vast majority of offspring confirmed to be worker-produced were

64 male (Table 3-4). The effective mating frequency for queens was 1.07 males overall and

1.59 in polyandrous nests.

Sex and investment ratios

Reproductive brood nests had, overall, a balanced sex ratio: 0.465 or 1:1.2 (SD =

0.349, N = 27) with nests weighted equally or 0.455 (1:1.2) with individuals weighted equally. Monogynous queenright reproductive brood nests (Type 1, Table 3-3) tended to have more females than males with a ratio of 0.255 or 1:2.9 (SD = 0.228, N = 12) averaged over nests. The sex ratio in all nests where the queen was present (type 1and type 2 nests combined) was 0.284 or 1:2.5 (SD = 0.253, N = 15). The sex ratio where the queen was present but workers produced some brood (type 2 nests) was 0.400 (SD =

0.369, N = 3) or 1:1.5. Queenless nests (Type 3) tended to have more males than females with a ratio of 0.742 or 2.9:1 (SD = 0.304, N= 9) over nests.

Female pupae were heavier, weighing an average of 102.0 mg (SD = 21.8, N = 59) compared to 62.4 mg (SD = 14.0, N = 45) for male pupae. This makes the population wide cost ratio for females over males 1.63:1. The sex investment ratios are then 1:2.0 overall, 1:4.7 in queenright monogynous (type 1) nests, 1:2.5 in queenright polygynous

(type 2) nests, 1:4.1 in queenright nests overall (type 1and type 2) and 1.8:1 in

65 queenless (type 3) nests. The sex ratio in queenright nests was significantly more female biased than the sex ratio in queenless nests (t = -3.72, p = 0.003).

Drifting workers

One of the summer nests (B108) contained a small and slightly worn female that was unrelated to the nest's queen or to the offspring produced in the nest at the time of excavation and had no ovary development. A second nest (B101) had three worker brood females present who were unrelated to its five native workers, two of these drifters were likely full siblings. It was unlikely that this was due to errors in nest excavation because all worker brood females other than the replacement queen were found together near the nest entrance. This nest was incredibly productive as it had twice the number of workers and 45% more offspring than the next most productive nest. The unrelated workers had some ovary development but did not contribute genetically to the brood.

Discussion

Relatedness estimates

Relatedness estimates (Table 3-2) were significantly different from expected under conditions where all nests are headed by a singly mated monogynous queen in

66 four comparisons. First worker to worker relatednesses in first brood nests were lower than expected because in two of the ten there was an additional patriline. Worker- worker relatedness in brood one nests was not significantly different than expected when these polyandrous nests were excluded (r = 0.73, SE = 0.02, CI = 0.05). Second, queen to male comparisons in reproductive brood nests were lower than expected because of frequent oviposition by workers. This value could have been lower still since workers produced most of the males only when the queen was not present and comparisons could not be made when the queen's genotype was not available. Third, worker to gyne and gyne to gyne relatednesses in the reproductive brood were significantly lower than expected because of multiple mating and worker-laid gyne- destined eggs. Finally, three of the second brood nests were type 4 nests, meaning that the pedigree could not be reconstructed from the available allelic data. One of these

(B311-2) had high relatedness but parentage was not straightforward, two or three related individuals may have been laying eggs but no adults were present upon excavation. Two type 4 (B301, B311) nests had low relatedness between nestmate females, with their confidence limits reaching into the negatives (Figure 3-1). Nest B311 had a worker that was the most likely parent of the young males and two female offspring present in the nest may have been her half siblings. Nest B301 had a group of eight males and one female that were likely laid by the same individual, the remaining two nestmates, one worn worker and one uneclosed adult female, appeared to be unrelated to the other nest inhabitants but were likely full siblings of one another.

67 Parasitism (nest usurpation) or excavation error may have accounted for the genetic patterns these three type 4 nests which all came from a very densely populated area. All relatedness estimates were as expected for a monogamous haplodiploid organism when polygynous, polyandrous and type 4 nests were excluded from the relatedness analyses

(Table 3-2).

Polyandry

Polyandry occurred occasionally in this population with a second patriline detected in four (10.8%) of the nests studied. Even when the queen had mated with more than one male the majority of the offspring were fathered by only one leading to low effective mating frequencies. Polyandry effectively lowers relatedness between female nestmates making it less rewarding for them to help in the nest according to inclusive fitness theory. The small number of patrilines, small effective mating frequency and low occurrence of multiple mating in this species agrees well with the results of similar studies in other primitively eusocial halictines (Crozier et al. 1987; Packer and

Owen 1994; Mueller et al., 1994; Richards et al., 1995; Paxton et al., 2002). This also matches the predictions of inclusive fitness theory and the finding that monogamy is prevalent in the early evolution of eusociality (Hughes et al., 2008).

68 Worker Drifting

Worker drifting, where conspecifics from unrelated, nests will join a nest and behave as a worker, occurred in 7.4% of the second brood nests studied. Curiously, drifting has been reported in a number of other eusocial halictines (Mueller et al., 1994;

Soro et al., 2009; Ulrich, 2009) indicating that it is a common occurrence. It is understandable from the results in this study why drifting workers may be accepted into the nest: drifters come into the nest at no cost since they were not found to produce any of the nest's offspring. Evidence of oophagy was found in dissected crops of the dominant egg layer in each of the two nests that contained alien workers (B101, B108) and in the crop of one of the workers in alien-containing nest B108. Oophagy was only suspected in one additional worker outside of these two nests indicating that eggs laid by alien workers may be recognized and removed or that the presence of aliens causes nest residents to increase their policing behaviour. Why workers would switch nests in this population is not obvious from our results. Ulrich et al. (2009) suggest that workers in a population of Halictus scabiosae move from overpopulated nests to neighbouring ones near the end of the season for hibernation and by joining smaller nests they increase their chance of becoming the dominant egg layer in that nest in the next season (Ulrich, 2009). This is unlikely to be the reason in H.farinosus in Green Canyon as they do not overwinter beneath the natal nest (Nye, 1980) and unlike H. scabiosae they are not pleometrotic. In addition, drifters in the Green Canyon population appeared to join the most active nests since the nests with alien workers in this study were the two

69 largest in the population. As occurs in some honey bee and bumble bee species

(Birmingham et al. 2004; Lopez-Vaamonde et al. 2004; Nanork et al. 2005), drifting worker brood females may aim to increase their chances of opposition by moving to a larger nest where the dominant would have less control. However, while drifters in some Bombus species have a greater number mature eggs than native workers

(Birmingham et al. 2004), aliens in H.farinosus had the same amount of ovarian development as the nests' natives (t = 0.40, p = 0.70, N = 11).

Replacement queens and succession

In 56% (N = 9) of the queenless type 3 nests the worker-produced offspring could all have belonged to a single worker. In three of the other four nests the worker who laid the youngest offspring was present but the mother of the older worker-laid progeny was missing. This indicates that a worker takes over whenever the dominant member of the nest dies and becomes her replacement, a phenomenon known as serial polygyny

(Bourke and Franks 1995; Paxton et al., 2002). Dissections show that the nine replacement queens in type 3 nests, deduced by their genotype as being the likely mother of the youngest offspring, do not differ from queens in terms of ovary development (t = 0.023, p = 0.819, df = 48) but have significantly more developed ovaries compared to all other workers (t = 3.44, p = 0.001, df = 38) and compared to other orphaned workers (t = 2.56, p = 0.018, df = 22). This pattern resembles that found in Augochlorella aurata (as A. striata Mueller et al., 1994) and Lasioglossum laevissimum

(Packer and Owen, 1994} where one worker replaced the queen and subsequently dominated reproduction.

It is not readily apparent how the replacement queen is determined. In other social halictids size (Michener et al., 1971) and especially age (Michener et al., 1971;

Plateaux-Quenu, 1978; Plateaux-Quenu, 1985) have been shown to be important factors in determining which bee dominates oviposition if the nest is orphaned. The line of succession in H. farinosus is not determined by size since replacement queens were smaller than each of their worker nestmates in two of the three nests where we could compare sizes (Figure 3-2). In addition, replacement queens did not appear to be older than their fellow workers based upon wing and mandible wear scores. In fact, replacement queens found in nests with other workers (n = 5) always had the lowest wing wear among their sisters and four of the nine replacement queens had virtually no wing wear (score of 0 or 1) indicating that they had done little, if any, foraging.

Replacement queens might simply arise based on which bee displays the most queen­ like and least worker-like behaviour at the time of orphaning (Michener, 1990).

Sex ratio and sex allocation

71 The overall sex ratio in the reproductive brood in this population is very close to

50% but the sex ratio is split with significant differences between queenright and orphaned nests (p = 0.0004, t = 4.218, df = 22). Queens produced the majority of females (96.6%) and the workers produced most of the males in this population (60.8%).

All replacement queens (N = 9) were mated and their spermathecae after mating resembled that of the queens', so inability to produce diploid gynes is an unlikely explanation for male biased sex ratios in orphaned nests. Similarly, Mueller (1991) found that queenright colonies of Augochlorella aurata (=striata) were significantly more female biased than queenless colonies. Split sex ratios are also known from Halictus rubicundus where eusocial (queen containing) nests produced mostly females and parasocial (orphaned) nests produced mostly males (Yanega, 1989). However, in

Halictus ligatus queens produced most of the males while abandoned colonies headed by replacement queens produced mostly gynes (Richards et al., 1995). According to

Richards et al. (1995), queens in H. ligatus lay as many male destined eggs as possible before workers are able to coerce them into producing gynes, and orphaned workers continue to produce females presumably to balance the overall sex ratio. While the reproductive brood of H.farinosus tended to be protandrous, its queens produce few males before switching to gyne production when workers are present. Theory predicts extreme differences in sex ratios between nests where the cost to produce one sex or the other differs (Grafen, 1986). This is predicted to occur to the point where different nest types specialize in producing one sex or the other where their sex investment

72 strategy will depend on the social structure of the family (Grafen, 1986; Boomsma,

1991). This is likely the cause of the extreme sex and investment ratios in queenright nests in this population which are highly female biased with an average investment ratio of nearly 1:5 (males: females) in monogynous queenright nests. No relatedness asymmetry is present when replacement queens produce daughters and sons (or nieces and nephews in the case of subordinate workers). Thus, when given the choice, workers in orphaned nests should prefer to produce whichever sex has the best chance of reproduction (Fisher, 1930; Trivers and Hare, 1976; Queller and Strassman, 1998) and given the sex ratio asymmetry of the queen's brood, workers in orphaned nests should prefer to produce males. This pattern is precisely what we see in this population with queenright colonies (type 1and 2) specializing in producing females and replacement queen colonies (type 3) specializing in producing males.

Conclusion

The results of this study agree with the predictions of inclusive fitness theory: monandry and high relatedness among nestmates in a primitively eusocial organism.

Nestmates were close kin as a result of a high incidence of monogamous foundresses and low effective mating frequency. Despite comparatively small queen to worker size disparity (Chapter 2) and high levels of worker ovary development and matedness, H. farinosus queens still produced the vast majority of offspring while they were alive. Only

73 when a nest had been orphaned did first brood female workers make a significant contribution to the reproductive brood. Overall the workers were as closely related to the reproductive brood as the queen. Workers may capitalize on the relatedness asymmetries presented by haplodiploidy by coercing the queen into producing females while producing males themselves. The fact that queenright nests specialized in producing females and orphaned nests headed by replacement queens produced mostly males indicates that haplodiploidy may be playing a role in the evolutionary maintenance of the helping behaviour by ensuring that working females are maximally related to the brood in all nest types.

74 Tables and Figures

Table 3-1: Primers for microsatellite loci (developed by Soro and Paxton, 2009) and variability in all individuals of H. farinosus. Including observed and expected heterozygosities, the inbreeding coefficient (Fis) and hardy-weinberg estimations for all loci. M13 tails were added to the forward primer for incorporation of a fluorescent dye during PCR. Z o

Name Primer Sequence (5'- Anneali Length n He or H-W 3') ng Range alleles test p temp.f (bp) value C) Rub02 f : TGTAAAACGACGGCCAGT 58 183-211 15 0.893 0.881 0.0133 0.082 CCAGCCGGCCAACGTTGC R:CGGAGCTGAAAACTCAATTACA G Rub04 F: TGTAAAACGACGGCCAGT 58 207-215 4 0.576 0.619 -0.0754 0.068 CGGACGTTTTTCAATGTTTTTC R: CGTCCGACTGCATTCTCTTTG Rub55 F: TGTAAAACGACGGCCAGT 58 144-172 14 0.834 0.852 0.0464 0.407 GCTATAAAAGGCGAAACGGGTG R:CTCCTATCCGGTTGACATTGCC Rub30 F: TGTAAAACGACGGCCAGT 65 168-197 14 0.846 0.921 -0.0885 0.121 GATCCGCnTCAACCGTCCG R: 6TGAGCTGGGTCCGGCGAG Rub 59 F: TGTAAAACGACGGCCAGT 60 203-232 10 0.785 0.790 -0.0042 0.458 GT6ACCAGGTGCGCTCGTTAC R: CCGTGTCCCCAGCTCCGTTrC Rub73 F: TGTAAAACGACGGCCAGT 60 191-244 14 0.811 0.781 0.0377 0.662 GCTTTGTTTCTCACTATCGTCCC R: CGCGCAAAGTTCCCAGGGGTG

75 Table 3-2: Estimates of relatedness between different categories of offspring in the worker and reproductive broods, given separately for monogamous nests. Expected relatedness is the predicted value is for a nest where there is a sole, singly mated, foundress who produces all offspring.

Relationship (y, x) Observed Expected SE 95% CI N relatedness relatednes s Worker Brood Queen to worker 0.523 0.50 0.043 0.110 47 Queen to male 1.000 1.00 0.000 0.000 14 Worker to worker 0.644* 0.75 0.022 0.056 74 Worker to male 0.745 0.50 0.137 0.353 10 Reproductive brood Queen to worker 0.446 0.50 0.026 0.067 53 Queen to male 0.794* 1.00 0.023 0.060 134 Queen to gyne 0.474 0.50 0.025 0.064 246 Worker to worker 0.628* 0.75 0.032 0.082 68 Worker to male 0.550 0.50 0.029 0.076 240 Worker to gyne 0.594* 0.75 0.040 0.104 263 Gyne to gyne 0.652* 0.75 0.024 0.061 496 Monogamous nests only Queen to worker 0.530 0.50 0.035 0.090 54 Queen to maleA 1.000 1.00 0.000 0.000 49 Queen to gyne 0.495 0.50 0.026 0.067 137 Worker to worker 0.720 0.75 0.014 0.037 74 Worker to maleA 0.578 0.75 0.039 0.099 54 Worker to gyne 0.749 0.75 0.024 0.062 111 •Significant difference rom expected by confidence interval limits includes reproductive brood individuals only

76 Table 3-3: Summary of kin structure in the population's reproductive brood nests where nests are divided into types.

Queen Non queen- Reproductiv presen produced Origin of non queen- Reproductiv N Type e type t brood produced brood eskew nests 1 Monogynous Y» n/a n/a Full (S = 1.0) 12 2 Polygynous Y Males worker produced High (S > 0.76) 3 Males and 3 Polygynous N females worker produced Highly variable 9 Males and unknown/cannot be 4 Polygynous N females determined Unknown 3 * While it is generally true that the queen is present, she was missing in two nests found to be completely monogynous

77 Table 3-4: Sociobiological details in all nests based on genotype, sibship patterns and genetic relatedness data.

Worker brood Nests

Nest Que Mon Monan N N alien Proporti Reproducti Sex Ratio Propor Queen' Notes en ogyn dry queens worker on of ve skew * (males/ tiori of s Pres V worker offspring total) female effecti ent laid by sin ve queen worker mating offspri freque ng ncy N070 Y Y Y n/a n/a 1.000 1.000 0.00 n/a 1 2-1 N070 Y Y N n/a n/a 1.000 1.000 0.00 n/a 1.39 2-2 N070 Y Y Y n/a n/a 1.000 1.000 0.00 n/a 1 4-3 N070 Y Y Y n/a n/a 1.000 1.000 0.00 n/a 1 6-1 N070 Y Y n/a n/a n/a 1.000 1.000 1.00 n/a n/a »» 8-2 N070 Y Y n/a n/a n/a 1.000 1.000 0.83 n/a n/a ** 9-4 N071 Y Y Y n/a n/a 1.000 1.000 0.25 n/a 1 2-1 N071 Y Y N n/a n/a 1.000 1.000 0.00 n/a 1.80 2-3 N071 Y Y Y n/a n/a 1.000 1.000 0.00 n/a 1 5-1 N072 Y Y Y n/a n/a 1.000 1.000 0.00 n/a 1 5-1 Reproductive brood nests Nest Que Mon Monan Queen Allen Proportio Reproducti Sex Ratio Propor Queen' Type, en ogyn dry worker worker nof ve skew* (males/t tk>n of s Notes Pres * (queen offspring otal) female effect) ent ) laid by sin ve queen worker mating offspri freque ng ncy N081 NA N Y 3 0 0.800 0.837 0.50 0.000 1 3,"* 1-1 N081 Y N Y 2 0 0.769 0.759 0.82 0.000 1 2 1-3 OrlOb Y Y Y 1 0 1.000 1.000 0.00 n/a 1 1

Or4b Y Y Y 1 0 1.000 1.000 0.33 n/a 1 1

B101 N N Y 5 3 0.000 0.915 0.98 0.021 1 3,***

B106 Y N Y 1 0 0.957 0.909 0.26 0.000 1 2

B107 Y N Y 5 0 0.969 0.984 0.12 1.000 1 2

B108 Y Y Y 4 1 1.000 1.000 0.22 n/a 1 1

B211 Y Y Y 1 0 1.000 1.000 0.11 n/a 1 1

B301 N N n/a 1 0 0.80 n/a 4,#

78 B302 NA N Y 1 0 0.400 0.273 0.56 0.167 1 3, *** B304 Y Y Y 2 0 1.000 1.000 0.32 n/a 1 1 B306 Y Y Y 2 0 1.000 1.000 0.14 n/a 1 1 B311 N N n/a 1 0 0.75 n/a 4,# B311- N N n/a 1 0 0.67 n/a 4,# 2 B313 NA Y Y 2 0 1.000 1.000 0.86 n/a 1 1 C208 NA N Y 2 0 0.875 0.903 0.13 1.000 1 3 D303 NA Y Y 3 0 1.000 1.000 0.25 n/a 1 1 D401 N N n/a 1 0 0.000 1.000 l.OO 0.000 n/a 3, ** E401 N N Y 2 0 0.000 1.000 0.82 0.182 n/a 3,** E407 Y Y Y 0 0 1.000 1.000 0.29 n/a 1 1 E4H Y Y N 1 0 1.000 1.000 0.16 n/a 1.80 1 F203 Y Y Y 2 0 1.000 1.000 0.40 n/a 1 1 F205 N N n/a 1 0 0.000 1.000 1.00 0.000 n/a 3,** F211 Y Y Y 0 0 1.000 1.000 0.00 n/a 1 1

G402 NA N N 1 0 0.545 0.016 0.70 0.000 1.39 3

G408 NA N Y 2 0 0.000 1.000 1.00 0.000 1 3 * From equation S3 in Crozier and Pamilo, 1996

•• All males or only one female - n patrilines cannot be determined

*•* More than one worker has produced offspring. In 0811-1, C208 and B302 the first replacement queen was not found, one worker (second replacement queen) was the mother of the youngest offspring.

# Kin structure cannot be determined

A Queen's alleles reliably deduced

79 1.25

*0.75 J—I- T30> m w 0.5 u S s0.25 (9

ZZZZZZZZZO O CO CO CO CO CD OOOOOOClCDWOOIimm-n 7! O O ooooooooo^7t;st;s;N OQOl-'>-»l->OoOO'-'PUWWUWWNU)^^^^ ti OOOMMPKJMPOCTMO^JOOH K)5o>h>l-'U)00U)t-'vlMW •-*£i oNJ o00 w4»cnroMUiuippO' (Nj -0.25 MMWHHWHI-'HW Nest

Figure 3-1: Genetic relatedness among H. farinosus females (excluding the queen) in each nest with 95% confidence intervals, calculated by Relatedness 5.0.8 (Queller and Goodknight, 1989).

80 0.25

c £ 0.2 £ 1 | o 0.15 I_ T3* 0.1 wII 01 41 • Head width = s 0.05 • Wing length oi aj e E §Ol fl)8 it Q. BlOl E401 5 2 -0.05 fl

-0.1 Nest

Figure 3-2: Size differences (mm) between replacement queens and the average size of the nests remaining workers. Only three nests (BlOl, E401, G408) had an identifiable replacement queen who was present with other workers. The replacement queen was generally smaller than the remaining workers.

81 References

Abbot, P., et al. 2010. Inclusive fitness theory and eusociality. Nature 471: El - E4.

Ascher, J. S., Pickering, J. 2011. Bee species guide (Hymenoptera: Halictidae). [Internet] http://www.discoverlife.org/mp/20q ?guide=Halictus

Bartomeus, I., Ascher, J. S., Wagner, D., Danforth, B. N., Colla, S., Kombluth, S., Winfree, R. 2011. Climate-associated phenological advances in bee pollinators and bee-pollinated plants. Proceedings of the National Academy of Sciences 108(51): 20645 - 20649.

Birmingham, A. L., Hoover, S. E., Winston, M. L., Ydenberg, R. C. 2004. Drifting bumble bee (Hymenoptera: Apidae) workers in commercial greenhouses may be social parasites. Canadian Journal of Zoology 82:1843-1853.

Boomsma, J. J. 2009. Lifetime monogamy and the evolution of eusociality. Philosophical Transactions of the Royal Society 364:3191 - 3207.

Bourke, A.F.G. 1988. Worker reproduction in the higher eusocial Hymenoptera. Quarterly Review of Biology 63: 291 - 311.

Bourke, A. F. G., Franks, N. R. 1995. Social Evolution in Ants. Princeton, NJ: Princeton University Press.

Brady, S. G., Sipes, S., Pearson, A., Danforth, B. N. 2006. Recent and simultaneous origins of eusociality in halictid bees. Proceedings of the Royal Society 273:1643 -1649.

Cameron, S. A., Mardulyn, P. 2001. Multiple molecular data sets suggest independent origins of highly eusocial behaviour in bees (Hymenoptera: Apinae). Systematic Biology 50 (2): 194-214.

Cardinal, S., Danforth, B. N. 2011. The antiquity and evolutionary history of social behaviour the bees. PLOS one 6(6): e21086.

Chapuisat, M., Keller, L. 1999. Testing kin selection with sex allocation data in eusocial Hymenoptera. Heredity 82:473 - 478.

Coelho, B. W. T. 2002. The biology of the primitively eusocial Augochloropsis iris (Schrottky, 1902) (Hymenoptera, Halictidae). Insectes Sociaux 49:181 -190.

82 Cook, J. M., Crozier, R. H. 1995. Sex determination and population biology in the Hymenoptera. Trends in Ecology and Evolution 10(7): 281- 286.

Cornwallis C. K., West, S. A., Davis, K. E., Griffin, A. S. 2010. Promiscuity and the evolutionary transition to complex societies. Nature 466: 969 - 972.

Crespi, B. J. 1992. Eusociality in Australian gall thrips. Nature 359:724 - 726.

Crozier, R. H. 2008. Advanced eusociality, kin selection and male haploidy. Australian Journal of Entomology 47: 2 - 8.

Crozier, R, H., Pamilo, P. 1996. Reproductive skew simplified. Oikos 75 (3): 533 - 535.

Crozier, R. H., Smith, B. H.; Crozier, Y. C. 1987. Relatedness and population structure of the primitively eusocial bee Lasioglossum zephyrum (Hymenoptera, Halictidae) in Kansas. Evolution 41:902 - 910.

Danforth, B. N. 2002. Evolution of sociality in a primitively eusocial lineage of bees. Proceedings of the National Academy of Sciences 99: 286 - 290.

Danforth, B. N., Sauquet, H., Packer, L. 1999. Phylogeny of the bee genus Halictus (Hymenoptera: Halictidae) based on parsimony and likelihood analyses of nuclear EF-la sequence data. Molecular Phylogenetics and Evolution 13: 605 - 618.

Danforth, B. N., Conway, L., Ji, S. 2003. Phylogeny of eusocial Lasioglossum reveals multiple losses of eusociality within a primitively eusocial clade of bees (Hymenoptera: Halictidae). Systematic Biology 52: 23 - 36.

Darwin, C., 1859. On the origin of species by means of natural selection. London, UK: John Murray.

Downs, S. G., Ratnieks, F. L. W. 1999. Recognition of conspecifics by honeybee guards uses nonheritable cues acquired in the adult state. Animal Behaviour 58:643 - 648.

Eickwort, G. C. 1985. The nesting biology of the sweat bee Halictus farinosus in California with notes on H. ligatus (Hymenoptera: Halictidae). Pan-Pacific Entomologist 61:122-137.

Eickwort, G. C., Eickwort, J. M., Gordon, J., Eickwort, M. A. 1996. Solitary behaviour in a high altitude population of the social sweat bee Halictus rubicundus (Hymenoptera: Halictidae). Behavioural Ecology and Sociobiology 38:227 - 233.

83 Fisher, R. A. 1930. The genetical theory of natural selection. Oxford: Oxford University Press.

Frank, S. A. 1998. Foundations of Social Evolution. Princeton, NJ: Princeton University Press.

Gadagkar, R. 1990. Origin and evolution of eusociality: A perspective from studying primitively eusocial wasps. Journal of Genetics 69 (2): 113 -125.

Gardiner, A., Foster, K. R. 2008. Evolution and ecology: history and concepts. In Ecology of Social Evolution (Korb, J., Heinze, J. eds.) Springer- Verlag, Berlin, p. 1-36.

Gardiner, A., West, S. A., Barton, N. H. 2007. The relation between multilocus population genetics and social evolution theory. The American Naturalist 169 (2): 207 -226.

Gathmann, A. Tscharntke, T. 2002. Foraging ranges of solitary bees. Journal of Animal Ecology 71: 757 - 764.

Gibbs, J., Albert, J., Packer, L. 2012. Dual origins of social parasitism in North American Dialictus (Hymenoptera: Halictidae) confirmed using a phylogenetic approach. Cladistics 28:195-207.

Grafen, A. 1986. Split sex ratios and the evolutionary origins of eusociality. Journal of Theoretical Biology 122:95 -121.

Hamilton, W. D., 1964a. The genetical evolution of social behavior: I. Journal of Theoretical Biology 7:1-17.

Hamilton, W. D., 1964b. The genetical evolution of social behavior: I. Journal of Theoretical Biology 7:17 - 52.

Hamilton, W. D. 1972. Altruism and related phenomena, mainly in social insects. Annual Review of Ecology and Systematics 3:193 - 232.

Hammond, R. L., Keller, L. 2004. Conflict over male parentage in social insects. PLOS Biology 2:1472 -1482.

Hines, H. M., Hunt, J. H., O'Connor, T. K., Gillespie, J. J., Cameron, S. A. 2007. Multigene phylogeny reveals eusociality evolved twice in vespid wasps. Proceedings of the National Academy of Sciences 104:3296 - 3299.

84 Hirata, M., Higashi, S. 2008. Degree-day accumulation controlling allopatric and sympatric variations in the sociality of sweat bees, Lasioglossum (Evylaeus) baleicum (Hymenoptera: Halictidae). Behavioural Ecology and Sociobiology 62:1239-1247.

Hughes, W. H. O., Olroyd, B. P., Beekman, M., Ratnieks, F. L. 2008. Ancestral monogamy shows kin selection is key to the evolution of eusociality. Science 320:1213 -1216.

Jacob, S., McClintock, M. K., Zelano, B., Ober, C. 2002. Paternally inherited HLA alleles are associated with women's choice of male odor. Nature Genetics 30:175 -179.

Johnstone, R. A. 2000. Models of reproductive skew: A review and synthesis. Ethology 106: 5 - 26.

Johnstone R.A., Cant M.C. 1999. Reproductive skew and the threat of eviction: a new perspective. Proceedings: Biological Sciences 266:275 - 79.

Jones, O. R., Wang, J. 2010. COLONY: a program for parentage and sibship inference from multilocus genotype data. Molecular Ecology Resources 10:551- 555.

Keller, L. 1997. Indiscriminate altruism: unduly nice parents and siblings. Trends in Ecology and Evolution 12:99 -103.

Knerer, G. 1980. Biology and social behaviour of bees of the genus Halictus (Hymenoptera: Halictidae). Zoologische Jahrbuecher Abteilung fuerSystematik Oekologie und Geographie der Tiere 107:511 - 536.

Linksvayer, T. A., Wade, M. J. 2005. The evolutionary origin and elaboration of sociality in the aculeate hymenoptera: maternal effects, sib-social effects, and heterochrony. Quarterly Review of Biology 80:317 - 336.

Lopez-Vaamonde, C., Koning, J. W., Brown, R. M., Jordan, W. C., Bourke, A. F. G. 2004. Social parasitism by male-producing reproductive workers in a eusocial insect. Nature 430: 557-560.

Manning, C. J., Wakeland, E. K., Potts, W. K. 1992. Communal nesting patterns implicate MHC genes in kin recognition. Nature 360:581-583.

Mateo, J. M. 2002. Kin recognition abilities and nepotism as a function of sociality. Proceedings: Biological Sciences 269: 721- 727.

85 Maynard Smith, J., 1964. Group selection and kin selection. Nature 201 (4924): 1145 - 1147.

McLeish, M. J. Chapman, T. M., Crespi, B. J. 2006. Inbreeding ancestors: the role of sibmating in the social evolution of gall thrips. Journal of Heredity 97:31 - 38.

McLeish, M. J., Chapman, T. M. 2007. The origins of soldiers in gall-inducing thrips of Australia (Thysanoptera: Phlaeothripidae). Australian Journal of Entomology 46: 300- 304.

Mehdiabadi, N. J., Jack, C. N., Farnham, T. T., Piatt, T. G., Kalla, S. E., Shaulsky, G., Queller, D.C., Strassman, J. S. 2006. Kin preference in a social microbe. Nature 442:881 -882.

Michener, C.D. 1974. The Social Behavior of Bees. The Belknap Press of Harvard University Press, Cambridge, Massachusetts.

Michener, C. D. 1990. Reproduction and castes in social halictine bees. In: Social Insects: An Evolutionary Approach to Castes and Reproduction (Engels, W., Buschinger, A. Eds.) Springer-Verlag pp. 75 -119.

Michener, C.D. 2007. The Bees of the World. John's Hopkins University Press, Baltimore, Maryland.

Michener, C. D., Grimaldi, D. A. 1988. The oldest fossil bee: Apoid history, evolutionary stasis, and antiquity of social behaviour. Proceedings of the National Academy of Sciences USA 85: 6424 - 6426.

Michener, C. D., Brothers, D. J., Kamm, D. R. 1971. Interactions in colonies of primitively social bees: Artificial colonies of Lasioglossum zephyrum. Proceedings of the National Academy of Sciences USA 68:1241-1245.

Minckley, R. L., Wcislo, W. T., Yanega, D., Buchmann, S. L. 1994. Behaviour and phenology of a specialist bee (Dieunomia) and sunflower (Helianthus) pollen availability. Ecology 75(5): 1406 -1419.

Moilanen A., Sundstrom, L., Pedersen, J. S. 2004. MATESOFT: a program for deducing parental genotypes and estimating mating system statistics in haplodiploid species. Molecular Ecology Notes 4: 795 - 797.

86 Moreau, C. S., Bell, C. D., Vila, R., Archibald, B., Pierce, N. E. 2006. Phylogeny of the ants: diversification in the age of angiosperms. Science 312:101 -104.

Mueller, U. G. 1991. Haplodiploidy and the evolution of facultative sex ratios in a primitively eusocial bee. Science 254:442 - 444.

Mueller U. G., Eickwort GC, Aquadro CF. 1994. DNA fingerprinting analysis of parent offspring conflict in a bee. Proceedings of the National Academy of Sciences USA 91: 5143-5147.

Muralidharan, K., Shaila, M. S., Gadagkar, R. 1986. Evidence of multiple mating in the primitively eusocial wasp Ropalidia marginata (Lep.) (Hymenoptera: Vespidae). Journal of Genetics 65 (3): 153 -158.

Nanork, P., Paar, J., Chapman, N. C., Wongsiri, S., Oldroyd, B. P. 2005. Asian honeybees parasitize the future dead. Nature 437:829.

Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583 - 590.

Nowak, M. A., Tarnita, C. E., Wilson, E. O. 2010. The evolution of eusociality. Nature 446: 1057 -1062.

Nye, W. P. 1980. Notes on the biology of Halictus (Halictus) farinosus Smith (Hymenoptera: Halictidae). U S Department of Agriculture Science and Education Administration Agricultural Research Results 11:1- 29.

Ordway, E. 1965. Caste differentiation in Augochlorella (Hymenoptera, Halictidae). Insectes Sociaux 12: 291- 308.

Ozaki, M., Wade-Katsumata, A., Fujikawa, K., Iwasaki, M., Yokohari, F. Satoji, Y., Nisimura, T., Yamaoka, R. 2005. Ant nestmate and non-nestmate discrimination by a chemosensory sensillum. Science 309:311 - 314.

Packard, A. S. 1868. The home of the bees (Concluded). The American Naturalist 1(11): 596-606.

Packer, L. 1986. The social organisation of Halictus ligatus (Hymenoptera; Halictidae) in southern Ontario. Canadian Journal of Zoology 64:2317 - 2324.

87 Packer, L. 1990. Solitary and eusocial nests in a population of Augochlorella striata (Provancher) (Hymenoptera: Halictidae) at the northern edge of its range. Behavioural Ecology and Sociobiology 27:339 - 344.

Packer, L. 1992. The social organization of Lasioglossum (Dialictus) laevissimum in southern Alberta. Canadian Journal of Zoology 70:1767 -1774.

Packer, L. 1997. The relevance of phylogenetic systematic to biology. Examples from medicine and behavioural ecology. Memoirs du Museum National d'Histoire Naturelle special volume: The origin of biodiversity in insects: phylogenetic tests of evolutionary scenarios. P. Grandcolas (Ed.) 173:11- 29.

Packer, L., Knerer, G. 1985. Social evolution and its correlates in bees of the subgenus Evylaeus (Hymenoptera: Halictidae). Behavioral Ecology and Sociobiology. 17:143 -149.

Packer, L., Owen, R. E. 1994. Relatedness and sex ratio in a primitively eusocial halictine bee. Behavioral Ecology and Sociobiology 34:1 -10.

Packer, L., Sampson, B., Lockerbie, C., Jessome, V. 1989. Nest architecture and brood mortality in four species of sweat bee (Hymenoptera: Halictidae) from Cape Breton Island. Canadian Journal of Zoology 67:2864 - 2870.

Pamilo, P. 1991. Evolution of colony characteristics in social insects. I. Sex allocation. The American Naturalist 137 (1): 83 -107.

Pamilo, P., Crozier, R. H. 1996. Reproductive skew simplified. Oikos 75(3): 533 - 535.

Parker, F. D. 1981. Sunflower pollination: abundance, diversity, and seasonality of bees and their effect on seed yields. Journal of Apical Research 20:49 - 61.

Parker, F.D. 1982. Efficiency of bees in pollinating onion flowers.Journal of the Kansas Entomological Society 55:171 -175.

Paxton, R. J., Thoren, P. A., Tengo, J., Estoup, A., Pamilo, P. 1996. Mating structure and nestmate relatedness in a communal bee, Andena jacobi (Hymenopter, Andrenidae), using microsatellites. Molecular Ecology 5: 511- 519.

Paxton, R. J., Ayasse, M., Field, J., Soro, A. 2002. Complex sociogenetic organization and reproductive skew in a primitively eusocial sweat bee, Lasioglossum malachurum, as revealed by microsatellites. Mol. Ecol. 11:2405 -16.

88 Pereboom, J. J. M., Biesmeijer, J. C. 2003. Thermal constraints for stingless bee foragers: the importance of body size and coloration. Oecologia 137:42-50.

Pesenko, Y. A. 1984. A subgeneric classification of bees of the genus Halictus sensu- stricto (Hymenoptera: Halictidae). Entomologicheskoe Obozrenie 62: 340 - 357.

Plateaux-Quenu, C. 1978. Les sexues de remplacement chez Evylaeus calceatus (Scop.). Insectes Sociaux 25:227 - 236.

Plateaux-Quenu, C. 1983. Le volume d'un pain d'abeille influence-t-il le sexe de I'oeuf pondu sur lui? Etude experimental portant sur la premiere couvee d'Evylaeus calceatus (Scop.) (Hym., Halictinae). Annates Des Sciences Naturelles Zoologie 13 (5): 41 - 52.

Plateaux-Quenu, C. 1985. Seconde couvee d'Evylaeus calceatus (Scop.) (Hym., Halictinae) - les fondatrices sont-elles seules capable d'engendrer des fondatrices? Annates Des Sciences Naturelles Zoologie 13 (7): 13 - 21.

Queller, D. C. 1992. Quantitative genetics, inclusive fitness, and group selection. The American Naturalist 139 (3): 540 - 558.

Queller, D. C. 1994. Genetic relatedness in viscous populations. Evolutionary Ecology 8: 70-73.

Queller, D. C., Goodknight, K. F. 1989. Estimating relatedness using genetic markers. Evolution 43 (2): 258 - 275.

Queller, D. C. Strassmann, J. E. 1998. Kin selection and social insects. Bioscience 48:165 -175.

Ratnieks, F. L. W. Foster, K. R., Wenseleers, T. 2006. Conflict resolution in insect societies. Annual Reviews in Entomology 51:581- 608.

Reeve H.K., Emlen ST, Keller, L. 1998. Reproductive sharing in animal societies: reproductive incentives or incomplete control by dominant breeders? Behavioural Ecology 9: 267-278.

Reeve, H. K., Keller, L. 2001. Tests of reproductive skew models in social insects. Annual Reviews in Entomology 46: 347 - 385.

Richards, M. H. 2001. Nesting biology and social organization of Halictus sexcinctus (Fabricius) in southern Greece. Canadian Journal of Zoology 79:2210 - 2219.

89 Richards, M. H. 2004. Annual and social variation in foraging effort of the obligately eusocial sweat bee, Halictus ligatus (Hymenoptera: Halictidae). Journal of the Kansas Entomological Society 77 (4): 484 - 502.

Richards, M.H., Packer, L. 1994. Trophic aspects of caste determination in Halictus ligatus, a primitively eusocial sweat bee. Behavioural Ecology and Sociobiology 34: 385 - 391.

Richards, M.H., Packer, L. 1995. Annual variation in survival and reproduction of the primitively eusocial sweat bee Halicus ligatus (Hymenoptera: Halictidae). Canadian Journal of Zoology 73:933 - 941.

Richards, M. H., Packer, L. 1996. The socioecology of body size variation in the primitively eusocial sweat bee, Halictus ligatus (Hymenoptera: Halictidae). Oikos 77(1): 68-76.

Richards, M. H., Packer, L., Seger, J. 1995. Unexpected patterns of parentage and relatedness in a primitively eusocial bee. Nature 373:239 - 241.

Richards, M. H., von Wettberg, E. J., Rutgers, A. C. 2003. A novel social polymorphism in a primitively eusocial bee. Proceedings of the National Academy of Science USA 100 (12): 7175-7180.

Richards, M. H., Vickruck, J. L., Rehan, S. M. 2010. Colony social organisation of Halictus confusus in Southern Ontario, with comments on Sociality in the Subgenus H. (Seladonia). Journal of Hymenoptera Research 19:144 -158.

Roulston, T. H., Cane J. H. 2002. The effect of pollen protein concentration on body size in the sweat bee Lasioglossum zephyrum (Hymenoptera: Apiformes). Evolutionary Ecology 16:49 - 65.

Rousset, F. 2008. GENEPOP'007: a complete re-implementation of the GENEPOP software for Windows and Linux. Molecular Ecology Resources 8(1): 103 -106.

Russell, A. F., Hatchwell, B. J. 2001. Experimental evidence for kin-biased helping in a cooperatively breeding vertebrate. Proceedings of the Royal Society of London 268: 2169 - 2174.

Sakagami, S., Michener, C. D. 1962. The nest architecture of the sweat bees (Halictinae): a comparative study of behaviour. University of Kansas Press, Lawrence, Kansas.

90 Sakagami, S. F., Munakata, M. 1972. Distribution and bionomics of a transpalaeartic eusocial halictine bee Lasioglossum (Evylaeus) calceatum in northern Japan, with reference to its solitary life cycle at high altitude. Jour. Fac. Sci. Hokkaido Univ. Ser. VI, Zool. 18(3): 411-439.

Schmitz, J., Mortiz, R. F. A. 1998. Molecular phylogeny of Vespidae (Hymenoptera) and the evolution of sociality in wasps. Molecular Phylogenetics and Evolution 9:183 -191.

Schuelke, M. 2000. An economic method for fluorescent labeling of PCR fragments. Nature Biotechnology 18:233 - 234.

Schwarz, M. P. 1994. Female-biased sex ratios in a facultatively social bee and their implications for social evolution. Evolution 48:1684 -1697.

Schwarz, M. P., Richards, M. H., Danforth, B. N. 2007. Changing paradigms in insect social evolution: insights from halictine and allodapine bees. Annual Review of Entomology 52:127 -150.

Sellars, R. 2004. The social organization of Halictus (Halictus) farinosus Smith in northern Utah (Masters Thesis, York University).

Seppa, P., Fogelqvist, J., Gyllenstrand, N., Lorenzi, M. C. 2011. Colony kin structure and breeding patterns in the social wasp, Polistes biglumis. Insectes Sociaux 58: 345 - 355.

Sharp, S. P., McGowan, A., Wood, M. J., Hatchwell. B. J. 2005. Learned kin recognition cues in a social bird. Nature 434:1127 -1130.

Sheffield, C. S., Kevan, P. G., Pindar, A., Packer, L. 2012. Bee (Hymenoptera: Apoidea) diversity within apple orchards and old fields in the Annapolis Valley, Nova Scotia. The Canadian Entomologist. Forthcoming.

Sitdikov, A. A. 1987. Nesting of the bee Halictus quadricinctus (F.) (Hymenoptera, Halictidae) in the udmurt ASSR. Entomologicheskoye Obozreniye 3:529 - 539.

Soro, A., and Paxton, R. J. 2009. Characterization of 14 polymorphic microsatellite loci for the facultatively eusocial sweat bee Halictus rubicundus (Hymenoptera, Halictidae) and their variability in related species. Molecular Ecology Resources 9:150 -152.

91 Soro A., Ayasse, M., Zobel, M. U., Paxton, R. J. 2009. Complex sociogenetic organization and the origin of unrelated workers in a eusocial sweat bee, Lasioglossum maiachurum. Insectes Sociaux 56:55 - 63.

Soro, A., Field, J., Bridge, C., Cardinal, S. C., Paxton, R. J. 2010. Genetic differentiation across the social transition in a socially polymorphic sweat bee, Halictus rubicundus. Molecular Ecology 19:3351 - 3363.

Soucy, S. L. 2002. Nesting biology and socially polymorphic behaviour of the sweat bee Halictus rubicundus (Hymenoptera: Halictidae). Annals of the Entomological Society of America 95:57 - 65.

Soucy, S., Danforth, B. N. 2002. Phylogeography of the socially polymorphic sweat bee Halictus rubicundus (Hymenoptera: Halictidae). Evolution 56(2): 330 - 341.

Starr C.K. 1984. Sperm competition, kinship, and sociality in the aculeate Hymenoptera. In: Sperm Competition and the Evolution of Animal Mating Systems (Smith R.L. Ed). Academic Press, New York, USA. pp 427 - 464.

Sundstrom, L., Chapuisat, M., Keller L. 1996. Conditional manipulation of sex ratios by ant workers: a test of kin selection theory. Science 274:993 - 995.

Taylor, P. D. 1992. Altruism in viscous populations - an inclusive fitness model. Evolutionary Ecology 6:352 - 356.

Trivers, R. L., Hare, H. 1976. Haplodiploidy and the evolution of social insects. Science 191: 249-263.

Ulrich, Y., Perrin, N., Chapuisat, M. 2009. Flexible social organization and high incidence of drifting in the sweat bee, Halictus scabiosae. Molecular Ecology 18:1791-1800.

Utah State University climate centre website. 2011. http://climate.usurf.usu.edu/products/data.php

Wang, J. 2004. Sibship reconstruction from genetic data with typing errors. Genetics 166:1963 -1979.

West, S.A., Pen, I., and Griffin, A.S. 2002. Cooperation and competition between relatives. Science 296:72 - 75.

92 West, S. A., Griffin, A. S., Gardner, A. 2007a. Social semantics: Altruism, cooperation, mutualism, strong reciprocity and group selection. Journal of Evolutionary Biology 20: 415-432.

West, S. A., Griffin, A. S., Gardner, A. 2007b. Evolutionary explanations for cooperation. Current Biology 17(16): 661- 672.

Wilson, D. S., Pollock, G. B., Dugatkin, L. A. 1992. Can altruism evolve in purely viscous populations? Evolutionary Ecology 6:331- 341.

Wilson, D. S., Wilson, E. O. 2007. Rethinking the theoretical foundation of sociobiology. The Quarterly Review of Biology 82(4): 327 - 348.

Wilson, E. 0.1971. The Insect Societies. Cambridge MA. The Belknap Press of Harvard University Press.

Wilson, E. O., 1975. Sociobiology: the new synthesis. Cambridge, MA. Harvard University Press.

Wilson, E. 0.2005. Kin selection as the key to altruism: its rise and fall. Social research 72:159-166.

Wilson, E. O. 2008. One giant leap: how insects achieved altruism and colonial life. Bioscience 58 (1): 17 - 25.

Wilson, E. O., Holldobler, B. 2005. Eusociality: origin and consequences. Proceedings of the National Academy of Sciences 102(38): 13367 -13371.

Yanega, D. 1988. Social plasticity and early diapausing females in a primitively social bee. Proceedings of the National Academy of Sciences 85:4374 - 4377.

Yanega, D. 1989. Caste determination and differential diapauses within the first brood of Halictus rubicundus in New York (Hymenoptera: Halictidae). Behavioural Ecology and Sociobiology 24:97 -107.

Zalom, F. G., P. B. Godell, L. T. Wilson, W. W. Barnett, and W. J. Bently. 1983. Degree days: the calculation and use of heat units in pest management. Division of Agriculture and Natural Resources, University of California, Leaflet 21373.

93 This is to advise you that as the representative of the Dean of Graduate Studies of York and after having reviewed the text and verified the insertion of the last corrections suggested by the members of the thesis committee during the thesis examination, I authorize Naba Al Najjar to file her thesis entitled "LE VICE MONSTRUEUX, LE MONSTRE VERTUEUX: une analyse narrative et sociale de Justine, ou les malheurs de/a vertude Donatien Alphonse Frangois de Sade (1740-1814)".

I am forwarding this message to the two thesis supervisors as well as to the student, whom I congratulate for this very impressive thesis.

Marie-Christine Pioffet