Bee genetics and conservation Amro Zayed

To cite this version:

Amro Zayed. Bee genetics and conservation. Apidologie, Springer Verlag, 2009, 40 (2), ￿10.1051/apido/2009026￿. ￿hal-00892012￿

HAL Id: hal-00892012 https://hal.archives-ouvertes.fr/hal-00892012 Submitted on 1 Jan 2009

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Apidologie 40 (2009) 237–262 Available online at: c INRA/DIB-AGIB/EDP Sciences, 2009 www.apidologie.org DOI: 10.1051/apido/2009026 Review article

Bee genetics and conservation*

Amro Zayed

Institute for Genomic Biology, University of Illinois at Urbana-Champaign, 1206 West Gregory Drive, Urbana, 61801, USA

Received 29 August 2008 – Revised 5 January 2009 – Accepted 15 January 2009

Abstract – The emerging threat of pollinator decline has motivated research on bee conservation biology in order to both understand the causes of declines and to develop appropriate conservation strategies. The ap- plication of genetics to the conservation of diploid animals has proven to be important for both overcoming genetic threats to population viability and for providing tools to guide conservation programs. However, the haplodiploid bees have several unusual genetic properties of relevance to their conservation, which warrant special attention. Here I review how haplodiploidy and complementary sex determination affect genetic parameters pertinent to the viability and future evolutionary potential of bee populations. I also review how genetic tools can improve the conservation management of bees. I find that bees are especially prone to extinction for genetic reasons, and that genetics can provide invaluable tools for managing bee populations to circumvent pollinator decline. haplodiploid / complementary sex determination / inbreeding depression / diploid males / extinction

1. INTRODUCTION purpose of examining the causes of pollinator decline and developing appropriate conserva- There are nearly 20,000 known species of tion strategies to circumvent it (e.g. Dias et al., bees worldwide (Michener, 2000), and they 1999;Kremenetal.,2007; NRC, 2007; Byrne are integral components of terrestrial ecosys- and Fitzpatrick, 2009). tems due to their indispensable role as pol- linators (Allen-Wardell et al., 1998; Klein Several extrinsic factors have been pro- et al., 2007;Kremenetal.,2007). Over the posed to explain the observed declines in past decade, multiple lines of evidence have bee populations: habitat fragmentation and shown that both native and managed pollina- loss, agricultural intensification, overuse of tors are experiencing a seemingly world-wide pesticides, pathogen spill-over from managed decline (Biesmeijer et al., 2006; Fitzpatrick pollinators, invasive species, and global cli- et al., 2006;Kosioretal.,2007; NRC, 2007; mate change (Cane, 2001; Cane and Tepedino, Colla and Packer, 2008;Goulsonetal.,2008; 2001;Kremenetal.,2002; Müller et al., Grixti et al., 2009), fueling both ecological 2006; NRC, 2007;Murrayetal.,2009). and economic concerns (Kremen et al., 2002; Genetic aspects of bee declines were com- Fontaine et al., 2006;Vamosietal.,2006; pletely neglected in the early syntheses (e.g. Olesen et al., 2007;Pauw,2007). This emerg- Dias et al., 1999) despite the overwhelming ing threat has mobilized the academic commu- evidence that genetic factors can play im- nity through the establishment of international portant roles in species extinction (Saccheri research initiatives and partnerships with the et al., 1998; Frankham et al., 2002;Spielman et al., 2004; Hanski and Saccheri, 2006). Corresponding author: A. Zayed, Several factors likely contributed to the ini- [email protected] tial oversight of possible genetic aspects of * Manuscript editor: Robert Paxton bee declines, including: (1) Prior theoretical

Article published by EDP Sciences 238 A. Zayed arguments suggesting that genetics is rela- plausible and best supported genetic threats to tively unimportant in the conservation biol- the viability of small bee populations: Com- ogy of haplodiploids (Box I). (2) Difficulties in plementary sex determination, inbreeding de- extrapolating knowledge gained from diploid pression, and loss of genetic diversity and con- organisms to the haplodiploid bees. (3) Lack sequent evolutionary potential. I do not discuss of studies examining how specific bee genetic the hypothesized effects of the accumulation and life history traits (e.g. haplodiploidy, com- of deleterious mutations on population vi- plementary sex determination, sociality, etc.) ability (Lynch et al., 1995), given limited impact population viability. (4) Lack of ge- empirical support of ‘mutational meltdowns’ netic resources for bees. However, theoretical in sexually-reproducing populations (reviewed and technical advances over the past decade by Frankham, 2005). Box I contains a glos- have overturned longstanding views regard- sary of terms relevant to bee conservation ge- ing the immunity of bees to genetic threats in netics. The terminology relating to inbreeding small populations and prompted research on is historically a very murky subject (Jacquard, their conservation genetics. 1975; Templeton and Read, 1994; Keller and Research in conservation genetics can be Waller, 2002). Following Glémin (2003)and conceptually divided into two general types Leberg and Firmin (2008), here I use ‘sys- of inquiry: (1) How does genetics contribute tematic inbreeding’ to describe species with to population decline and extinction? (2) How mating systems characterized by ‘inbreeding’ can molecular tools be utilized to learn about (Box I). I use inbreeding by drift to de- an organism’s characteristics relevant to con- scribe the increasing probability of relatedness servation and management (e.g. taxonomy, among mates in small populations (Box I). The natural history, and demography)? In this re- term inbreeding depression (Box I) can also view, I summarize and synthesize studies on be ambiguous. Defined broadly, inbreeding de- how genetics can contribute to bee declines pression, or reduced fitness of inbred individ- (Sect. 2), focusing on how haplodiploidy and uals, can encompass the effects of comple- genetic sex determination affect parameters mentary sex determination that are unique to relevant to population viability and evolution- some members of the Hymenoptera. However, ary potential. I also review how genetic and inbreeding depression is classically ascribed genomic tools can be used in conservation to dominance and overdominance (Box I) – and management of bee populations (Sect. 3). which are common to both diploids and hap- To ensure that my review is accessible to lodiploids. Here I treat the unique effects of the largest possible readership, I focus on complementary sex determination on the fit- presenting general concepts without delving ness of haplodiploids as distinct from inbreed- into detailed mathematical treatments. For the ing depression caused by dominance and over- mathematically inclined, I recommend several dominance. reviews on haplodiploid population genetics (Crozier, 1977; Hedrick and Parker, 1997), and conservation genetics (Frankham et al., 2002; 2.1. Complementary sex determination Gaggiotti, 2003; Hedrick, 2004; Frankham, and bee declines: death by diploidy 2005). Bees belong to the insect order Hy- menoptera, a group characterized by hap- 2. GENETIC ASPECTS OF BEE lodiploidy (i.e. Females are diploid while DECLINES males are usually haploid; Box I). The pre- sumed ancestral sex determination mechanism Genetic threats can reduce the viability of in the Hymenoptera is single locus comple- small wildlife populations over the short term mentary sex determination (sl-CSD; Box I), by reducing fitness, as well as over the long where sex is determined by genotype at a term by limiting evolutionary potential (Box I) single gene (Cook, 1993; van Wilgenburg and adaptability. Here I discuss the three most et al., 2006; Heimpel and de Boer, 2008). Bee conservation genetics 239

Ayabe et al., 2004; Liebert et al., 2004, 2005), with one rare exception in an aculeate wasp (Cowan and Stahlhut, 2004). In such cases, the production of diploid males indirectly in- creases female mortality by constraining their mates to the production of haploid males and triploid daughters. Therefore, diploid male production increases female mortality over one or two generations, given inviable or ef- Figure 1. Diploid male production. A matched mat- fectively sterile diploid males respectively. Ul- ing results when a female mates with a male shar- timately, increased female mortality caused by ing a sex-determining allele in common (allele E). diploid male production will result in reduced Half of the female’s diploid progeny – all intended population growth rates in bee populations to be female – will be homozygous at the sex deter- mination locus and will develop into diploid males (Stouthamer et al., 1992; Pamilo and Crozier, (genotype EE). 1997). In social bees, diploid male production will effectively increase female mortality for both reproductive and worker castes, and can Heterozygotes at the sex-determination locus thus reduce both population and colony growth develop into diploid females from fertilized rates (Cook and Crozier, 1995). In cases where eggs, while hemizygotes develop into hap- colony survival is a function of the size of the loid males from unfertilized eggs. In cases worker force, diploid male production can sig- where a female mates with a haploid male nificantly increase colony mortality (Plowright that shares a sex-determining allele in com- and Pallett, 1979; Ross and Fletcher, 1986). mon (i.e. matched mating; Box I) half of the Diploid male production is thus expected to be female’s diploid progeny will be homozygous costly in both solitary and social bees. at the sex-determination locus and will con- The number of sex-determining alleles, sequently develop into diploid males instead which indirectly controls the frequency of of females (Fig. 1). Diploid males have been diploid male production, is an increasing func- observed in 4 families and at least 27 species tion of population size (Fig. 2). Mutation pro- of both solitary and social bees (Tab. I), and vides a constant but slow (e.g. 10−6,Kerr, the sex determination locus has been molec- 1997) input of novel sex-determining alleles ularly identified in the honey bee Apis mellif- into the gene pool, increasing genetic diversity era (Beye et al., 2003) and genetically mapped at the sex-determination locus. Given the fit- in the bumble bee Bombus terrestris (Gadau ness costs associated with the production of et al., 2001). homozygotes at the sex-determination locus, The production of diploid males represents which develop into diploid males, the locus a large genetic cost in bee populations. Diploid experiences strong negative frequency depen- males are mostly either inviable or sterile dent selection (Box I): New sex-determining (Agoze et al., 1994; Duchateau et al., 1994; alleles introduced by mutation will initially Holloway et al., 1999; Liebert et al., 2004, rise in frequency since individuals carrying 2005; Heimpel and de Boer, 2008). Female them have higher relative fitness (i.e. they Hymenoptera only fertilize their eggs when are less likely to participate in matched- attempting to produce daughters. The pro- matings). However, as selection increases the duction of diploid males from fertilized eggs frequency of the new allele, relative to pre- therefore acts to increase female mortality. A existing alleles, the relative fitness of indi- secondary cost to diploid male production is viduals carrying the new allele will diminish also incurred if diploid males are viable and as they participate in more matched-matings. achieve matings. Viable diploid males produce Negative frequency dependent selection there- diploid sperm and thus females mating with fore acts to homogenize the frequencies of sex- them produce inviable fertilized eggs or ster- determining alleles in a population: The equi- ile triploid daughters (Krieger et al., 1999; librium allele frequency at the sex-determining 240 A. Zayed

Table I. Bee species where diploid males have been reported.

Species Reference Family Andrenidae Andrena scotica (Paxton et al., 2000) Family Halictidae Augochlorella striata (Packer and Owen, 1990) Halictus poeyi (Zayed and Packer, 2001) Lasioglossum leucozonium (Zayed et al., 2007) Lasioglossum zephyrum (Kukuk and May, 1990) Family Megachilidae Megachile rotundata (McCorquodale and Owen, 1994) Family Apidae Apis cerana (Woyke, 1979) Apis mellifera (Adams et al., 1977; and others) Bombus atratus (Plowright and Pallett, 1979) Bombus florilegus (Takahashi et al., 2008) Bombus impatiens Zayed et al. unpublished Bombus muscorum (Darvill et al., 2006) Bombus sylvarum (Ellis et al., 2006) Bombus terrestris (Duchateau et al., 1994) Eufriesea magrettii (Lopez-Uribe et al., 2007) Euglossa imperialis (Zayed et al., 2004) Euglossa mandibularis (Takahashi et al., 2001) Euglossa meriana (Roubik et al., 1996) Euglossa piliventris (Lopez-Uribe et al., 2007) Euglossa sapphirina (Roubik et al., 1996) Euglossa tridentata (Roubik et al., 1996) Eulaema cingulata (Lopez-Uribe et al., 2007) Melipona compressipes (Kerr, 1987) Melipona scutellaris (Carvalho, 2001) Melipona quadrifasciata (Camargo, 1979) Scaptotrigona postica (Paxton et al., 2003) Trigona carbonaria (Green and Oldroyd, 2002) locus is 1/k,wherek represents the number 1/k. The frequency of diploid male pro- of sex-determining alleles in the population duction is thus equivalent to the equilib- (Adams et al., 1977; Yokoyama and Nei, 1979; rium frequency of sex-determining alleles, Owen and Packer, 1994). In a random mat- and thus diploid male production is inversely ing population, the frequency of diploids that proportional to the number of sex-determining are male, measured as the proportion of ho- alleles in the population (Fig. 2). Ultimately, mozygotes at the sex-determination locus, is the number of sex-determining alleles in a k(1/k)2(i.e. number of sex-determining alle- population is a function of mutation, nega- les multiplied by the expected homozygos- tive frequency dependent selection, and ge- ity of an allele at equilibrium frequency) = netic drift (Yokoyama and Nei, 1979; Cornuet, Bee conservation genetics 241

cal population sizes and growth rates, the pro- duction of diploid males increased the risk of extinction by more than an order of magnitude on average (Zayed and Packer, 2005). The pro- duction of effectively sterile diploid males in- creased the extinction risk over that caused by inviable diploid males, because of the former’s secondary effects on population growth rates. In diploid species, inbreeding depression is the biggest genetic threat to short-term popula- tion viability (Brook et al., 2002b;Spielman et al., 2004). Zayed and Packer (2005)also Figure 2. Smaller populations maintain less alle- compared the probability of extinction caused les at the sex determination locus, and consequently by inbreeding depression in diploid popula- produce higher frequencies of diploid males, when tions with that caused by diploid male produc- compared to larger populations. The graph was ob- tion in haplodiploid populations. Given identi- tained assuming a mutation rate of 10−6 using Cor- cal starting population parameters, the risk of nuet’s (1980) mutation-selection-drift equilibrium extinction caused by diploid male production model. in haplodiploid populations was an order of magnitude higher than that caused by inbreed- ing depression in diploid populations, mak- 1980). Smaller populations are more affected ing complementary sex determination in hap- by drift and are thus expected to maintain lodiploids the largest known genetic threat to fewer sex-determining alleles, and have higher population viability (Zayed and Packer, 2005; frequencies of diploid male production com- Hedrick et al., 2006). pared to larger populations (Fig. 2). Zayed and Packer (2005) attributed this in- Complementary sex determination and creased extinction risk to the effects of the diploid male production are expected to have “Diploid Male Vortex”. The production of a negative impact on the population viabil- diploid males can initiate a positive feed- ity of bees. The production of diploid males back cycle that leads to rapid extinction. is expected to be higher in smaller popula- This occurs because the production of diploid tions (Fig. 2), and is thus expected to reduce males initially reduces population sizes and growth rates in these populations (Stouthamer growth rates. Demographic and environmental et al., 1992; Pamilo and Crozier, 1997). Zayed stochasticity combined with increased genetic and Packer (2005) explored the relationship drift in a smaller population further reduce the between diploid male production and extinc- number of sex-determining alleles which leads tion using simulation models of population to higher levels of diploid male production. viability. Population viability analysis is a This diploid male vortex will continue to re- very useful framework to examine how deter- duce the proportion of females in the popula- ministic and stochastic events impact extinc- tion, further reducing growth rates. In isolated tion risk (Brook et al., 2002a). Briefly, Zayed populations with small sizes and or low repro- and Packer (2005) modelled hypothetical hap- ductive rates, negative population growth rates lodiploid populations assuming a wide range can be rapidly achieved, ultimately leading to of population sizes, and growth rates. First, extinction. simulations were conducted assuming no com- Although it is difficult to attribute specific plementary sex determination to estimate the causes to population declines and extinction in risk of extinction due to demographic factors nature, several lines of evidence suggest that alone. Then, simulations incorporating com- diploid male production can contribute, and plementary sex determination were performed has contributed, to bee declines: to examine the effects of diploid male produc- 1. Higher levels of extinction are observed tion on extinction risk. Given initially identi- in laboratory populations of parasitoid wasps 242 A. Zayed with versus without complementary sex deter- Carvalho’s (2001) study demonstrates a direct mination (Stouthamer et al., 1992;Wuetal., causal link between diploid male production, 2003), as expected from Zayed and Packer’s and extinction in bees. (2005) simulations. Furthermore, when labo- The above lines of evidence suggest that ratory populations of parasitoid wasps go ex- diploid male production can be, and likely tinct, they usually do so with high male-biased has been, involved in the global decline of sex ratios (Simmonds, 1947), as predicted by bees. Habitat loss and fragmentation are ex- Zayed and Packer’s (2005) simulations. pected to reduce levels of genetic variation 2. The simulation models developed by in natural populations (Gilpin, 1991; Hedrick Zayed and Packer (2005) show that diploid and Gilpin, 1997; Whitlock and Barton, male production can bring-about extinction 1997), including allelic variation at the sex- over realistic demographic parameters for determination locus, making smaller bee pop- bees. In simulations, the net reproductive out- ulations more susceptible to extinction through put of female bees was an important pre- the diploid male vortex. Furthermore, any dictor for their susceptibility to extinction extrinsic factor that acts to reduce bee pop- through the diploid male vortex. Many soli- ulation growth rates (e.g. overuse of pesti- tary bee species have very low life-time fe- cides, pathogen/parasite infections, competi- cundity (Michener and Rettenmeyer, 1956; tion from invasive species) will also, through Danks, 1971;Elseetal.,1978; Minckley et al., the dependence of diploid male production 1994; Franzén and Larsson, 2007). Franzén on population size, initiate the diploid male and Larsson (2007) recently summarized the vortex (Fig. 3). Bee populations are thus ex- reproductive output of several Andrena species pected to decline faster, and recover slower, which averaged 5.8 eggs per lifetime. At these than expected based on ecological predictions levels, the risk of extinction caused by the alone due to the synergistic negative effects diploid male vortex are extremely high for of diploid male production on population size a small population unless the environment is and growth (Fig. 3). conducive to rapid population growth (Zayed Stochastic models supporting the role of and Packer, 2005). diploid male production in bee declines have 3. The production of diploid males has been assumed monandry (i.e. females singly mate), empirically shown to increase colony mortal- random mating, and no adaptations to re- ity for some social Hymenoptera, both in the duce the costs of diploid male production laboratory (Plowright and Pallett, 1979)andin (Zayed and Packer, 2005). The first two as- the field (Ross and Fletcher, 1986). sumptions are well supported by empirical 4. Diploid males have been found in data (Eickwort and Ginsberg, 1980; Estoup many declining and/or endangered solitary and et al., 1995; Strassmann, 2001; Green and social bee species (Carvalho, 2001; Zayed Oldroyd, 2002;Palmeretal.,2002;Paxton et al., 2004; Darvill et al., 2006; Ellis et al., et al., 2003; Cameron et al., 2004;Paxton, 2006; Takahashi et al., 2008). For example, 2005; Beveridge et al., 2006; Kraus et al., diploid males, and triploid queens have been 2008), with some exceptions (e.g. Page, 1980; observed in small populations of the rare Paxton et al., 2000). Polyandry is expected and locally distributed bumble bee Bombus to reduce between-family variation in the florilegus (Takahashi et al., 2008). In the production of diploid males when compared stingless bee Melipona scutellaris, small iso- to monandry. However, the total frequency lated populations maintain low numbers of of diploid male production will remain un- sex-determining alleles and rapidly go extinct changed (Cook and Crozier, 1995). Lower (Carvalho, 2001), presumably due to the ef- variance in diploid male frequencies between fects of the diploid male vortex. However, families should slightly reduce the effects of these populations can be ‘rescued’ through the drift (Frankham, 1995a), although similar ex- intentional introduction of mated queens from tinction risks should be experienced under other populations, which increases the number both polyandry and monandry given equal of sex-determining alleles (Carvalho, 2001). population-wide frequencies of diploid male Bee conservation genetics 243

random mating (Eickwort and Ginsberg, 1980) – is expected to increase diploid male produc- tion and the risk of extinction over the pre- dictions of Zayed and Packer’s (2005) mod- els that assume random mating. For example, 50% of matings between siblings given mo- nandry will be matched even in a population with a large number of sex-determining alle- les. Hymenopteran species with mating sys- tems characterized by systematic inbreeding tend not to have sl-CSD (van Wilgenburg et al., 2006; Heimpel and de Boer, 2008), and this may be also expected of bees with similar mat- ing systems. However, sl-CSD has never been plausibly rejected in any bee species exam- ined so far (van Wilgenburg et al., 2006), al- though the available data are somewhat lim- ited in taxonomic scope with a predominance of studies on the Family Apidae (Tab. I). Fre- quencies of diploid males estimated from sam- ples of adults in the systematically inbreed- ing Andrena scotia were lower than expected given sl-CSD, suggesting that this species may have multiple-locus CSD or a related adapta- tion to reduce levels of diploid male produc- Figure 3. Complementary sex determination syn- tion (Paxton et al., 2000). However, a lower ergistically interacts with extrinsic factors resulting frequency of diploid males, sampled as adults, in faster decline and slower recovery of bee popula- does not necessarily rule out sl-CSD given tions. I stochastically modelled (Zayed and Packer, the potential for high pre-adult mortality of 2005) a bee population (N = 10 000 bees) experi- diploid males (Cook and Crozier, 1995;van encing negative growth rates (assumed r = −0.28) Wilgenburg et al., 2006; Heimpel and de Boer, caused by an extrinsic environmental factor. At gen- 2008). eration 50, the extrinsic factor was removed allow- Finally, it has been proposed that hy- = . ing the population to recover (assumed r 0 28). menopterans with sl-CSD should evolve adap- The population size (a), averaged over 100 simula- tations to reduce the cost of diploid male pro- tion iterations, and the probability of extinction (b), duction (Cowan and Stahlhut, 2004;Paxton, are presented for populations without diploid male 2005; van Wilgenburg et al., 2006). Such (DM) production, as well as with DM production assuming inviable or effectively sterile DMs. The adaptations include the evolution of more sex production of diploid males reduced growth rates loci, restored fertility of diploid males, se- below that caused by the extrinsic factor resulting lective fertilization of non-matched sperm in in faster decline and slower recovery (a). polyandrous females, avoidance of matched matings through sex-allele signaling, and oth- ers (reviewed by Cook and Crozier, 1995; production. Polyandry in social bees may van Wilgenburg et al., 2006). Evidence for serve to reduce the cost of diploid male pro- most of these hypotheses is either rare (Cowan duction on colony mortality, but this largely and Stahlhut, 2004 report reproductive diploid depends on the shape of the relationship be- males in an aculeate wasp; De Boer et al., tween colony fitness and diploid male produc- 2007 report multiple locus CSD in a parasitoid tion (Cook and Crozier, 1995). Systematic In- wasp), or completely lacking (Cook and breeding (Box I) – a rare phenomenon in bees Crozier, 1995; van Wilgenburg et al., 2006). In which have several adaptations that promote the largely random-mating bees, it is unclear if 244 A. Zayed adaptations reducing the cost of diploid male small populations (Frankham, 1995b, 1998; production would arise in normally large pop- Saccheri et al., 1998; Westemeier et al., 1998; ulations where diploid males are rare and the Hedrick and Kalinowski, 2000; Brook et al., costs of diploid male production are low. Em- 2002b; Gaggiotti, 2003; Reed and Frankham, pirical work is needed to investigate the possi- 2003;Spielmanetal.,2004; Frankham, 2005; bility that bees have adaptations which reduce Vilas et al., 2006). the cost of diploid male production, and theo- The exact cause of inbreeding depression retical work is needed to derive conditions un- is often debated among evolutionary biolo- der which such adaptations can arise. The lim- gists (Charlesworth and Charlesworth, 1999; ited evidence suggests that such adaptations do Hedrick and Kalinowski, 2000; Crnokrak and not exist, or are not widespread. Barrett, 2002; Keller and Waller, 2002; Leberg In summary, there is compelling theoreti- and Firmin, 2008). Two major hypotheses can cal and empirical evidence that diploid male mechanistically explain inbreeding depres- production can substantially reduce the via- sion. The dominance hypothesis (Box I) posits bility of small bee populations. The diploid that inbreeding depression is caused by the ex- male vortex is also expected to synergistically pression of deleterious recessive alleles nor- interact with other deterministic factors caus- mally sheltered in heterozygous individuals. ing faster extinction rates than would be ex- The overdominance (Box I) hypothesis states pected. Endangered bee species targeted for that inbreeding depression is caused by lower conservation should be managed to reduce fitness of homozygous versus heterozygous frequencies of diploid male production. This genotypes. Inbreeding increases the frequency can be achieved by promoting gene flow be- of homozygous genotypes, and reduced fit- tween populations (Zayed, unpubl. data), and ness of inbred individuals can be explained attempting to maintain a high level of allelic by both the dominance and overdominance diversity at the sex determination locus in bee hypotheses. Empirical data support the view meta-populations. that, generally, inbreeding depression results mostly from dominance with a minor contribu- tion from overdominance (Charlesworth and 2.2. Haplodiploidy, inbreeding Charlesworth, 1999; Hedrick and Kalinowski, depression, and bee declines 2000; Keller and Waller, 2002), although overdominance can be a major component Inbreeding depression (i.e. reduction in fit- of inbreeding depression for some traits ness due to inbreeding) is considered the (Charlesworth and Charlesworth, 1999). major threat to the short term viability of In haplodiploids, inbreeding depression can small populations of diploid organisms (Brook result from dominance, overdominance, and et al., 2002b; Gaggiotti, 2003;Spielmanetal., from increased homozygosity at the sex deter- 2004; Frankham, 2005). In small closed pop- mination locus (see Sect. 2.1). Here I ignore ulations, inbreeding by drift is an unavoid- complementary sex determination as a source able phenomenon as relatedness between in- of inbreeding depression in haplodiploids dividuals is expected to increase over time since classically, inbreeding depression is de- (Hedrick, 2000; Frankham et al., 2002). In- fined as reductions in fitness of inbred in- breeding increases the frequency of homozy- dividuals due to dominance and overdomi- gous genotypes and this is often associated nance only. Several attempts have been made with a reduction in fitness traits (Hedrick, to theoretically predict the relative extent 2000; Frankham et al., 2002). Inbreeding de- of inbreeding depression in haplodiploids pression has been observed for a large number (Crozier, 1976b, 1985; Werren, 1993; Hedrick of mostly diploid organisms (Ralls and Ballou, and Parker, 1997). In haplodiploids, reces- 1983; Ralls et al., 1988; Keller and Waller, sive lethal and mildly deleterious alleles are 2002; Armbruster and Reed, 2005)whereit constantly exposed to selection and are more has been both theoretically and empirically likely to be purged in haploid males, and shown to increase the risk of extinction in the equilibrium frequency of these alleles is Bee conservation genetics 245 therefore expected to be lower in haplodiploid plementary sex determination are appropriate versus diploid populations. If recessive dele- organisms to investigate the consequences of terious mutations are the sole cause of in- haplodiploidy on inbreeding depression with- breeding depression, then males, and females, out the confounding effects of complemen- in a haplodiploid population should exhibit tary sex determination. Henter (2003)and higher, and lower, genetic load than either sex Antolin (1999) reviewed estimates of in- of a diploid population, respectively (Werren, breeding depression in Hymenopteran para- 1993). Assuming the dominance hypothesis sitoids for several fitness traits. Both authors and an equal sex ratio, the genetic load of showed that haplodiploids without comple- haplodiploids should be on average approxi- mentary sex determination experience sub- mately 25% lower than diploid populations, stantial inbreeding depression, although lower although exact values depend on levels of than that experienced by diploids. Further- selection against deleterious alleles, degree more, even some haplodiploid species with of dominance, and whether systematic in- systematically-inbreeding mating systems still breeding is practiced (Werren, 1993; Hedrick experience inbreeding depression (Luna and and Parker, 1997). Deleterious mutations in Hawkins, 2004;Schremphetal.,2006). These genes with effects limited only to females results indicate that haplodiploidy does not of- are not expected to be purged by selection fer complete immunity from inbreeding de- on haploid males. The genetic load for sex- pression, and that both overdominance and limited genes should thus be substantial in fe- dominance in female-limited genes likely con- male haplodiploids (Crozier, 1976b), but will stitute a significant source of inbreeding de- still be 25% lower than the sex-limited load pression in haplodiploids. Inbred haplodiploid of a diploid population (Werren, 1993). Fi- populations, including those of bees, are thus nally, haplodiploidy is not expected to re- expected to suffer from reduction in fitness due duce the inbreeding depression caused by to inbreeding depression. overdominance. The effects of haplodiploidy on inbreeding depression are clearly depen- The effects of inbreeding depression were dent on many important, but often unknown, not incorporated into Zayed and Packer’s parameters which include the genetic basis (2005) population viability models, which of inbreeding depression (dominance versus showed the potential of high rates of extinction ff overdominance), the deleterious e ects and due to the production of diploid males in small relative dominance of mutations (e.g. lethal bee populations. Given the theoretical expec- versus mild deleterious, recessive versus par- tations and empirical evidence, it is likely that tially recessive), and the contribution of sex- inbreeding depression can interact synergisti- limited mutations to the genetic load. Al- cally with the sex determination load (Box I) though it seems clear that haplodiploidy can to increase extinction risk over the already reduce the genetic load under many circum- high values predicted by Zayed and Packer’s stances, the magnitude of the expected re- (2005) models. Modelling the effects of in- ductions are small, implying that haplodiploid breeding depression on population viability populations should still experience significant is possible following methods established for inbreeding depression. diploid organisms (Lacy, 1993; Brook et al., Lowered fitness of inbred individuals has 2002b), although parametrizing such models been documented in several social bees would be difficult given lack of empirical data (e.g. Bienefeld et al., 1989;Gerloff and for bees. Future studies which attempt to quan- Schmid-Hempel, 2005). However, most stud- tify inbreeding depression in inbred but not ies of inbreeding depression in bees con- match-mated bees (i.e. eliminating the con- found the negative effects of diploid male founding effects of the sex-determination load) production with those caused by dominance will provide important empirical data needed and overdominance, and only the latter two for a formal examination of the effects of contribute to inbreeding depression as classi- inbreeding depression on the viability of bee cally defined. Parasitoid wasps without com- populations. 246 A. Zayed

2.3. Haplodiploidy, genetic diversity, esized to reduce Ne by limiting both the and bee declines population density and dispersal opportuni- ties of specialist when compared to generalist Genetic diversity is needed for popula- bees (Packer et al., 2005). The hypothesized tions to adapt to their changing environments lower Ne of haplodiploid versus diploid in- (Fisher, 1930; Lynch and Walsh, 1998). The sects, social versus solitary bees, and special- population’s effective size, Ne (Box I), is an ist versus generalist bees have been supported important determinant of standing levels of by comparative studies of genetic diversity. genetic diversity. Populations with small Ne Haplodiploid insects – mostly hymenopter- experience stronger drift, and consequently ans – have less neutral genetic variation when maintain less genetic variation than those with compared to Drosophila (Hedrick and Parker, large Ne. As a result, small populations are be- 1997) and Lepidopteran insects (Packer and lieved to have limited potential evolutionary Owen, 2001), while eusocial bees tend to ex- responses to future changes in their environ- hibit lower levels of neutral genetic variation ment (e.g. to novel pathogens, pesticides and when compared to solitary bees (Packer and contaminants, introduced species, habitat frag- Owen, 2001). However, both of the previ- mentation, climate change, etc.) thereby in- ously mentioned meta-analyses do not correct ff creasing their long-term risk of extinction (re- for phylogeny (e.g. systematic di erences be- viewed by Frankham et al., 2002; Gaggiotti, tween hymenopterans and lepidopterans, other 2003; Hedrick, 2004; Frankham, 2005). Since than haplodiploidy, may be responsible for the ff a haplodiploid population will have less gene observed di erences), and their results should copies compared to an equivalent diploid pop- thus be interpreted with caution. Lower levels ulation, the former will have lower Ne (except of neutral genetic variation in specialist versus when sex ratios are extremely female biased, generalist bees have been documented in sev- see Crozier, 1976a). All other factors being eral studies that correct for the confounding ef- equal – including mutation rates – a hap- fects of phylogeny (Packer et al., 2005; Zayed lodiploid population is therefore expected to et al., 2005). have lower levels of neutral genetic variation Although there is some empirical support (Box I) than a diploid population. Comple- that haplodiploidy, eusociality, and oligolecty mentary sex determination further reduces the reduce Ne, this does not necessarily imply that already lower Ne of haplodiploids by reducing the haplodiploid bees in general, and euso- the number of females produced every gen- cial or specialist bees in particular, are more eration (i.e. due to diploid male production), at risk of extinction due to their lowered abil- and biasing the secondary sex ratio in favor ity to adapt to their changing environments. of haploid males when compared to an iden- The evolutionary potential of a population is tical haplodiploid population without sl-CSD best predicted by measuring levels of additive (Zayed, 2004). In addition to haplodiploidy genetic variance (Box I) in quantitative traits and complementary sex determination, sev- affecting fitness, a very difficult parameter to eral other ecological attributes of bees con- quantify in natural populations (Falconer and tribute to lower Ne (Packer and Owen, 2001), Mackay, 1996;Pfrenderetal.,2000; Reed and of which the best known are eusociality and Frankham, 2001; Gaggiotti, 2003; Hedrick, oligolecty. In eusocial bees, only a small pro- 2004). Neutral genetic variation often poorly portion of the population is actually reproduc- reflects additive genetic variance mostly due to tive (i.e. queens and males) thereby reducing the effects of selection and differences in mu- Ne by orders of magnitude over a comparable tation rates (Pfrender et al., 2000; Reed and solitary population with the same census size Frankham, 2001; Gaggiotti, 2003; Hedrick, (Pamilo and Crozier, 1997). Furthermore, the 2004). For example, the mutation input for a production of males by workers in some social polygenic trait should be larger than that ob- species also serves to increase drift and lower served at molecular markers. Further, variation genetic diversity (Owen, 1985). Oligolecty, at neutral markers is not expected to reflect or diet-specialization, has also been hypoth- variation in a quantitative trait where selection Bee conservation genetics 247

– instead of drift – may be the primary evolu- species characteristics relevant to wildlife con- tionary force. Therefore, lower neutral genetic servation management. In this section, I pro- variation in bees should not be taken as evi- vide a brief review of how molecular markers dence that bees are more at risk of extinction have been utilized to ascertain important pop- due to reduced evolutionary potential when ulation characteristics and processes in bees, compared to other taxa. and how conservation genetics can be used to Although there is no empirical evidence help overcome pollinator decline. suggesting that losses of adaptive genetic di- versity are contributing to the observed de- 3.1. Resolving taxonomic uncertainty clines in bee populations, the threat is both conceivable and warrants further investiga- In order to implement conservation actions, tion. For example, pathogens and parasites knowledge about the taxonomic status of the have been linked to declines of both man- targeted species is needed. It is obvious that aged and native bees (Colla et al., 2006; Cox- taxonomic uncertainty due to the presence of Foster et al., 2007; Otterstatter and Thomson, cryptic species can hinder conservation efforts 2008). In honey bees, there is a clear genetic (Frankham et al., 2002). Molecular markers basis for resistance to parasites (e.g. for re- have been previously used to resolve cryp- sistance to Varroa mites, Harbo and Harris, tic bee species (e.g. Blanchetot and Packer, 1999;LeConteetal.,2007). Several quan- 1992; Carman and Packer, 1997; Danforth titative trait loci (QTL) have been found to et al., 1998;Francketal.,2004; Kuhlmann control hygienic behavior of worker honey et al., 2007; Tavares et al., 2007;Murrayetal., bees (Lapidge et al., 2002) – a behavioral trait 2008). This is commonly achieved by observ- which limits parasite and pathogen loads in ing high DNA sequence divergence, or large a colony. Similarly, several QTLs have been genetic distances (Box I), between species found to affect parasite infection intensity and when contrasted against divergence/distance general immune response in the bumble bee measures within species. For example, in a Bombus terrestris (Wilfert et al., 2007a,b). survey of genetic diversity in the morpho- Therefore, it is plausible that loss of genetic logically distinct bee ‘species’ Halictus liga- variation at loci affecting pathogen / parasite tus from eastern North America, Carman and resistance and general immune response may Packer (1997) found fixed differences at more reduce a bee population’s ability to cope with than 7 out of 34 allozyme loci (Box I) be- infections from novel pathogens or parasites. tween populations in southern South Carolina Future research is clearly needed to estimate and Florida when compared to all other pop- levels of additive genetic variance in fitness- ulations, suggesting the presence of a cryptic related traits and examine its consequences on species (H. poeyi). This was later confirmed by population viability in natural bee populations. sequencing approximately 800 bp from several Experimental studies which attempt to directly mtDNA genes from a few individuals from estimate the effects of reduced adaptive diver- northern (H. ligatus) and southern (H. poeyi) sity on population viability after controlling populations: sequence divergence between H. for other confounding factors (e.g. inbreed- ligatus and H. poeyi exceed 4% while in- ing depression, and complementary sex deter- traspecific sequence divergence was very low mination) are particularly needed (e.g. Vilas (< 0.6%) (Danforth et al., 1998). et al., 2006). The application of molecular markers has helped in resolving taxonomic uncertainty of several endangered bees. For example, us- 3. USE OF GENETICS IN BEE ing a combination of allozymes, microsatel- CONSERVATION MANAGEMENT lite (Box I) and RAPD markers, Tavares et al. The application of molecular and popula- (2007) found high levels of genetic differenti- tion genetics to the field of conservation bi- ation between several geographic populations ology has provided a wealth of information of the endangered stingless bee Melipona ru- on fundamental population parameters and fiventris, indicating the presence of a cryptic 248 A. Zayed species which should be considered sepa- 1999). Indeed, the availability of microsatellite rately for conservation purposes. Similarly, data has stimulated the development of novel Quezada-Euán et al. (2007) surveyed genetic statistical methods aimed at characterizing a diversity at several microsatellite markers and species’ demography (e.g. population struc- 678 bp of the mitochondrial gene COI in the ture, detection of first generation migrants and endangered stingless bee Melipona beecheii admixed individuals; testing for recent popula- and found high levels of genetic differentiation tion size changes) (reviewed by Excoffier and between populations in the Yucatan peninsula Heckel, 2006). The growing availability of mi- and Costa Rica. Quezada-Euán et al. (2007) crosatellite markers combined with non-lethal recommended that movement of colonies be- techniques to sample DNA (Holehouse et al., tween southern Mexico and Central America 2003; Chaline et al., 2004) should open the should be reconsidered given that the two re- door for population genetic studies of com- gions may harbor two cryptic species. mon, declining, and even endangered bees. The discovery of a nearly universal DNA Population genetic surveys of natural bee barcode for animals (Hebert et al., 2003)isex- populations have almost always detected pected to greatly improve efforts to conserve biologically-significant levels of population bees. The DNA barcode represents a short structure. Packer and Owen (2001)reviewed (658 bp) sequence of the mitochondrial gene allozyme-based studies and found that bees, COI which shows high divergence between along with other Hymenoptera, tend to have closely related species of most animal taxa, significantly higher estimates of genetic differ- allowing for an efficient and nearly univer- entiation when compared to lepidopteran in- sal DNA identification system (Hebert et al., sects, suggesting that haplodiploidy promotes 2003; but see Moritz and Cicero, 2004). Con- genetic differentiation, possibly though its ef- servation workers can quickly identify a tar- fect on reducing effective population size. geted species by sequencing its DNA barcode More recent studies using microsatellites have and comparing it to a database of DNA bar- corroborated the earlier findings (e.g. Widmer codes from known taxa. COI has an estab- and Schmid-Hempel, 1999; Danforth et al., lished history of delineating closely related 2003; Herrmann et al., 2007;Stowetal., bee species (Danforth et al., 1998; Danforth, 2007; Zayed and Packer, 2007), with some 1999;Dicketal.,2004; Quezada-Euán et al., exceptions (Estoup et al., 1996; Beveridge 2007;Murrayetal.,2008), with some rare ex- and Simmons, 2006). Population genetic stud- ceptions (Kuhlmann et al., 2007). The applica- ies of endangered bumble bees also discov- tion of DNA barcoding has already proved to ered significant levels of genetic differentia- be useful in resolving uncertainty in the taxo- tion (Darvill et al., 2006; Ellis et al., 2006), nomically difficult Lasioglossum subgenus Di- even at small geographic scales (< 10 km). alictus (Gibbs, 2009), and a campaign to bar- The available evidence suggests that popula- code the bees of the world has been initiated tion genetic structure is a nearly universal fea- (see www.bee-bol.org). ture of bee populations. This may be expected since nest-building and central place foraging 3.2. Estimating species characteristics can promote population subdivision by reduc- relevant to conservation ing gene flow. Neutral molecular markers have been uti- Molecular markers have also been used lized to great effect in estimating several to estimate colony densities, foraging ranges, critical and often unknown demographic char- and effective population sizes for several acteristics of species targeted for conservation bee species. In a monoandrous haplodiploid (Frankham et al., 2002; Hedrick, 2004). Be- species, sisters are expected to be 75% related cause of their high variability, microsatellite to each other. If female bees are sampled as markers (Box I) have high statistical power they forage, it is possible to detect full sisters to estimate demographic parameters, and have based on their high degree of relatedness as thus emerged as the gold standard in popu- inferred from genotyping a large number of lation genetic surveys (Luikart and England, hypervariable markers such as microsatellites Bee conservation genetics 249

(Chapman et al., 2003). By estimating the with ecological studies, as the recent work number of colonies contributing foragers (i.e. by Herrmann et al. (2007) demonstrates. sisters) at any particular site, and by making Herrmann et al. (2007) sampled workers of some assumptions regarding the number of the common bumble bee Bombus pascuorum colonies not observed, it is possible to estimate from 13 sites that differed in land-use types the number and density of nests at the sampled in an agricultural landscape in Germany, and sites (Darvill et al., 2004). Also, the foraging the sampled workers were genotyped at 8 mi- range of a species can be estimated by exam- crosatellite loci. Hermann et al. (2007) were ining the geographic distances separating for- able to estimate important demographic and agers from the same colony (Chapman et al., genetic parameters for each site (e.g. num- 2003; Darvill et al., 2004). The above meth- ber of colonies, colony density, relative colony ods have been used to great effect in estimating size, estimates of genetic differentiation, and nest densities and foraging ranges for common inbreeding coefficients). The authors then used bumble bees (Chapman et al., 2003; Darvill generalized linear models to examine the re- et al., 2004; Knight et al., 2005). Sistership- lationship between demographic parameters, based methods, as outlined above, can also be genetic parameters, and land-use character- used to estimate the effective number of breed- istics. Herrmann et al. (2007) found that ing individuals in a population (equals 1.5X the proportion of mass flowering crops posi- the number of nests assuming monandry; Ellis tively affected bumble bee abundance, and that et al., 2006). positive inbreeding coefficients (i.e. higher fre- Although more data are clearly needed, es- quency of homozygotes than expected assum- timates of effective population sizes of bees ing Hardy-Weinberg equilibrium) had a neg- tend to be surprisingly low. Zayed and Packer ative effect on colony size. The latter result (2007), using changes in microsatellite allele represents strong evidence that increased ho- frequencies over time (Wang, 2005), estimated mozygosity adversely affects fitness of bee current Ne to range from 200 to 1000 for populations, and this effect is most likely me- several populations of the solitary sweat bee diated through the production of diploid males Lasioglossum (Sphecodogastra) oenotherae, (Sect. 2.1) and possibly inbreeding depres- a specialist on evening primroses. Using a sion (Sect. 2.2). Herrmann et al.’s (2007)work sistership-based method, Ellis et al. (2006) clearly demonstrates that the integration of estimated the breeding population size (i.e. population genetic and ecological studies is which provides an upper limit for Ne) of the likely to yield findings of great significance endangered bumble bee Bombus sylvarum to for examining the ecological and genetic pro- range 21 to 72. Low historical Ne, inferred cesses which underpin declines in bee popula- from allozymes, have also been observed for tions. the common sweat bee Halictus poeyi (Zayed and Packer, 2001), and the declining orchid bee Euglossa imperialis (Zayed et al., 2004). 3.3. Detecting declining bee species More estimates of Ne for solitary and social using genetic methods bees are badly needed before any broad con- clusions can be made, but the available data The first step to conserving a species in- seem to suggest that bees persist in somewhat volves demonstrating that it is on the de- isolated populations with comparatively small cline. Biodiversity data (presence/absence of effective sizes. Low Ne and limited gene flow species, or population census data) have been can exacerbate both genetic and demographic mostly used for detecting declines in bee pop- threats to population viability (see Sect. 2), ulations (Roubik, 2001; Williams et al., 2001; and conservation actions developed for bees Biesmeijer et al., 2006; Grixti and Packer, must be mindful of these characteristics. 2006; Colla and Packer, 2008;Grixtietal., Although population genetic surveys are 2009). However, natural variation in the abun- very insightful on their own, they be- dance of bees over space and time complicates come extremely illuminating when combined the detection of population decline (Roubik, 250 A. Zayed

2001; Williams et al., 2001). Population ge- diploid male production estimated from sam- netic data, on the other hand, reflect both the pling adults are unbiased, preliminary studies current state of the population as well as its of diploid male survival will be needed before history, making such data more resistant to field monitoring programs are established. the temporal and spatial variation found in The recent characterization of the sex- ecological datasets. There are several genetic determination locus in the honey bee (Beye approaches to detecting recent reductions in et al., 2003;Hasselmannetal.,2008) can also effective population sizes. One approach in- aid in developing ways to directly assess the volves detecting population bottlenecks. Bot- number of sex-determining alleles in a popu- tleneck tests rely on the fact that reductions in lation. In principal, PCR can be used to am- population size create some disparity among plify the sex-determination locus in samples the different ways of quantifying genetic diver- of bees, followed by cloning and sequencing sity (Cornuet and Luikart, 1996;Excoffier and of the amplified PCR products to examine se- Heckel, 2006). For example, following a bot- quence diversity at that locus. This approach tleneck, allelic richness is reduced faster when has one major advantage over detecting popu- compared to heterozygosity: An elevated het- lation declines through frequencies of diploid erozygosity compared to that expected based males, since the latter requires that diploid on allelic richness can indicate recent reduc- males be sampled in an unbiased way, which tions in population size (Cornuet and Luikart, may be difficult if they have low survivorship. 1996). Applying this method to microsatel- However, it remains to be demonstrated that lite data, bottlenecks were detected in sev- the sex-determination gene characterized in eral populations of two endangered and declin- the honey bee is sufficiently well-conserved to ing bumble bee species (Darvill et al., 2006; allow for constructing universal PCR primers Ellis et al., 2006). Using simulation-based ap- for routine amplification of the locus in other proaches, it is also possible to estimate past bee taxa. We also still lack knowledge on how population sizes, the magnitude of change in alleles are ‘functionally’ encoded at the sex- population size, and the relative timing of bot- determination locus, and thus the use of se- tlenecks (Excoffier and Heckel, 2006), provid- quence diversity as a surrogate for allelic di- ing potential clues regarding the causes of the versity at the sex determination locus may not declines. be appropriate. Further studies are needed to Another method of detecting declines of examine the utility of using PCR-based meth- bee populations involves using frequencies of ods to quantify allelic diversity at the sex de- diploid males (Zayed et al., 2004). The fre- termining gene, and its potential use in detect- quency of diploid males is a negative function ing declines in bee populations. of effective population size (Fig. 2; Sect. 2.1), The above approaches for genetically de- and declining populations are expected to tecting population declines can prove very use- produce higher frequencies of diploid males. ful in bee conservation management, however, Therefore, increased levels of diploid male they can sometimes be undermined by demo- production can theoretically indicate bee pop- graphic history. For example, if a population ulation declines (Zayed et al., 2004). Ge- experienced a drastic bottleneck in the past, netic monitoring of bee populations can be but has rapidly recovered since, both of the undertaken by first establishing baseline data proposed methods would suggest that the pop- of diploid male production in populations ulation is in decline. For example, the soli- of interest, followed by routine sampling of tary sweat bee Lasioglossum leucozonium and males – preferably through non-lethal DNA the bumble bee Bombus terrestris exhibit sig- extraction protocols – to measure changes in nificant signs of recent bottlenecks and high that parameter over time. This, of course, as- frequencies of diploid male production even sumes that bees are not able to adaptively though they are actually increasing in their in- avoid producing diploid males as the num- vasive ranges in North America (Zayed et al., ber of sex determining alleles declines (see 2007) and Tasmania (Schmid-Hempel et al., Sect. 2.1). Also, to ensure that frequencies of 2007) respectively. However, in both cases, Bee conservation genetics 251 the genetic data also indicated that the stud- can also be used in ‘forensic’ applications. ied populations were invasive. The risk of mis- For example, Cox-Foster et al. (2007)used diagnosis can further be reduced if the genetic pyrosequencing technology (Margulies et al., data are interpreted in light of the ecology and 2005; 454 Life Sciences) to sequence RNA natural history of the species in question. from honey bees (Apis mellifera) and royal Finally, the above mentioned approaches jelly found in healthy colonies, and colonies can be greatly leveraged by genetic analyses afflicted with colony collapse disorder (CCD). of ‘historic’ specimens from museum collec- By matching the generated sequences to pub- tions. Population genetic studies using spec- lic sequence databases, the study found that imens from museum collections can be used the presences of several pathogens, including to contrast historical levels of genetic diver- Israeli acute paralysis virus, were correlated sity and differentiation with present-day lev- with CCD. Although Cox-Foster et al.’s (2007) els (Wandeler et al., 2007). A recent study meta-genomic study is not statistically appro- has demonstrated that old bumble bee speci- priate for establishing cause and effect, it pro- mens (as old as 100 years) can be genotyped vides important knowledge which can be used at microsatellite loci using PCR (Strange et al., to develop and test causal hypotheses regard- 2009). Population genetic analyses of museum ing the role of pathogens in CCD of honey specimens have also proved useful in quantify- bees. ing demographic changes in several common Global analysis of gene expression, through and declining Midwestern bumble bee species the use of microarrays, has provided invalu- (Lozier and Cameron, 2009). able insights into nearly all fields of biol- ogy (Schena et al., 1995; Brown and Botstein, 1999; Gibson, 2002; Neumann and Galvez, 3.4. The next frontier of conservation 2002; Cowell and Hawthorn, 2007;van’t biology: Genomics and bee Veer and Bernards, 2008), and has great po- conservation tential for bee conservation management. Af- ter a species’ transcriptome is characterized The emerging field of Genomics has the po- through complete or partial sequencing, short tential to contribute considerably to the con- DNA probes for the discovered transcripts servation management of wildlife populations. can be synthesized on glass slides called mi- Three particular developments have clear util- croarrays. The relative mRNA abundance for ity in bee conservation management: (1) New all genes in the genome – at a specific tis- and cost-effective sequencing technologies; sue at a specific time – can thus be mea- (2) Global gene expression profiling using mi- sured through hybridization of fluorescently- croarrays; and (3) Population genomic studies labelled transcripts isolated from the tissues of using Single Nucleotide Polymorphism (SNP) interest with the probes on the microarray. markers (Box I). New sequencing technologies Microarray experiments are very effective (Shendure et al., 2004; Margulies et al., 2005; in detecting gene expression differences across Strausberg et al., 2008) can generate large experimental groups (e.g. healthy or diseased amounts of data at a fraction of the cost of tissue), and are thus naturally suited to foren- traditional Sanger sequencing, although the se- sic and diagnostic applications. For exam- quences have shorter read length. These tech- ple, microarray experiments on healthy and nologies can be used to re-sequence the known cancerous organs have indicated many differ- genomes of model organisms, or for de novo ences in gene expression levels between the sequencing of the genomes or transcriptomes two (Cowell and Hawthorn, 2007;Nevinsand of non-model organisms. This is an impor- Potti, 2007). Knowledge of differential gene tant first step for the development of plat- expression has been used to better understand forms for measuring global gene expression, the molecular processes underlying cancer by and for developing high-throughput genotyp- identifying potentially causal candidate genes, ing platforms for population genomic analy- in addition to enabling more accurate diag- sis (see below). New sequencing technologies noses through gene expression profiles (e.g. 252 A. Zayed over expression of certain genes can be used applications with relevance to conservation to identify cancer) (Wadlow and Ramaswamy, management (Morin et al., 2004). Following 2005). The same principles can be applied to the discovery of SNPs through sequencing, the field of wildlife conservation. Microarray customized SNP chips can be created to geno- experiments comparing gene expression pro- type hundreds to thousands of SNPs, allow- files between individuals from healthy versus ing for the characterization of genetic diver- declining populations can be used to predict sity across the genome in natural populations the ‘health’ of a population, and to test hy- (Twyman, 2004; Kim and Misra, 2007). The potheses regarding the causes of declines. For data can be used to conduct traditional popula- example, differential expression of genes in- tion genetic analyses (see Sects. 3.1 and 3.2), volved in activating the innate immune sys- as well as to discover areas of the genome tem in a declining population may suggest that experiencing selection (Morin et al., 2004; pathogen infections and not some other fac- Nielsen, 2005;BiswasandAkey,2006; Sabeti tor (e.g. pesticides) is associated with the de- et al., 2006; Nielsen et al., 2007). A recent sur- cline. Microarrays have been developed for vey of SNP diversity in native and introduced the honey bee Apis mellifera (Whitfield et al., Apis mellifera populations clearly demon- 2002), and have recently been used to ex- strates the utility of the approach (Whitfield amine differences in gene expression associ- et al., 2006; Zayed and Whitfield, 2008;De ated with Varroa mite parasitism in both sus- la Rúa et al., 2009). Whitfield et al. (2006) ceptible and tolerant colonies (Navajas et al., used a panel of ∼1100 SNPs to genotype ∼350 2008). Navajas et al. (2008) found that Var- honeybees from across their native and in- roa parasitism affects the expression of sev- troduced ranges. The results supported earlier eral genes associated with immune function findings of four major geographic populations and neural development. Workers from tol- of the honey bee, but uncovered that the honey erant colonies showed differential expression bee actually originated in Africa, and not Asia for several genes affecting neural sensitivity as previously believed. Zayed and Whitfield and olfaction, suggesting that behavior un- (2008) then interrogated the dataset for signa- derlies tolerance to Varroa. Navajas et al.’s tures of selection based on differences in levels (2008) study demonstrates the utility of mi- of genetic differentiation measured by SNPs croarrays in identifying genes associated with in functional versus nonfunctional and pre- particular parasites/pathogens in bees, which sumably neutral regions of the genome. They can then lead to the development of specific found strong signatures of positive selection management tools. Although genomic tools and adaptive evolution associated with the an- are currently only available for the honey bee cient and recent-invasive expansion of honey Apis mellifera, there are plans to generate ge- bees out of Africa into West Europe and the nomic resources for other bees. For exam- New World respectively. The ability of SNP ple, Dr. Gene Robinson (University of Illi- datasets to both estimate demographic param- nois at Urbana Champaign) is leading efforts eters and uncover instances of adaptive evolu- to sequence partial brain transcriptomes of 12 tion will undoubtedly move conservation biol- species spanning the entire phylogeny of bees ogy a step closer to understanding the genetic (G.E. Robinson, unpubl. data). The availabil- basis underlying fitness traits in natural pop- ity of the honey bee genome sequence (The ulations (Mitchell-Olds et al., 2007; Ellegren Honeybee Genome Sequencing Consortium, and Sheldon, 2008), and possibly provide the 2006), combined with the expected develop- means to quantify the evolutionary potential ment of genomic resources for other bees, will of small endangered populations (Morin et al., lay the ground work for developing microarray 2004). technology that can be used for investigating the causes of bee declines in both model and 4. CONCLUSIONS non-model species. Finally, population genomic studies using Bees are indispensable components of ter- SNP markers can be used for a variety of restrial ecosystems and their conservation is Bee conservation genetics 253 essential for both ecological and economic and/or isolated populations. (2) When sub- reasons. Efforts to conserve declining bee jected to an extrinsic factor causing decline, populations can only be as effective as our bee populations will do so at a faster rate knowledge of their biology, and the causes than expected based on the direct effect of the contributing to their declines. The application extrinsic factor. (3) Bee populations should re- of genetics to bee conservation biology can be cover at a slower rate than expected following of great use in aiding bee conservation man- the removal of extrinsic factors causing de- agement, and in identifying genetic threats to clines. Bee populations targeted for conserva- the short term viability of bee populations. tion should be managed to reduce frequencies Population genetic surveys can be, and have of diploid male production, and to a lesser ex- indeed been used, to provide knowledge about tent, inbreeding depression. several important parameters of relevance to bee conservation biology, including popula- tion structure, gene flow, effective population sizes, colony densities, and foraging ranges. ACKNOWLEDGEMENTS Genetic approaches can also be used for rapid species identification, for resolving taxonomic I am most grateful to Laurence Packer for mo- conflict, and for indicating bee populations in tivating and guiding my research on bee conser- decline. They also provide more robust and vation genetics. I would like to thank Laurence Packer, Rob Paxton, Richard Frankham, Jennifer timely estimates of population size in com- ff parison to the vagaries of census methods. As Grixti, and Je Lozier for providing helpful com- ments on the manuscript. I also thank Jennifer more genetic and genomic resources are de- Grixti, and Jamie Strange for providing copies of veloped for bees, it will become progressively their manuscript slated for publication, and to Gene easier and cheaper to gather population ge- Robinson, Jason Gibbs, Laurence Packer, and Jeff netic data from both common and endangered Lozier for allowing me to refer to their unpublished bee species for comparative analyses and con- research. Finally, I wish to thank Jason Alexander servation studies. Integrating population ge- Zayed for providing inspiration and a stimulating netic surveys with ecological studies, although environment to write this review. My work is sup- rarely undertaken, can provide greater insight ported by a Natural Sciences and Engineering Re- into bee conservation biology than either ap- search Council of Canada Postdoctoral Fellowship, proach practiced alone. and a postdoctoral fellowship from the Institute for The production of inviable or sterile diploid Genomic Biology, University of Illinois at Urbana- males, a necessary by-product of complemen- Champaign. tary sex determination, is a large threat to the short-term viability of small bee populations, as indicated both by theory and mostly in- Génétique des abeilles et conservation des direct empirical data. Inbreeding depression, espèces. caused by dominance or overdominance, al- Apoidea / haploidiploïdie / dépression consan- though expected to be lower in haplodiploids guinité / mâle diploïde / extinction / détermina- when compared to diploids, should still con- tion sexe complémentaire / abeille tribute to reduced fitness and reduced via- bility of small bee populations. Inbreeding depression is expected to interact synergis- Zusammenfassung – Bienengenetik und Arten- schutz. In dieser Arbeit betrachten wir, welche tically with diploid male production causing Rolle die Genetik im Rückgang der Bienenpopula- greater rates of extinction than the already high tionen spielt und wie die Werkzeuge der Genetik in rates caused by the latter. These genetic fac- Bienenschutzprojekten eingesetzt werden können. tors will also interact synergistically with ex- Genetische Marker haben sich als höchst nützlich trinsic factors causing declines in bee popu- erwiesen in der Bestimmung wichtiger demogra- phischer Parameter von Bienenpopulationen (z.B. lations. The above mentioned genetic threats Populationsgrössen, Dichte, Vernetzung, Ausbrei- to population viability imply that: (1) Bees tungsraten und Sammelradien), sowie zur Klärung are highly susceptible to extinction in small taxonomischer Probleme und zur Erkennung von 254 A. Zayed

Rückgängen in Populationen. Neuere Fortschritte Single locus complementary sex deter- in der Bienengenomik sollten ebenfalls einen Bei- mination: Sex is determined by genotype at trag zur Erleichterung von Untersuchungen über die a single locus: heterozgyotes, homozygotes Ursachen von Rückgängen in Bienenpopulationen leisten. Die Integration der Kentnisse zur Genetik and hemizygotes develop into diploid females, und Genomik der Bienen mit denen ihrer Ökologie diploid males, and haploid males respectively. sollte daher deutliche Verbesserungen in Bienen- Diploid species: A species where both schutzprogrammen zur Folge haben. Ein wichtiger sexes have two copies of each autosomal chro- Faktor, der kleine Bienenpopulationen bedroht, ist mosome. die durch Homozygotie am Geschlechtslokus aus- ff gelöste Produktion nicht lebensfähiger, steriler di- E ective population size,Ne: The size of ploider Männchen anstelle von Weibchen (Abb. 1). an ‘ideal’ population (a random mating popu- Das Vorkommen diploider Männchen ist für minde- lation of constant size with Poisson variation sten 27 Arten beschrieben (Tab. I) und die Frequenz in family sizes) that would have the same ge- dieser Männchen nimmt in kleinen Populationen netic parameters (e.g. genetic drift) as the ac- durch Homozygotie am Geschlechtsbestimmungs- lokus bedingt durch genetische Drift zu (Abb. 2). tual population under study. Simulationsstudien lassen erwarten, dass die Pro- Evolutionary potential: The ability of duktion diploider Männchen die Populationswachs- populations to adapt to future changes in their tumsraten vor allem bei solitären Bienen reduzieren environments. kann, die eine niedrige Fekundität in ihrer Gesamt- Frequency dependent selection: A form lebenszeit aufweisen. Das Problem der Produktion diploider Männchen geht so Hand in Hand mit ne- of natural selection where the relative fit- gativen Umweltfaktoren und führt zu einem schnel- ness of genotypes is a function of their fre- leren Rückgang und einer langsameren Erhohlung quency. Negative and positive frequency de- von Populationen (Abb. 3). Sie spielt damit eine pendent selection denotes higher fitness of, wichtige Rolle im weltweiten Rückgang von Bie- and selection for, rare or common genotypes nen. Ausser der Produktion diploider Männchen könn- respectively. The sex determination locus in te auch eine Inzuchtdepression das Überleben klei- the Hymenoptera is under negative frequency ner Bienenpopulationen gefährden. Obwohl zu er- dependent selection since the fitness of indi- warten wäre, dass die Haplodoploidie die geneti- viduals carrying rare sex-determining alleles is sche Bürde bei Bienen reduziert, zeigen sowohl higher as they are less likely to produce diploid theoretische Studien als auch empirische Befunde, dass selbst haplodiploide Organismen eine erhebli- males. che Inzuchtdepression durchlaufen können. Schutz- Genetic distance and differentiation:Pa- programme für gefährdete und zurückgehende Bie- rameters that measure the extent of genetic nenpopulationen sollten daher bestrebt sein, die differences between populations, usually over Produktion diploider Männchen und Inzucht zu space. Populations with similar gene pools reduzieren. (e.g. similar allele frequencies) have smaller Haplodiploidie / Komplementäre Geschlechts- genetic distances and levels of differentiation determinierung / Inzuchtdepression / diploide than populations with different gene pools. Männchen / Artensterben Genetic drift: Changes in allele frequen- cies due to the effects of random sampling of gametes in finite populations. Genetic drift is stronger in smaller populations, and results in BOX I: GLOSSARY OF GENETIC reduced genetic diversity. TERMS Genetic load: The reduction in fitness of a population from the maximum possible due to Additive genetic variance: The contribu- the deleterious effects (i.e. load) of alleles. The tion of genetic variance in a quantitative trait genetic load can be due to the deleterious ef- due to the effects of substituting one allele for fects of mutations (i.e. mutational load) or due another at a locus. to lower fitness of homozygous genotypes rel- Allozymes: Alternative forms of a particu- ative to heterozygous genotypes (i.e. balance lar protein as visualized on a gel, mostly result- load). ing from genetic variation at non-synonymous Sex determination load: The reduction of sites in a gene’s protein coding sequence. fitness of a haplodiploid population from the Bee conservation genetics 255

maximum possible due to the deleterious ef- a population of Apis mellifera, Genetics 86, 583– fects of homozygosity at the hymenopteran 596. complementary sex determination locus. Agoze M.E., Drezen J.M., Renalt S., Preiquet G. Haplodiploid species: A species where fe- (1994) Analysis of the reproductive potential of males are diploid (i.e. have two copies of each diploid males in the wasp Diadromus pulchellus (Hymenoptera: Ichneumonidae), Bull. Entomol. chromosome), and males are haploid (i.e. have Res. 84, 213–218. one copy of each chromosome). Inbreeding: Several phenomena leading to Allen-Wardell G., Bernhardt P., Bitner R., Burquez A., Buchmann S.L., Cane J.H., Cox P.A., Dalton mating between individuals related by de- V., Feinsinger P., Ingram M., Inouye D.W., Jones scent, and resulting in increased homozygos- C.E., Kennedy K., Kevan P.G., Koopowitz H., ity. Inbreeding in a mating-system sense, know Medellin R., Medellin-Morales S., Nabhan G.P. as systematic inbreeding, results when re- (1998) The potential consequences of pollinator declines on the conservation of biodiversity and latedness between mates are higher than the stability of food crop yields, Conserv. Biol. 12, 8– average relatedness of mates chosen at ran- 17. dom from the same population. Inbreeding Antolin M.F. (1999) A genetic perspective on mating by drift,orpanmictic inbreeding, results systems and sex ratios of parasitoid wasps, Res. from the increasing probability of relatedness Popul. Ecol. 41, 29–37. among randomly chosen mates in a finite pop- Armbruster P., Reed D.H. (2005) Inbreeding de- ulation over time. pression in benign and stressful environments, Inbreeding depression: Reduction in the Heredity 95, 235–242. mean fitness of a trait due to inbreeding. At- Ayabe T., Hoshiba H., Ono M. (2004) Cytological ev- tributed to the increased expression of reces- idence for triploid males and females in the bum- sive or partially recessive deleterious alleles blebee, Bombus terrestris, Chromosome Res. 12, through increased homozygosity (Dominance 215–223. or Partial dominance hypothesis),ortothe Beveridge M., Simmons L.W. (2006) Panmixia: an ex- reduction in the frequency of fitter heterozy- ample from Dawson’s burrowing bee (Amegilla gous genotypes (Overdominance hypothe- dawsoni) (Hymenoptera: Anthophorini), Mol. sis). Ecol. 15, 951–957. Matched mating: A mating between a Beveridge M., Simmons L.W., Alcock J. (2006) male and female that share a sex-determining Genetic breeding system and investment patterns allele in common. Half of the diploid progeny within nests of Dawson’s burrowing bee (Amegilla dawsoni) (Hymenoptera: Anthophorini), Mol. produced by a matched-mated female will be Ecol. 15, 3459–3467. homozygous at the sex-determination locus, Beye M., Hasselmann M., Fondrk M.K., Page R.E., and develop into diploid males. Omholt S.W. (2003) The gene csd is the primary Microsatellite: A sequence of DNA com- signal for sexual development in the honey bee posed of short (2-6 bases) tandem repeats (e.g. and encodes a new SR-type protein, Cell 114, CACACACACA). Mutation at microsatellite 419–429. loci results in alleles with different repeat Bienefeld K., Reinhardt F., Pirchner F. (1989) ff numbers. Inbreeding e ects of queen and workers on colony traits in the honey bee, Apidologie 20, 439–450. Neutral genetic variation: Genetic varia- tion that does not affect fitness. Biesmeijer J.C., Roberts S.P.M., Reemer M., Ohlemuller R., Edwards M., Peeters T., Schaffers Single nucleotide polymorphism, SNP: A A.P., Potts S.G., Kleukers R., Thomas C.D., specific position in a DNA sequence which Settele J., Kunin W.E. (2006) Parallel declines in contains alternative bases. pollinators and insect-pollinated plants in Britain and the Netherlands, Science 313, 351–354. Biswas S., Akey J.M. (2006) Genomic insights into positive selection, Trends Genet. 22, 437–446. REFERENCES Blanchetot A., Packer L. (1992) Genetic variabil- ity in the social bee Lasioglossum marginatum Adams J., Rothman E.D., Kerr W.E., Paulino Z.L. and a cryptic underscribed sibling species, as de- (1977) Estimation of the number of sex alleles and tected by DNA fingerprinting and allozyme elec- queen matings from diploid male frequencies in trophoresis, Insect Mol. Biol. 1, 89–97. 256 A. Zayed

Brook B.W., Burgman M.A., Akçakaya H.R., O’Grady Cook J.M. (1993) Sex determination in the J.J., Frankham R. (2002a) Critiques of PVA ask Hymenoptera: a review of models and evidence, the wrong questions: throwing the heuristic baby Heredity 71, 421–435. out with the numerical bath water, Conserv. Biol. 16, 262–263. Cook J.M., Crozier R.H. (1995) Sex determination and population biology of the Hymenoptera, Trends Brook B.W., Tonkyn D.W., O’Grady J.J., Frankham R. Ecol. Evol. 10, 281–286. (2002b) Contribution of inbreeding to extinction risk in threatened species, Conserv. Ecol. 6, 16. Cornuet J.M. (1980) Rapid estimation of the number of sex alleles in panmictic honeybee populations, Brown P.O., Botstein D. (1999) Exploring the new J. Apic. Res. 19, 3–5. world of the genome with DNA microarrays, Nat. Genet. 21, 33–37. Cornuet J.M., Luikart G. (1996) Description and power analysis of two tests for detecting recent pop- Byrne A., Fitzpatrick Ú. (2009) Bee conservation pol- ulation bottlenecks from allele frequency data, icy at the global, regional and national levels, Genetics 144, 2001–2014. Apidologie 40, 194–210. Cowan D.P., Stahlhut J.K. (2004) Functionally repro- Camargo C.A. (1979) Sex determination in bees. XI. ductive diploid and haploid males in an inbreed- Production of diploid males and sex determination ing hymenopteran with complementary sex deter- in Melipona quadrifasciata, J. Apic. Res. 18, 77– mination, Proc. Natl Acad. Sci. USA 101, 10374– 84. 10379. Cameron E.C., Franck P., Oldroyd B.P. (2004) Genetic Cowell J.K., Hawthorn L. (2007) The application of structure of nest aggregations and drone congrega- microarray technology to the analysis of the can- tions of the southeast Asian stingless bee Trigona cer genome, Curr. Mol. Med. 7, 103–120. collina, Mol. Ecol. 13, 2357–2364. Cox-Foster D.L., Conlan S., Holmes E.C., Palacios Cane J.H. (2001) Habitat fragmentation and native G., Evans J.D., Moran N.A.,QuanP.-L.,Briese bees: a premature verdict? Conserv. Ecol. 5, 3. T., Hornig M., Geiser D.M., Martinson V., vanEngelsdorp D., Kalkstein A.L., Drysdale A., Cane J.H., Tepedino V.J. (2001) Causes and extent of Hui J., Zhai J., Cui L., Hutchison S.K., Simons declines among native North American inverte- J.F., Egholm M., Pettis J.S., Lipkin W.I. (2007) A brate pollinators: detection, evidence, and conse- Metagenomic Survey of Microbes in Honey Bee quences, Conserv. Ecol. 5, 1. Colony Collapse Disorder, Science 318, 283–287. Carman G.M., Packer L. (1997) A cryptic species Crnokrak P., Barrett S.C. (2002) Perspective: purging allied to Halictus ligatus Say (Hymenoptera; the genetic load: a review of the experimental evi- Halictidae) detected by allozyme electrophoresis, dence, Evolution 12, 2347–2358. J. Kans. Entomol. Soc. 69, 168–176. Crozier R.H. (1976a) Counter-intuitive property of ef- Carvalho G.A. (2001) The number of sex alleles (CSD) fective population size, Nature 262, 384. in a bee population and its practical importance (Hymenoptera: Apidae), J. Hymenoptera Res. 10, Crozier R.H. (1976b) Why Male-Haploid and Sex- 10–15. Linked Genetic Systems Seem to Have Unusually Sex-Limited Mutational Genetic Loads, Evolution Chaline N., Ratnieks F.L.W., Raine N.E., Badcock 30, 623–624. N.S., Burke T. (2004) Non-lethal sampling of honey bee, Apis mellifera, DNA using wing tips, Crozier R.H. (1977) Evolutionary genetics of the hy- Apidologie 35, 311–318. menoptera, Annu. Rev. Entomol. 22, 263–288. Chapman R.E., Wang J., Bourke A.F.G. (2003) Crozier R.H. (1985) Adaptive consequences of male- Genetic analysis of spatial foraging patterns and haploidy, in: Helle W., Sabelis M.W. (Eds.), resource sharing in bumble bee pollinators, Mol. Spider Mites: their biology, natural enemies Ecol. 12, 2801–2808. and control, Elsevier Science Publisher B.V. Amsterdam, pp. 201–222. Charlesworth B., Charlesworth D. (1999) The genetic basis of inbreeding depression, Genet. Res. 74, Danforth B.N. (1999) Phylogeny of the bee genus 329–340. Lasioglossum (Hymenoptera: Halictidae) based on mitocondrial COI sequence data, Syst. Colla S.R., Packer L. (2008) Evidence for de- Entomol. 24, 377–393. cline in eastern North American bumblebees (Hymenoptera: Apidea), with special focus on Danforth B.N., Mitchell P.L., Packer L. (1998) Bombus affinis Cresson, Biodivers. Conserv. 17, Mitochondrial DNA differentiation between 1379–1391. two cryptic Halictus (Hymenoptera: Halictidae) species, Ann. Entomol. Soc. Am. 91, 387–391. Colla S.R., Otterstatter M.C., Gegear R.J., Thomson J.D. (2006) Plight of the bumblebee: Pathogen Danforth B.N., Ji S., Ballard L.J. (2003) Gene spillover from commercial to wild populations, flow and population structure in an oligolectic Biol. Conserv. 129, 461–467. desert bee, Macrotera (Macroteropsis) portalis Bee conservation genetics 257

(Hymenoptera: Andrenidae), J. Kans. Entomol. nental and island populations of Bombus terrestris Soc. 76, 221–235. (Hymenoptera: Apidae) in Europe, Mol. Ecol. 5, Danks H.V. (1971) Nest mortality factors in stem- 19–31. nesting aculeate Hymenoptera, J. Anim. Ecol. 40, Excoffier L., Heckel G. (2006) Computer programs for 79–82. population genetics data analysis: a survival guide, Darvill B., Knight M.E., Goulson D. (2004) Use of Nat. Rev. Genet. 7, 745–758. genetic markers to quantify bumblebee foraging Falconer D.S., Mackay T.F.C. (1996) Introduction to range and nest density, Oikos 107, 471–478. Quantitative Genetics, Longman, Essex, UK. Darvill B., Ellis J.S., Lye G.C., Goulson D. (2006) Fisher R.A. (1930) The Genetic Theory of Natural Population structure and inbreeding in a rare Selection, Dover, New York. and declining bumblebee, Bombus muscorum (Hymenoptera: Apidae), Mol. Ecol. 15, 601–611. Fitzpatrick U., Murray M.G., Paxton R.J., Breen J., Cotton D., Santorum V., Brown M.J.F. (2006) De Boer J.G., Ode P.J., Vet L.E.M., Whitfield J., Rarity and decline in bumblebees - A test of causes Heimpel G.E. (2007) Complementary sex deter- and correlates in the Irish fauna, Biol. Conserv. mination in the parasitoid wasp Cotesia vestalis 136, 185–194. (C. plutellae), J. Evol. Biol. 20, 340–348. Fontaine C., Dajoz I., Meriguet J., Loreau M. ff De la Rúa P., Ja é R., Dall’Olio R., Muñoz I. Serrano (2006) Functional Diversity of Plant-Pollinator J. (2009) Biodiversity, conservation and current Interaction Webs Enhances the Persistence of threats to European honeybees, Apidologie 40, Plant Communities, PLoS Biol. 4, e1. 263–284. Franck P., Cameron E., Good G., Rasplus J.Y., Oldroyd Dias B.S.F., Raw A., Imperatriz-Fonseca V.L. (1999) B.P. (2004) Nest architecture and genetic differen- International pollinators initiative: The São Paulo tiation in a species complex of Australian stingless declaration on pollinators., Brazilian Ministry of bees, Mol. Ecol. 13, 2317–2331. the Environment, Brasilia. [online] http://www. cbd.int/doc/case-studies/agr/cs-agr-pollinator-rpt. Frankham R. (1995a) Effective population size/adult pdf (accessed on 2 February 2009). population size ratios in wildlife: a review, Genet. Res. 66, 95–107. Dick C.W., Roubik D.W., Gruber K.F., Bermingham E. (2004) Long-distance gene flow and cross-Andean Frankham R. (1995b) Inbreeding and extinction: a dispersal of lowland rainforest bees (Apidae: threshold effect, Conserv. Biol. 9, 792–799. Euglossini) revealed by comparative mitochon- Frankham R. (1998) Inbreeding and extinction: island drial DNA phylogeography, Mol. Ecol. 13, 3775– populations, Conserv. Biol. 12, 665–675. 3785. Frankham R. (2005) Genetics and extinction, Biol. Duchateau M.J., Hoshiba H., Velthuis H.H.W. (1994) Conserv. 126, 131–140. Diploid males in the bumblebee Bombus ter- restris: sex determination, sex alleles and viability, Frankham R., Ballou J.D., Briscoe D.A. (2002) Entomol. Exp. Appl. 71, 263–269. Introduction to conservation genetics, Cambridge University Press, Cambridge. Eickwort G.C., Ginsberg H.S. (1980) Foraging and mating behavior in Apoidea, Annu. Rev. Entomol. Franzén M., Larsson M. (2007) Pollen harvesting and 25, 421–446. reproductive rates in specialized solitary bees, Ellegren H., Sheldon B.C. (2008) Genetic basis of Ann. Zool. Fenn. 44, 405–414. fitness differences in natural populations, Nature Gadau J., Gerloff C.U., Krüger N., Chan H., Schmid- 452, 169–175. Hempel P., Wile A., Page R.E.J. (2001) A linkage Ellis J.S., Knight M.E., Darvill B., Goulson D. (2006) analysis of sex determination in Bombus terrestris Extremely low effective population sizes, genetic (L.) (Hymenoptera: Apidae), J. Hered. 87, 234– structuring and reduced genetic diversity in a 242. threatened bumblebee species, Bombus sylvarum Gaggiotti O.E. (2003) Genetic threats to population (Hymenoptera: Apidae), Mol. Ecol. 15, 4375– persistence, Ann. Zool. Fenn. 40, 155–168. 4386. Gerloff C.U., Schmid-Hempel P. (2005) Inbreeding de- Else G., Felton J., Stubbs A. (1978) The Conservation pression and family variation in a social insect, of Bees and Wasps, Nature Conservancy Council, Bombus terrestris (Hymenoptera: Apidae), Oikos Peterborough. 111, 67–80. Estoup A., Scholl A., Pouvreau A., Solignac M. Gibbs J. (2009) Integrative taxonomy identifies (1995) Monoandry and polyandry in bumble bees new (and old) species in the Lasioglossum (Hymenoptera; Bombinae) as evidenced by highly (Dialictus) tegulare (Robertson) species group variable microsatellites, Mol. Ecol. 4, 89–94. (Hymenoptera, Halictidae), Zootaxa 2032, 1–38. Estoup A., Solignac M., Cornuet J.M., Goudet J., Gibson G. (2002) Microarrays in ecology and evolu- Scholl A. (1996) Genetic differentiation of conti- tion: a preview, Mol. Ecol. 11, 17–24. 258 A. Zayed

Gilpin M.E. (1991) The genetic effectivesizeofa Heimpel G.E., de Boer J.G. (2008) Sex determina- metapopulation, Biol. J. Linn. Soc. Lond. 42, 165– tion in the Hymenoptera, Annu. Rev. Entomol. 53, 175. 209–230. Glémin S. (2003) How are deleterious mutations Henter H.J. (2003) Inbreeding depression and hap- purged? Drift versus nonrandom mating, lodiploidy: experimental measures in a parasitoid Evolution 57, 2678–2687. and comparisons across diploid and haplodiploid insect taxa, Evolution 57, 1793–1803. Goulson D., Lye G.C., Darvil B. (2008) Decline and conservation of bumble bees, Annu. Rev. Herrmann F., Westphal C., Moritz R.F.A., Steffan- Entomol. 53, 191–208. Dewenter I. (2007) Genetic diversity and mass re- sources promote colony size and forager densities Green C.L., Oldroyd B.P. (2002) Queen mating fre- of a social bee (Bombus pascuorum) in agricul- quency and maternity of males in the stingless bee tural landscapes, Mol. Ecol. 16, 1167–1178. Trigona carbonaria Smith, Insectes Soc. 49, 196– 202. Holehouse K.A., Hammond R.L., Bourke A.F.G. (2003) Non-lethal sampling of DNA from bumble Grixti J.C., Packer L. (2006) Changes in the bee fauna bees for conservation genetics, Insectes Soc. 50, (Hymenoptera: Apoidea) of an old field site in 277–285. southern Ontario, revisited after 34 years, Can. Entomol. 138, 147–164. Holloway A.K., Heimpel G.E., Strand M.R., Antolin M.F. (1999) Survival of diploid males in Bracon Grixti J.C., Wong L.T., Cameron S.A., Favret C. sp. near hebetor (Hymenoptera: Braconidae), (2009) Decline of bumble bees (Bombus)inthe Ann. Entomol. Soc. Am. 92, 110–116. North American Midwest, Biol. Conserv. 142, 75– 84. Jacquard A. (1975) Inbreeding: One word, several meanings, Theor. Popul. Biol. 7, 338–363. Hanski I., Saccheri I. (2006) Molecular-Level ff Variation Affects Population Growth in a Keller L.F., Waller D.M. (2002) Inbreeding e ects in Butterfly Metapopulation, PLoS Biol. 4, e129. wild populations, Trends Ecol. Evol. 17, 230–241. Harbo J.R., Harris J.W. (1999) Heritability in honey Kerr W.E. (1987) Sex determination in bees. XXI. bees (Hymenoptera: Apidae) of characteristics Number of XO-heteroalleles in a natural pop- associated with resistance to Varroa jacobsoni ulation of Melipona compressipes fasciculata (Mesostigmata: Varroidae), J. Econ. Entomol. 92, (Apidae), Insectes Soc. 34, 274–279. 261–265. Kerr W.E. (1997) Sex determination in honey bees Hasselmann M., Gempe T., Schiott M., Nunes-Silva (Apinae and Meliponinae) and its consequences, C.G., Otte M., Beye M. (2008) Evidence for the Braz. J. Genet. 20, 4. evolutionary nascence of a novel sex determina- Kim S., Misra A. (2007) SNP genotyping: tion pathway in honeybees, Nature 454, 519–522. Technologies and biomedical applications, Annu. Rev. Biomed. Eng. 9, 289–320. Hebert P.N.D., Ratnasingham S., DeWaard J.R. (2003) Barcoding animal life: cytochrome c oxidase sub- Klein A.M., Vaissière B.E., Cane J.H., Steffan- unit 1 divergences among closely related species, Dewenter I., Cunningham S.A., Kremen C., Proc. R. Soc. Ser. B Biol. Sci. 270, S96–S99. Tscharntke T. (2007) Importance of pollinators in changing landscapes for world crops, Proc. R. Soc. Hedrick P.W. (2000) Genetics of Populations, Second Lond., Ser. B Biol. Sci. 274, 303–313. edn, Jones and Bartlett Publishers, Sudbury, Massachusetts. Knight M.E., Martin A.P., Bishop S., Osborne J.L., Hale R.J., Sanderson R.A., Goulson D. (2005) Hedrick P.W. (2004) Recent developments in conser- An interspecific comparison of foraging range and vation genetics, For. Ecol. Manage. 197, 3–19. nest density of four bumblebee (Bombus) species, Hedrick P.W., Gilpin M.E. (1997) Genetic effective Mol. Ecol. 14, 1811–1820. size of a metapopulation, in: Hanski I.A., Gilpin Kosior A., Celary W., Olejniczak P., Fijal J., Król M.E. (Eds.), Metapopulation biology: ecology, ge- W., Solarz W., Plonka P. (2007) The decline of netics, and evolution, Academic Press Inc. San the bumble bees and cuckoo bees (Hymenoptera: Diego, USA, pp. 165–181. Apidae: Bombini) of western and central Europe, Hedrick P.W., Parker J.D. (1997) Evolutionary genetics Oryx 41, 79–88. and genetic variation of haplodiploids and X-lined Kraus F.B., Weinhold S., Moritz R.F.A. (2008) Genetic genes, Annu. Rev. Ecol. Syst. 28, 55–83. structure of drone congregations of the stingless Hedrick P.W., Kalinowski S.T. (2000) Inbreeding de- bee Scaptotrigona mexicana, Insectes Soc. 55, 22– pression in conservation biology, Annu. Rev. Ecol. 27. Syst. 31, 139–162. Kremen C., Williams N.M., Thorp R.W. (2002) Crop Hedrick P.W., Gadau J., Page R.E.J. (2006) Genetic sex pollination from native bees at risk from agricul- determination and extinction, Trends Ecol. Evol. tural intensification, Proc. Natl Acad. Sci. USA 21, 55–57. 99, 16812–16816. Bee conservation genetics 259

Kremen C., Williams N.M., Aizen M.A., Gemmill- Lynch M., Conery J., Burger R. (1995) Mutation accu- Herren B., LeBuhn G., Minckley R., Packer L., mulation and the extinction of small populations, Potts S.G., Roulston T.a., Steffan-Dewenter I., Am. Nat. 146, 489–518. Vazquez D.P., Winfree R., Adams L., Crone E.E., Greenleaf S.S., Keitt T.H., Klein A.-M., Regetz J., Lynch M., Walsh B. (1998) Genetics and Analysis Ricketts T.H. (2007) Pollination and other ecosys- of Quantitative Traits, Sinauer Associates Inc., tem services produced by mobile organisms: a Sunderland, MA, USA. conceptual framework for the effects of land-use Margulies M., Egholm M., Altman W.E., Attiya S., change, Ecol. Lett. 10, 299–314. Bader J.S., Bemben L.A., Berka J., Braverman Krieger M.J.B., Ross K.G., Chang C.W., Keller L. M.S., Chen Y.-J., Chen Z., Dewell S.B., Du (1999) Frequency and origin of triploidy in the fire L.,FierroJ.M.,GomesX.V.,GodwinB.C.,He ant Solenopsis invicta, Heredity 82, 142–150. W., Helgesen S., Ho C.H., Irzyk G.P., Jando Kuhlmann M., Else G.R., Dawson A., Quicke D.L.J. S.C., Alenquer M.L.I., Jarvie T.P., Jirage K.B., (2007) Molecular, biogeographical and phenolog- Kim J.-B., Knight J.R., Lanza J.R., Leamon J.H., ical evidence for the existence of three western Lefkowitz S.M., Lei M., Li J., Lohman K.L., European sibling species in the Colletes succinc- Lu H., Makhijani V.B., McDade K.E., McKenna tus group (Hymenoptera: Apidae), Org. Divers. M.P.,MyersE.W.,NickersonE.,NobileJ.R., Evol. 7, 155–165. Plant R., Puc B.P., Ronan M.T., Roth G.T., Sarkis G.J., Simons J.F., Simpson J.W., Srinivasan M., Kukuk P.F., May B. (1990) Diploid males in a Tartaro K.R., Tomasz A., Vogt K.A., Volkmer primitively eusocial bee, Lasioglossum (Dialictus) G.A., Wang S.H., Wang Y., Weiner M.P., Yu P., zephyrum, Evolution 44, 1522–1528. Begley R.F., Rothberg J.M. (2005) Genome se- Lacy R.C. (1993) Vortex: a computer simulation model quencing in microfabricated high-density picolitre for population viability analysis, Wildl. Res. 20, reactors, Nature 437, 376–380. 45–64. McCorquodale D.B., Owen R.E. (1994) Laying Lapidge K.L., Oldroyd B.P., Spivak M. (2002) sequence, diploid males and nest usurpation Seven suggestive quantitative trait loci in- in the leafcutter bee, Megachile rotundata fluence hygienic behavior of honey bees, (Hymenoptera: Megachilidae), J. Insect. Behav. 7, Naturwissenschaften 89, 565–568. 731–738. Le Conte Y., De Vaublanc G., Crauser D., Jeanne F., Rousselle J.C., Becard J.M. (2007) Honey bee Michener C.D. (2000) The bees of the world, colonies that have survived Varroa destructor, The Johns Hopkins University Press, Baltimore, Apidologie 38, 566–572. Maryland. Leberg P.L., Firmin B.D. (2008) Role of inbreeding Michener C.D., Rettenmeyer C.W. (1956) The ethol- depression and purging in captive breeding and ogy of Andrena erythronii with comparative data restoration programmes, Mol. Ecol. 2008, 334– on other species, Univ. Kans. Sci. Bull. 37, 645– 343. 684. Liebert A.E., Johnson R.N., Switz G.T., Starks P.T. Minckley R.L., Wcislo W.T., Yanega D., Buchmann (2004) Triploid females and diploid males: under- S.L. (1994) Behavior and phenology of a special- reported phenomena in Polistes wasps? Insectes ist bee (Dieunomia) and sunflower (Helianthus) Soc. 51, 205–211. pollen availability, Ecology 75, 1406–1419. Liebert A.E., Sumana A., Starks P.T. (2005) Diploid males and their triploid offspring in the paper wasp Mitchell-Olds T., Willis J.H., Goldstein D.B. (2007) Polistes dominulus, Biol. Lett. 1, 200–203. Which evolutionary processes influence natural genetic variation for phenotypic traits? Nat. Rev. Lopez-Uribe M.M., Almanza M.T., Ordonez M. Genet. 8, 845–856. (2007) Diploid male frequencies in Colombian populations of euglossine bees, Biotropica 39, Morin P.A., Luikart G., Wayne R.K. (2004) SNPs in 299–314. ecology, evolution and conservation, Trends Ecol. Lozier J.D., Cameron S.A. (2009) Comparative ge- Evol. 19, 208–216. netic analyses of historical and contemporary col- Moritz C., Cicero C. (2004) DNA barcoding: promise lections highlight contrasting demographic his- and pitfalls, PLoS Biol. 2, e279. tories for the bumble bees Bombus pensylvan- icus and B. impatiens in Illinois, Mol. Ecol., Müller A., Diener S., Schnyder S., Stutz K., Sedivy C., doi:10.1111/j.1365.294X.2009.04160x. Dorn S. (2006) Quantitative pollen requirements Luikart G., England P.R. (1999) Statistical analysis of of solitary bees: Implications for bee conserva- microsatellite DNA data, Trends Ecol. Evol. 14, tion and the evolution of bee–flower relationships, 253–256. Biol. Conserv. 130, 604–615. Luna M.G., Hawkins B.A. (2004) Effects of Murray T.E., Fitzpatrick U., Brown M.J.F., Paxton R.J. Inbreeding Versus Outbreeding in Nasonia vit- (2008) Cryptic species diversity in a widespread ripennis (Hymenoptera: Pteromalidae), Environ. bumble bee complex revealed using mitochondrial Entomol. 33, 765–775. DNA RFLPs, Conserv. Genet. 9, 653–666. 260 A. Zayed

Murray T.E., Kuhlmann M., Potts S.G. (2009) Pamilo P., Crozier R.H. (1997) Population biology of Conservation ecology of bees: populations, social insect conservation, Mem. Mus. Vict. 56, species and communities, Apidologie 40, 211– 411–419. 236. Pauw A. (2007) Collapse of a pollination web in small Navajas M., Migeon A., Alaux C., Martin-Magniette conservation areas, Ecology 88, 1759–1769. M.L., Robinson G.E., Evans J.D., Cros-Arteil S., Crauser D., Le Conte Y. (2008) Differential gene Paxton R.J. (2005) Male mating behaviour and mating expression of the honey bee Apis mellifera as- systems of bees: an overview, Apidologie 36, 145– sociated with Varroa destructor infection, BMC 156. Genomics 9, 301. Paxton R.J., Thorén P.A., Gyllenstrand N., Tengö J. Neumann N.F., Galvez F. (2002) DNA microarrays and (2000) Microsatellite DNA analysis reveals low toxicogenomics: applications for ecotoxicology? diploid male production in a communal bee with Biotechnol. Adv. 20, 391–419. inbreeding, Biol. J. Linn. Soc. Lond. 69, 483–502. Nevins J.R., Potti A. (2007) Mining gene expression Paxton R.J., Bego L.R., Shah M.M., Mateus S. (2003) profiles: expression signatures as cancer pheno- Low mating frequency of queens in the stingless types, Nat. Rev. Genet. 8, 601–609. bee Scaptotrigona postica and worker maternity of males, Behav. Ecol. Sociobiol. 53, 174–181. Nielsen R. (2005) Molecular signatures of natural se- lection, Annu. Rev. Genet. 39, 197–218. Pfrender M.E., Spitze K., Hicks J., Morgan K., Latta L., Lynch M. (2000) Lack of concordance between Nielsen R., Hellmann I., Hubisz M., Bustamante C., genetic diversity estimates at the molecular and Clark A.G. (2007) Recent and ongoing selection quantitative-trait levels, Conserv. Genet. 1, 263– in the human genome, Nat. Rev. Genet. 8, 857– 269. 868. NRC (2007) Status of pollinators in North America, Plowright R.C., Pallett M.J. (1979) Worker-male National Academies Press, Washington, DC. conflict and inbreeding in bumble bees (Hymenoptera: Apidae), Can. Entomol. 111, Olesen J.M., Bascompte J., Dupont Y.L., Jordano P. 289–294. (2007) The modularity of pollination networks, Proc. Natl Acad. Sci. USA 104, 19891–19896. Quezada-Euán J.J.G., Paxton R.J., Palmer K.A., Itza W.D.M., Tay W.T., Oldroyd B.P. (2007) Otterstatter M.C., Thomson J.D. (2008) Does Morphological and molecular characters re- Pathogen Spillover from Commercially Reared veal differentiation in a Neotropical social Bumble Bees Threaten Wild Pollinators? PLoS bee, Melipona beecheii (Apidae: Meliponini), ONE 3, e2771. Apidologie 38, 247–258. Owen R.E. (1985) The opportunity for polymorphism Ralls K., Ballou J. (1983) Extinction: lessons from and genic variation in social hymenoptera with zoos, in: Schonewald-Cox C.M., Chambers worker-produced males, Heredity 54, 25–36. S.M., MacBryde B., Thomas L. (Eds.), Owen R.E., Packer L. (1994) Estimation of the Genetics and Conservation: A Reference for proportion of diploid males in populations of Managing Wild Animal and Plant Populations, Hymenoptera, Heredity 72, 219–227. Benjamin/Cummings Menlo Park, CA. Packer L., Owen R.E. (1990) Allozyme variation, Ralls K., Ballou J.D., Templeton A. (1988) Estimates linkage disequilibrium and diploid male produc- of lethal equivalents and the cost of inbreeding in tion in a primitively social bee Augochlorella mammals, Conserv. Biol. 2, 185–193. striata (Hymenoptera: Halictidae), Heredity 65, 241–248. Reed D.H., Frankham R. (2001) How closely corre- lated are molecular and quantitative measures of Packer L., Owen R. (2001) Population genetic aspects genetic variation? A meta-analysis, Evolution 55, of pollinator decline, Conserv. Ecol. 5, 4. 1095–1103. Packer L., Zayed A., Grixti J.C., Ruz L., Owen R.E., Reed D.H., Frankham R. (2003) Correlation between Vivallo F., Toro H. (2005) Conservation genet- ics of potentially endangered mutualisms: reduced fitness and genetic diversity, Conserv. Biol. 17, 230–237. levels of genetic variation in specialist versus gen- eralist bees, Conserv. Biol. 19, 195–202. Ross K.G., Fletcher D.J.C. (1986) Diploid male pro- Page R.E. (1980) The evolution of multiple mating duction - a significant colony mortality factor behavior by honey bee queens (Apis mellifera), in the fire ant Solenopsis invicta (Hymenoptera: Genetics 96, 263–273. Formicidae), Behav. Ecol. Sociobiol. 19, 283–291. Palmer K.A., Oldroyd B.P., Quezada-Euán J.J.G., Roubik D.W. (2001) Ups and Downs in Pollinator Paxton R.J., May-Itza W.D.J. (2002) Paternity fre- Populations: When is there a Decline? Conserv. quency and maternity of males in some stingless Ecol.5,2.[online]http://www.consecol.org/vol5/ bee species, Mol. Ecol. 11, 2107–2113. iss1/art2 (accessed on 2 February 2009). Bee conservation genetics 261

Roubik D.W., Weigt L.A., Bonilla M.A. (1996) Takahashi N.C., Peruquetti R.C., Del Lama M.A., Population genetics, diploid males, and limits to de Oliveira Campos L.A. (2001) A reanaly- social evolution of euglossine bees, Evolution 50, sis of diploid male frequencies in euglossine 931–935. bees (Hymenoptera: Apidae), Evolution 55, 1897– 1899. Sabeti P.C., Schaffner S.F., Fry B., Lohmueller J., Varilly P., Shamovsky O., Palma A., Mikkelsen Takahashi J., Ayabe T., Mitsuhata M., Shimizu I., Ono T.S., Altshuler D., Lander E.S. (2006) Positive M. (2008) Diploid male production in a rare and natural selection in the human lineage, Science locally distributed bumblebee, Bombus florilegus 312, 1614–1620. (Hymenoptera, Apidae), Insectes Soc. 55, 43–50. Saccheri I., Kuussaari M., Kankare M., Vikman P., Tavares M.G., Dias L.A.D., Borges A.A., Lopes Fortelius W., Hanski I. (1998) Inbreeding and ex- D.M., Busse A.H.P., Costa R.G., Salomao tinction in a butterfly metapopulation, Nature 392, T.M.F., Campos L.A.D. (2007) Genetic di- 491–494. vergence between populations of the sting- less bee uruqu amarela (Melipona rufiventris Schena M., Shalon D., Davis R.W., Brown P.O. (1995) group, Hymenoptera, Meliponini): Is there a new Quantitative Monitoring of Gene Expression Melipona species in the Brazilian state of Minas Patterns with a Complementary DNA Microarray, Gerais? Genet. Mol. Biol. 30, 667–675. Science 270, 467–470. Templeton A.R., Read B. (1994) Inbreeding: one word, several meanings, much confusion, in: Loeschcke Schmid-Hempel P., Schmid-Hempel R., Brunner P.C., V., Tomiuk J., Jain S.K. (Eds.), Conservation Seeman O.D., Allen G.R. (2007) Invasion success Genetics, Birkhäuser-Verlag Basel, Switzerland, of the bumblebee, Bombus terrestris, despite a pp. 91–106. drastic genetic bottleneck, Heredity 99, 414–422. The Honeybee Genome Sequencing Consortium Schremph A., Aron S., Heinze J. (2006) Sex determi- (2006) Insights into social insects from the nation and inbreeding depresion in an ant with reg- genome of the honeybee Apis mellifera,Nature ular sib-mating, Heredity 97, 75–80. 443, 931–949. Shendure J., Mitra R.D., Varma C., Church G.M. Twyman R.M. (2004) SNP discovery and typing tech- (2004) Advanced sequencing technologies: meth- nologies for pharmacogenomics, Curr. Top. Med. ods and goals, Nat. Rev. Genet. 5, 335–344. Chem. 4, 1423–1431. Vamosi J.C., Knight T.M., Steets J.A., Mazer S.J., Simmonds F.J. (1947) Improvement of the sex ratio of Burd M., Ashman T.-L. (2006) Pollination de- a parasite by selection, Can. Entomol. 79, 41–44. cays in biodiversity hotspots, Proc. Natl Acad. Sci. Spielman D., Brook B.W., Frankham R. (2004) Most USA 103, 956–961. species are not driven to extinction before genetic van ’t Veer L.J., Bernards R. (2008) Enabling person- factors impact them, Proc. Natl Acad. Sci. USA alized cancer medicine through analysis of gene- 101, 15261–15264. expression patterns, Nature 452, 564–570. van Wilgenburg E., Driessen G., Beukeboom L. (2006) Stouthamer R., Luck R.F., Werren J.H. (1992) Genetics Single locus complementary sex determination in of sex determination and the improvement of Hymenoptera: an “unintelligent” design? Front. biological control using parasitoids, Environ. Zool. 3, 1–15. Entomol. 21, 427–435. Vilas C., San Miguel E., Amaro R., Garcia C. (2006) Stow A., Silberbauer L., Beattie A.J., Briscoe D.A. Relative contribution of inbreeding depression and (2007) Fine-scale genetic structure and fire- eroded adaptive diversify to extinction risk in created habitat patchiness in the Australian allo- small populations of shore Campion, Conserv. dapine bee, Exoneura nigrescens (Hymenoptera: Biol. 20, 229–238. Apidae), J. Hered. 98, 60–66. Wadlow R., Ramaswamy S. (2005) DNA microarrays Strange J.P., Knoblett J., Griswold T. (2009) DNA in clinical cancer research, Curr. Mol. Med. 5, amplification from pin-mounted bumble bees 111–120. (Bombus) in a museum collection: effects of frag- Wandeler P., Hoeck P.E.A., Keller L.F. (2007) Back to ment size and specimen age on successful PCR, the future: museum specimens in population ge- Apidologie, DOI: 10.1051/apido/2008070. netics, Trends Ecol. Evol. 22, 634–642. ff Strassmann J. (2001) The rarity of multiple mating by Wang J. (2005) Estimation of e ective population sizes females of social Hymenoptera, Insectes Soc. 48, from data on genetic markers, Philos. Trans. R. 1–13. Soc. Lond. B Biol. Sci. 360, 1395–1409. Werren J.H. (1993) The evolution of inbreeding in Strausberg R.L., Levy S., Rogers Y.-H. (2008) haplodiploid organisms, in: Thornhill N.W. (Ed.), Emerging DNA sequencing technologies for hu- The natural history of inbreeding and outbreed- man genomic medicine, Drug Discov. Today 13, ing: theoretical and empirical perspectives, The 569–577. University of Chicago Press Chicago, pp. 42–59. 262 A. Zayed

Westemeier R.L., Brawn J.D., Simpson S.A., Esker Determination In Hymenopteran Parasitoids And T.L., Jansen R.W., Walk J.W., Kershner E.L., Its Implications For Biological Control., Entomol. Bouzat J.L., Paige K.N. (1998) Tracking the long- Sin. 10, 81–93. term decline and recovery of an isolated popula- Yokoyama S., Nei M. (1979) Population dynamics of tion, Science 282, 1695–1698. sex-determining alleles in honey bees and self- Whitfield C.W., Band M.R., Bonaldo M.F., Kumar incompatibility alleles in plants, Genetics 91, 609– C.G., Liu L., Pardinas J.R., Robertson H.M., 626. Soares M.B., Robinson G.E. (2002) Annotated Zayed A. (2004) Effective population size in Expressed Sequence Tags and cDNA Microarrays Hymenoptera with complementary sex deter- for Studies of Brain and Behavior in the Honey mination, Heredity 93, 627–630. Bee, Genome Res. 12, 555–566. Whitfield C.W., Behura S.K., Berlocher S.H., Clark Zayed A., Packer L. (2001) High levels of diploid A.G., Johnston J.S., Sheppard W.S., Smith D.R., male production in a primitively eusocial bee Suarez A.V., Weaver D., Tsutsui N.D. (2006) (Hymenoptera: Halictidae), Heredity 87, 631– Thrice Out of Africa: Ancient and Recent 636. Expansions of the Honey Bee, Apis mellifera, Zayed A., Packer L. (2005) Complementary sex deter- Science 314, 642–645. mination substantially increases extinction prone- Whitlock M.C., Barton N.H. (1997) The effective size ness of haplodiploid populations, Proc. Natl Acad. of a subdivided population, Genetics 146, 427– Sci. USA 102, 10742–10746. 441. Zayed A., Packer L. (2007) The population genetics Widmer A., Schmid-Hempel P. (1999) The pop- of a solitary oligolectic sweat bee, Lasioglossum ulation genetic structure of a large temperate (Sphecodogastra) oenotherae (Hymenoptera: pollinator species, Bombus pascuorum (Scopoli) Halictidae), Heredity 99, 397–405. (Hymenoptera: Apidae), Mol. Ecol. 8, 387–398. Zayed A., Whitfield C.W. (2008) A genome-wide sig- Wilfert L., Gadau J., Baer B., Schmid-Hempel P. nature of positive selection in ancient and recent (2007a) Natural variation in the genetic architec- invasive expansions of the honey bee Apis mellif- ture of a host-parasite interaction in the bumblebee era, Proc. Natl Acad. Sci. USA 105, 3421–3426. Bombus terrestris, Mol. Ecol. 16, 1327–1339. Zayed A., Roubik D.W., Packer L. (2004) Use of Wilfert L., Gadau J., Schmid-Hempel P. (2007b) The diploid male frequency data as an indicator of pol- genetic architecture of immune defense and re- linator decline, Proc. R. Soc. Lond. Ser. B Biol. production in male Bombus terrestris bumblebees, Sci. 271, S9–S12. Evolution 61, 804–815. Zayed A., Packer L., Grixti J.C., Ruz L., Toro H., Williams N., Minckley R., Silviera F. (2001) Variation Owen R.E. (2005) Increased genetic differentia- in native bee faunas and its implications for detect- tion in a specialist versus a generalist bee: impli- ing community changes, Conserv. Ecol. 5, 7. cations for conservation, Conserv. Genet. 6, 1017– Woyke J. (1979) Sex determination in Apis cerana in- 1026. dica, J. Apic. Res. 18, 122–127. Zayed A., Constantin S.A., Packer L. (2007) Wu Z., Hopper K.R., Ode P.J., Fuester R.W., Chen Successful biological invasion despite a se- J., Heimpel G.E. (2003) Complementary Sex vere genetic load, PLoS ONE 2, e868.