Copyright

by

Natalia Beatriz Biani

2009

The Dissertation Committee for Natalia Beatriz Biani Certifies that this is the

approved version of the following dissertation:

Alliances and Struggles in the Miniature Ecosystem of a Socially

Flexible Bee

Committee:

Ulrich G. Mueller, Co-Supervisor

Larry E. Gilbert, Co-Supervisor

Mike C. Singer

Lauren Ancel Meyers

John L. Neff

Alliances and Struggles in the Miniature Ecosystem of a Socially Flexible Bee

by

Natalia Beatriz Biani, M.S.

Dissertation

Presented to the Faculty of the Graduate School of

The University of Texas at Austin

in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy

The University of Texas at Austin

August, 2009

Dedication

To my parents,

for their gift of life.

To Evangelina, Marianela and Juan Francisco, the most marvelous siblings.

To Juan, my love.

To our son, Ignacio, who is loved even before existing.

v

Acknowledgements

Doing a PhD. is not a work of months, not even a year… it involves many years,

many moments, many feelings, many struggles. And is not the work of a single person, it

is a relatively collaborative process.

This dissertation work would not have been accomplished without the guidance,

support, and encouragement of Bill Wcislo and Ulrich Mueller. I am infinitely grateful to

Bill, who in the midst of uncertainty suggested me to “watch those mites” in Megalopta

bee nests. Bill is not only a brilliant and passionate scientist, but more importantly, he is a

fantastic person. When I met him for the first time, we had a long talk, about graduate

school and science, and it was the most peaceful and friendly talk I have ever had with a

scientist. I am thankful to you for teaching me how to do science, for being always so

supportive, understandable and kind person and for granting me the privilege to get you

know you.

I am extremely thankful to Ulrich, who timidly accepted me as one of his

students, and rescue me from dropping this whole PhD. project. I am immensely grateful

to you for your essentially kind and clever advice and your intellectual and moral support

and friendship over these years. Thank you for allowing me to work in your lab and for

honoring me with the privilege of being a “muellerita”.

I thank Larry Gilbert, who opened the doors of UT and who introduced me to the

beauty of tropical forests. The other members of my committee, Lauren Meyers, Jack

Neff and Mike Singer were plentiful sources of support, advice and passion throughout

this learning process.

vi

I am indebted to Juan Carlos di Trani and Fernando Soley for training me on

finding and opening Megalopta’s nests. Field work was both possible and enjoyable

thanks to the invaluable help, company and friendship of Hermogenes Fernandez-Marin,

Adam Smith, Simon Tierney, Therany Gonzales, Marcos Rios, Cecilia Blundo, Gonzalo

Rivas, Piotr Lukasik, and Juan Carlos Santos. Greg Gilbert was a wonderful fungi mentor

and taught me the delicacies of these unique and ubiquitous organisms. Hermógenes

Fernández-Marín and Julia Sarco provided essential logistic help and advice in the

Gamboa mycology lab.

Hans Klompen and Jon Shik were essential with the identification of the mite

species. Andre Rodrígues suggested the best microbiology method to test the cleaning

effect of mites.

The Smithsonian community in Gamboa is the most stimulating, supporting and

fun environment one can ever hope for during field work. I am thankful to the ones I had

the privilege to overlap with. I am grateful to the Ryan Lab for accepting me as a “social

parasite” during our summers in Gamboa. I am obliged to the Smithsonian helpful and

friendly staff for their logistical support and for their help with the collection and

exportation permits. I want to give special thanks to Paola Galgani who has not only

allowed this research to be possible, but has been a wonderful friend. Oris Acevedo and

Belkys Jimenez are amazingly accommodating and caring persons. Oris, I will never

forget the afternoon we returned from BCI in a little boat, avoiding transatlantic ships

under a heavy tropical rain.

The family who I stayed with in Gamboa, were constant source of laugh, support

and gifted me with the thrilling experience of driving Emiliana all the way to Panama

City while she was in labor.

vii

The Mueller Lab is the most wonderful, supportive and fun atmosphere ever. I am

thankful to all of you: Barrett Klein and Christian Rabeling, for being such enjoyable,

funny and delightful friends and office mates, Rachelle Adams for her support during my

first years and beyond, Alexander Mikheyev for his help and advice with the AFLP data

and for his constant compliments about my baking skills, Ruchira Sen, Sze Huei, Heather

Ishak the sweet and loving ladies of the lab, and Mike Cooper for his help during

molecular work.

Sandy Monahan could not be sweeter and more compassionate. I thank her for her

friendship and for being always so helpful and kind. I will never forget the day you

invited me to visit the tower!

My friends in Austin and in Argentina are essential to me. I want to specially

thank my friend Natalia Ruiz Juri, with whom I came to Austin for the first time, and

who I lived with during the first mesmerizing and overwhelming years of grad school in a

foreign country. My friend Lidia Marte, la “Arania”, with whom I share innumerable

moments of joy, fear, nausea, hope, etc about grad school and life, who always gives me

the anthropologist perspective and the beauty of her art, and with who I share the

indescribable feelings of being a latinoamerican.

My friends from the EEB cohort of the fall of 2002 and beyond are fundamental

pillars enriching my life with theirs, and with their extravagant and delicious dishes from

all over the world: Samraat, Catalina, Barrett, Simone, Christian, Juanita, Deepa, Evan,

Sasha. I could not have made more marvelous friends. I thank Rachel Page, Ximena

Bernal and Rachelle Adams, with whom I shared the adventures of grad school, field

work, friendship and the amazing journey to motherhood.

I am grateful to the music and the blessed musicians I love, for they make our life

and our sometime tedious work in the lab, enjoyable and for they bring us awareness of

viii

unique memories. Borges, Galeano, Cortazar, Fontanarrosa, Maitena, Girondo, Gelman,

Benedetti, Neruda, Garcia Marquez, Les Luthiers, are essential to rescue me from the

hard questioning moments of grad school, and are sublime sources of inspiration.

I feel eternal gratitude to my parents, Ines and Ricardo, who gifted me with the

miracle of life, who taught me the love for the natural world and who instilled me the

value of learning. I am immensely thankful to Evi, Marianela y Juan Francisco with

whom I share the life, for their unique singularity, for their love and undoubtable

fraternity. I thank their beloved ones, for making them truly happy. I am grateful to my

grandmas for being so “grandmas”. I will never be able to thank all of them enough for

their continual, essential and unconditional support and immense love; for giving me the

most grandiose welcoming every time I went back to Rosario; and for the wonderful

miracle of being present, connected and sharing the whole year throughout time and

distance.

Juan has being my friend, my guide, my extra advisor, my field assistant, my lab

assistant, my personal statistical consultant, my fervent supporter, and above all, the love

of my life. I could not have done this without you. Ignacio, our son, is the most

wonderful being in my life and I thank him for being such a sweet and patient baby.

Finally, I thank the beautiful bees and the amazing mites for allowing me to peep

at their lives.

This work was supported by the Smithsonian Tropical Research Institute (STRI)

Short Term Fellowships and especially by the research grants from the Program of

Ecology, Evolution and Behavior since without those little fellowships, none of this

would have even started. I am grateful for the teaching assistantships that were granted to

me during all these years. Being an independent international student, I would have been

unable to support myself otherwise.

ix

Alliances and Struggles in the Miniature Ecosystem of a Socially

Flexible Bee

Publication No._____________

Natalia Beatriz Biani, PhD.

The University of Texas at Austin, 2009

Supervisor: Ulrich G. Mueller

Lawrence E. Gilbert

Cooperation is pervasive in nature but paradoxically also provides opportunity to

cheaters. My dissertation involves the study of both cooperation and conflict in two

species of Megalopta bees. Megalopta is a Neotropical genus of halictid bees whose

biology is characterized by complex life cycles that can range from solitary to eusocial.

These bees nest in dead wood and forage under dim light conditions. Megalopta’s nests

are inhabited by an extensive array of organisms and each nest therefore constitutes a

miniature ecosystem providing opportunities for cooperation and conflict, both within

and between species. I first delineate the social structure of M. genalis and M. ecuadoria

nests in several Panamanian populations and integrate the factors that play a role in the

behavioral decisions of females when joining a social group or not. Within a kin-selection

framework, I discuss how genetic relatedness plays a role in the formation of social nests.

Second, I investigate the conflict between host bees and a congener social parasite, and I

elucidate reproductive structures that are relevant for understanding the evolution of

x

parasitism. Finally, I describe a cleaning mutualism between Megalopta bees and their

mite associates. Bee-mite associations encompass a broad spectrum of interspecific

interactions. Some bee-mites are thought to perform cleaning services for their hosts in

exchange for suitable environments for reproduction and dispersal. Field observations

and experimental manipulation reveal a significant correlation between the presence of

mites and the absence of fungi inside the brood cells, as well as between the absence of

mites and increased bee mortality. This study therefore provides evidence of the sanitary

effect of mites in nests of Megalopta bees. This bee-mite association constitutes one of

the few examples of terrestrial cleaning mutualisms.

xi

Table of Contents

List of Tables ........................................................................................................ xii

List of Figures ...................................................................................................... xiii

Chapter 1: Introduction ...........................................................................................1

Chapter 2: Nestmate relatedness in the socially flexible bees Megalopta genalis and Megalopta ecuadoria ......................................................................................5

INTRODUCTION ..........................................................................................5

METHODS .....................................................................................................8

RESULTS .....................................................................................................12

DISCUSSION ...............................................................................................15

CONCLUSION .............................................................................................17

ACKNOWLEDGEMENTS ..........................................................................19

Chapter 3: Notes on the reproductive morphology of the parasitic bee Megalopta byroni (Hymenoptera: Halictidae), and a tentative new host record ............27

Chapter 4: Cleaner mites: sanitary mutualism in the miniature ecosystem of Neotropical bee nests ....................................................................................32

ABSTRACT ..................................................................................................32

INTRODUCTION ........................................................................................33

MATERIALS & METHODS .......................................................................36

RESULTS .....................................................................................................39

DISCUSSION ...............................................................................................41

ACKNOWLEDGMENTS ............................................................................45

References ..............................................................................................................50

Vita 63

xii

List of Tables

Table 2.1: AFLP scoring errors statistics. ..............................................................19

Table 2.2: Pairwise Fst values between populations of M. genalis ........................20

Table 2.3: Relatedness coefficient statistics for the M. genalis and M. ecuadoria

populations studied. ..........................................................................20

Table 2.4: Bivariate Pearson correlation coefficients, Spearman’s correlation

coefficients and significance level for several variables of M. ecuadoria

and M. genalis. ..................................................................................20

Table 2.5: Cluster analysis of the reproductive variables of M. genalis and M.

ecuadoria. .........................................................................................21

Table 2.6: Means and standard error of morphological variables measured in females

of M. ecuadoria and M. genalis. .......................................................21

Table 2.7: Mating frequency and number of patrilines estimated for each population

of M. ecuadoria and M. genalis. .......................................................22

xiii

List of Figures

Figure 2.1: Map of collection sites in Panama. ......................................................23

Figure 2.2: Bivariate Pearson correlation between the head width and reproductive

variables of M. ecuadoria and M. genalis and M. ecuadoria. ..........24

Figure 2.3: Frequency distribution of the proportion of shared markers between

reproductive individuals and their nest-mates (brood and non-

reproductive adults), and among unrelated individuals in each population

of M. ecuadoria and M. genalis. .......................................................25

Figure 4.1: Megalopta nest structure and cleaning effects of the mites. ...............46

Figure 4.2: Deutonymphs (dipersal stage of the mites) on thorax and metasoma of

Megalopta pupa. ...............................................................................47

Figure 4.3: Artificial cell with a bee larva. ............................................................47

Figure 4.5. Means ± 2 Standard Errors of Colony Forming Units in the four

experimental treatments. ...................................................................49

1

Chapter 1: Introduction

Despite the inherently selfish character of natural selection, cooperation is

pervasive in the natural world but, paradoxically, provides opportunity to cheaters

(Bronstein 2001; Sachs et al. 2004). Moreover, cooperation and conflict cross several

level of organization from within a single cell to within and between species (Maynard

Smith & Szathmary 1997). Cooperation is quite conspicuous in insect societies, however,

such alliances do not preclude the existence of intrasocietal conflicts (Ratnieks et al.

2006).

The most remarkable example of cooperation within species of eusocial

organisms is reproductive altruism, where sterile workers forego their own reproduction

in order to dedicate their lives to rearing siblings (Wilson 1971). Not coincidentally, one

of the most widespread intraspecific conflicts in hymenopteran social groups is over the

control of reproduction. There exists a vast literature regarding how relatedness

asymmetry, the differential degree of relatedness between males and females, causes

conflict in sex allocation, male parentage, and queen worker-conflict over reproduction

(reviewed in Bourke 2005; Ratnieks et al. 2006).

Mutualisms are reciprocally advantageous interactions among individuals of

different species. Such interspecific cooperation is ubiquitous in nature and is one of the

evolutionary driving forces in ecological communities (Maynard Smith 1989; Herre et al.

1999). More than 70% of all plant species rely on their relationships with mycorrhizal

2

fungi to obtain nutrients from the soil (Selosse et al. 2006). Similarly, most flowering

plants (90%) depend on animals for transporting pollen between flowers (Fægri & van

der Pijl 1979). The examples are countless.

Conflicts between species are best exemplified by predation, such as the deceptive

behavior of the bolas spider that not only spins their typical web, but also actively hunt

moths by sending specific pheromones that mimics those of their prey (Gemeno et al.

2000). Likewise, there are conflicts in animal communication when predator and parasite

eavesdroppers take advantage of the honest signals used by their preys to attract mates

(Bernal et al. 2007).

My dissertation work involves the study of both cooperation and conflict in two

species of Megalopta bees (Hymenoptera: Halictidae). Megalopta is a Neotropical genus

of sweat bees (Moure & Hurd 1987; Engel 2000). Like many other species in the family,

Megalopta is characterized by complex life cycles that can range from solitary to

eusocial. Unlike most bees, they forage under dim light conditions taking advantage of

dawn and crepuscular blooms (Wcislo et al. 2004). Nests are constructed in dead

branches or lianas intertwined in the vegetation, and consist of a single excavated tunnel

with individual brood cells made of chewed wood (see Sakagami & Michener 1962;

Wcislo et al. 2004). Brood cells are provisioned with a mass formed from nectar and

pollen. Following oviposition, the cell is sealed, and brood development from egg to

adult eclosion takes about 35 days. Fungi are common in bee nests, including inside cells

for developing brood (Batra 1965; Batra et al. 1973; Inglis et al. 1993; Hajek & Stleger

1994). Megalopta’s nests are inhabited by an extensive array of organisms and each nest

3

thus constitutes a miniature ecosystem providing opportunities for cooperative and

antagonistic interactions, both within and between species.

In Chapter 2, I delineate the social structure of M. genalis and M. ecuadoria nests

in several Panamanian populations. Given the social flexibility of these species, every

new female that emerges in a nest has the opportunity to decide between staying in the

natal nest, helping their nest-mates and even reproducing, or start a new nest of her own.

There are benefits and drawbacks in whatever decision the bee makes. Multifemale nests

have higher survivorship because they are better defended against ant predation (Smith et

al. 2003). Based on the framework of kin selection (Hamilton 1964), the cost and benefits

of group living are two of the factors that need to be taken into account in order to

understand the evolution of sociality. In Chapter 1, I outline the importance of the other

parameter involved in the evolution of social groups, the degree of relatedness within

nests. I calculated relatedness from information provided by AFLP markers, and discuss

the possible social outcomes in these populations as well as several aspects of the

reproductive biology of these bees.

Chapter 3 investigates the occurrence of brood parasitism in Megalopta, a

common phenomenon not only in nest-making Hymenoptera (Wcislo 1987; Michener

2000; Stuart 2002) but in vertebrate groups as well (Jones 1997; Hauber et al. 2004). M.

byroni is one of the congeneric parasites of these bees. It is extremely rare and its biology

is almost unknown. Here, I provide notes on reproductive structures that are relevant for

understanding the evolution of parasitism, and provide a tentative new host record for M.

4

byroni in Panama. I also discuss how specific aspect of the biology of this parasite might

contribute to its rarity and the appropriateness of classifying it as a social parasite, given

the social flexibility of the host bees.

Finally, Chapter 4 describes an astonishing example of cooperation between

Megalopta bees and their mite associates. One of the most common arthropods inhabiting

social insect nests are mites (Arachnida: Acari) (Eickwort 1990, 1993). Bee-mite

associations encompass a broad spectrum of interspecific interactions. Some bee-mites

are thought to perform cleaning services for their hosts in exchange for suitable

environments for reproduction and dispersal (Ordway 1964; Eickwort 1979). Using field

observation and experimental manipulation, I demonstrate that there is a significant

correlation between the presence of mites and the absence of fungi inside the brood cells,

as well as between the absence of mites and increased bee mortality. This study therefore

provides clear evidence of the sanitary effect of mites in nests of Megalopta bees. This

bee-mite association constitutes one of the few examples of terrestrial cleaning

mutualisms.

5

Chapter 2: Nestmate relatedness in the socially flexible bees Megalopta genalis and Megalopta ecuadoria

INTRODUCTION

Kin selection has recently been criticized as a sufficient explanation for the origin

and maintenance of eusociality (Wilson 2005; Wilson & Hölldobler 2005; Wilson 2008a;

Wilson 2008b), challenging a large body of evidence arguing for the role of genetic

relatedness in the evolution of interactions in socially complex organisms or microbial

communities (Jarvis 1981; Queller & Strassmann 1998; Faulkes & Bennett 2001; Griffin

& West 2003; Bourke 2005; Gilbert et al. 2007) . Those defending the utility of kin

selection argue that alternative explanations of eusociality (various forms of group

selection) merely constitute a semantic rephrasing of selective processes that are best

viewed as kin selection (Foster et al. 2006; West et al. 2007). A recent analysis of the

importance of relatedness at the transition from solitary to eusocial life concluded that

this transition occurred only in societies with singly-mated queens (high relatedness),

whereas multiple mating (low relatedness) is invariably a derived condition arising within

eusocial lineages. Hence, relatedness was essential in the evolution of eusociality

(Hughes et al. 2008).

The family of sweat bees (Hymenoptera, Halictidae) is one of most socially labile

families within the order Hymenoptera, with obligate eusociality arisen (and lost)

repeatedly within the family. Eusociality appear to have arisen five times among all the

bees and three of these origins occurred within the family Halictidae (Danforth 2002).

Moreover, sweat bees feature the greatest variability in life cycle and sociality of any bee,

6

sometimes even between and within populations of the same species (Wcislo 1997;

Wcislo & Danforth 1997; Michener 2000; Danforth 2002).

Megalopta is a neotropical genus of sweat bees (Hymenoptera: Halictidae)

(Moure & Hurd 1987) that exhibit diverse social life cycles with alternative behavioral

strategies between solitary and social life (Wcislo et al. 2004; Tierney et al. 2008),

providing a window to view stages in the evolution of sociality. Unlike the typical

ground-nesting species in the sweat bee family, Megalopta nests are constructed in dead

branches or lianas intertwined in the profuse vegetation of primary and secondary forests.

Megalopta nests are founded generally by a single female, however there is some

evidence that co-founding by more than one female occurs in about 25-30% of the nests

(Wcislo & Gonzalez 2006). At the beginning of the Panamanian dry season (mid

December-mid January), nests are mostly solitary. As the season progresses, some nests

become social (with more than one adult female), most likely because some of the

emerging females decide to stay as helpers; other females might disperse and start new

nests on their own. Males are more abundant at the beginning of the dry season, and most

females are inseminated by the second half of the wet season (Wcislo et al. 2004).

Solitary and social nests constitute two categorically distinct behaviors and are not

developmental stages of a continuum. That is, nests that are solitary at the beginning of

the season tend to remain solitary at the end of the season. Similarly, social nests remain

with several females at the end of the season (Smith et al. 2007). The founder female can

live an average of 97.1 ± 41.5 days, which is considerably longer than the mean egg to

adult development time of about 35 days (Wcislo & Gonzalez 2006). Another estimate

suggests that solitary females have one third chance of dying before their offspring

emerge (Smith et al. 2007). Unrelated individuals joining nests or usurping nests have not

7

been reported (Wcislo et al. 2004), and unfortunately nothing is known about the mating

behavior of these bees.

In M. genalis, protection against ant predation and the risk taken by foraging

adults seem to be strong selective forces favoring group living, consistent with the

Assured Fitness Returns model (Queller 1989; Gadagkar 1990; Smith et al. 2003). Under

this model, offspring depend on extended parental care in order to reach independence,

and given a high risk of adult mortality, social groups composed of foundresses and

helpers are favored over solitary nests (i.e., offspring survival is increased, or “assured”

by the care provided by helpers). Solitary nesters are likely to lose all their investment

because, if they die, they are not able to bring their offspring to independence. Smith et

al.’s (2003) study on Megalopta is one of the few to assess the costs and benefits of

social versus solitary life in a mass provisioning bee (Eickwort et al. 1996; Kukuk et al.

1998). The AFR model assumes some degree of relatedness between brood and workers

(Queller 1989).

We investigated the social status of several populations of M. genalis and M.

ecuadoria in Panama to assess the importance of within-nest relatedness in the evolution

of sociality of these bee species. Because the number of mates of a reproductive female

might influence the degree of relatedness within her nest, we estimated the effective

mating frequency and determined if females mate singly or multiply in each of these two

Megalopta species. We inferred reproductive dominance relationships in multifemale

nests from morphological data on several reproductive physiology variables.

8

METHODS

Nest collections: Nests were collected in Soberanía National Park, Barro Colorado

Island Natural Monument (BCI), and Bocas del Toro Province (Republic of Panamá),

during June to August of 2005, December 2005 to January 2006, March to July 2006, and

May to June of 2007. Nest collection sites were close to each other within each

population (about 2.5 Km. apart), but distant between populations (Fig. 1, see

Delineation of populations, below). Because Megalopta have crepuscular foraging habits

(Wcislo et al. 2004), nests were collected during the day in order to assure capture of all

resident bees inside the nests. Bees were individually preserved in 100% ethanol shortly

after collection. Ecological information on these nests is reported in Biani et al. (2009).

DNA extraction: A piece of muscle tissue was excised from the mesosoma to

extract DNA. Use of thoracic muscle minimizes contamination with DNA from potential

parasites (most frequently in the head or abdomen) or from food material in the crop or

gut in the metasoma. The DNA extraction was performed with a Bioneer AccuPrep

Genomic DNA Extraction Kit supplier, and the extracted DNA was diluted to 0.01µg/µl.

AFLP marker development: We used a modified version of the Applied

Biosystems (ABI) AFLP protocol for small plant genome

(http://www.appliedbiosystems.com, Protocol #: 4303146 Rev E). The DNA was

digested with EcoRI and Msel restriction enzymes. The EcoRI and Msel adaptor pairs

were annealed using 10 µl of 10XT4 DNA ligase buffer, 10 µl of 0.5 M NaCl, 5 µl

(1mg/ml) BSA, 100 Units Msel, 500 Units EcoRI, and 100 Units T4 DNA ligase. The

restriction-ligation reaction was incubated at 37°C for 3 hours instead of the

recommended 2 hours to maximize production of template for amplification. Each

restriction-ligation product was diluted in 60 µl of pure water instead of the

recommended 189 µl of TE buffer. For the pre-selective and selective amplification, we

9

used 2 µl of 10X Taq buffer, 1.6 Units of Taq, 10mM dNTPs, all from New England

Biolabs. The amplification of the pre-selective PCR was repeated for 25 cycles of 20s at

94°C, 30s at 56°C, and 2 min at 72°C. The pre-selective PCR template was diluted by

adding 40 µl of water to 10 µl of the pre-selective product.

Primers pairs for selective amplification were EcoR1-TC x Msel-CAA, EcoR1-

AA x Msel-CTC, and EcoR1-AT x Msel-CTG (Applied Biosystems). The samples were

amplified by 11 cycles starting with 20s at 94°C, 30s at 66°C, and 2 min at 72°C, but

decreasing the annealing temperature by 1°C each cycle until it reached 56°C. The final

selective PCR products were combined with 0.7 µl of Chimerex GeneFloTM 625 (Rox

label) as a size standard and 7.6 µl of Hi-Di formamide, denatured at 95°C for 5 minutes,

and subsequently analyzed in an ABI Prism 3100 Genotyper.

Scoring of AFLP markers: The AFLP banding pattern was scored with the

program GeneMarker (AFLP/Genotyping ProgramVersion 1.51. Softgenetics LLC®,

State College, PA, 16803). A scoring panel for each primer combination was created

where reliable alleles were first chosen from among all bands in an AFLP’s profile, then

all samples were scored under that common pattern. We followed the recommendations

of the software manual to create the panels and to run the program (SoftGenetics 2006).

The two species were analyzed separately. A total of 236 markers was scored with 3

primers combinations for both species.

We pursued the AFLP scoring strategy suggested by Bonin et al. (2007). We

minimized the risk of homoplasy (non-homologous fragments migrating to the same

position in the electrophoretic profile) by scoring only the fragments above 100 base

pairs and,. With the exceptions of M. ecuadoria population from BCI (12 individuals)

and M. genalis from Bocas del Drago (16 individuals), in all the other populations the

number of individuals scored is greater than 30 for each species, which is the

10

recommended number (Bonin et al. 2007). The scoring was very conservative because

we only used markers that could be scored unambiguously across all samples; markers

that could not be scored reliably as present/absent were excluded from the analysis.

Fifteen samples were processed blindly starting with DNA extraction. The

samples were also blindly scored with respect to colony and population. The percentage

of markers that differ between original and the blind rescoring correspond to the error

rate in scoring the AFLP markers. We were able to unambiguously score 236 markers

among the three primer combinations. We discarded 265 additional markers because they

could not be scored unambiguously. For the 236 reliable markers, the average scoring

error in the blind rescoring was 2.51 % (Table 1.1), which is comparable to error rates of

microsatellite markers (Okello et al. 2005; Morin et al. 2007).

Delineation of population boundaries: Before calculating within-nest relatedness,

we defined the populations within which mating can be assumed to occur at random,

using the program AFLPsurv (Vekemans et al. 2002). We used the option that assumes

Hardy-Weinberg conditions within each population and non-uniform prior distribution of

allele frequencies for the Bayesian analysis of the matrices (Zhivotovsky 1999).

Relatedness calculation: The coefficient of relatedness was calculated following

Reeve et al. (1992). Because this method obtains a single relatedness estimate without

confidence intervals, we performed a bootstrap analysis to estimate the standard error of

that single relatedness estimate. One hundred matrices were created in Mesquite by

bootstrap re-sampling of markers from the original matrix (Maddison & Maddison 2009),

calculating the average relatedness for each sub-sampled matrix, and then deriving the

standard error across these 100 bootstrap averages.

Morphological measurements: Several morphological variables were measured

for each adult bee collected, including head width, intertegula distance, largest ovariole

11

width, Dufour gland width, and number of yellow bodies (degenerating nursing cells that

indicate a recent egg laying history). These variables are known to correlate with the

reproductive status of bees (Michener 1974; Arneson & Wcislo 2003), and therefore

allow identification of the primary reproductive in each nest, as well as current and past

reproductive activity. Bees were measured and dissected using an ocular micrometer

mounted on a stereomicroscope. Bivariate Pearson correlation and non-parametric

Spearman correlation were performed to assess the relationship between morphometric

and reproductive variables. A cluster analysis determined which the reproductive

individuals in each nest were. All tests were performed in SPSS (SPSS 2006).

Estimation of mating frequency: Females that had the largest ovariole width, the

largest Dufour gland, and the largest number of yellow bodies were assumed to be the

primary reproductive in a given nest. The mating frequency was estimated in nests where

reproduction by a single female (rather than several possible females) could be verified

based on the above reproductive variables. AFLP markers present in the progeny but

absent in the primary reproductive female must have been inherited paternally. If these

sets of paternally-derived markers differ between brood, they might constitute an event of

polyandry. The frequency of the different sets of bands (pi) can then be used to estimate

the effective mating frequency in each nest [me=1/Σ(pi)2 (Starr 1984)]. We only

considered paternal sets of markers consisting of more than one marker in order to avoid

artificially created patterns that might be due to amplification or scoring errors of a single

band. Alternatively, dissimilar banding patterns among brood can be due to events of

queen supersedure (serial polygyny) when the original reproductive dies and another

female takes over the reproduction in the nest (Mueller 1996; Paxton et al. 2003; Bargum

et al. 2007). For the cases used in the mating-frequency analyses, we ruled out queen

supersedure by also calculating the percentage of markers shared among all individuals

12

within each nest (see below). An event of polygyny would be detected by dissimilar

banding pattern and low percentage of shared markers between the reproductive and

some of the brood.

Determination of potential intra-specific nest parasitism: We calculated the

proportion of markers that were shared between the primary reproductive and her

nestmates, as well as the proportion of shared markers among unrelated individuals in

each population. By comparing the frequency distribution of these shared markers, it is

potentially possible to detect cases of intra-specific nest parasitism. Identifying cases of

intraspecific next parasitism is critical because the calculation of the mating frequency

assumes correct identification of mother-daughter relationships within a given nest. We

only considered the nests where the primary reproductive could be unambiguously

identified and where workers did not seem to reproduce, based on the morphological and

physiological criteria described above. Moreover, for these nests we ruled out the

occurrence of nest usurpation, that is, when the primary reproductive is replaced by an

unrelated individual that takes over reproduction of the nest. Under supersedure, older

brood would have low proportion of band shared with respect to the identified primary

reproductive, whereas younger brood would share a higher proportion of bands.

Conversely, egg dumping (intra-specific nest parasitism) can be detected when a single

individual share a low proportion of markers with the primary reproductive.

RESULTS

Scoring of AFLP markers: The average number of fragments per individual was

70.6 and 85.4 for M. genalis and M. ecuadoria, respectively, and the corresponding total

13

number of segregating (polymorphic) fragments was 184 (78%) and 163 (69.1%),

respectively.

Delineation of population boundaries: We measured population-genetic

differences with the Fst statistic, which can be interpreted as the proportion of the total

genetic diversity that occurs among populations (Excoffier et al. 1992). The AFLPsurv

program creates a distribution of Fst through 1,000 random permutations of the matrix

under a null hypothesis of no population differentiation. For M. genalis, the lower and

upper 95% limits of the distribution of Fst were -0.0082 and 0.0300 respectively. The

observed Fst was 0.0963, which fell outside these limits and therefore allowed the

rejection of the null hypothesis of no population differentiation (p<0.001). The Fst

analyses defined three populations, Pipeline Road, BCI, and Bocas del Drago for M.

genalis (Bocas del Toro Province; Table 1.2, Fig. 1.1). Likewise, for M. ecuadoria the

observed Fst of 0.0943 fell outside the lower (-0.0147) and upper (0.0291) 95% limit;

hence we rejected the null hypothesis of no population differentiation (p<0.001). The two

defined populations were Pipeline Road and BCI (Fig. 1). The close proximity of the

collection sites within each population relative to the distance among populations

supports the interpretation that within-population gene flow is very likely for both

species. All subsequent analyses consider each population separately.

Within-nest relatedness: The genetic relatedness between individuals of the

same nest, as estimated by the method of Reeve et al. (1992), varied from 0.25 (± 0.002

SE) to 0.37 (± 0.002 SE) for the three populations of M. genalis, and from 0.31 (± 0.003

SE) to 0.37 (± 0.002 SE) for the two populations of M. ecuadoria (Table 1.3).

Morphological measurements: The Dufour gland width was significantly

correlated with head width in both species, while ovariole width was only significantly

correlated with head width in M. genalis, but not in M. ecuadoria (Fig. 1.2, Table 1.4).

14

Non-parametric correlation assessed the relationship between the number of yellow

bodies with the size of the bees. This correlation was significant only for M .genalis

(Table 1.4). The cluster analysis divided the individuals into reproductives and non-

reproductives based on values of the largest ovariole width, Dufour gland width, head

width and intertegula distance (Table 1.5). Most of the nests seemed to have a primary

reproductive (greatest ovariole development and largest Dufour gland development,

highest number of yellow bodies; Table 1.6). However, 19% (N=21) of both M.

ecuadoria and M. genalis nests may form more egalitarian societies, because all the adult

females in these nests showed comparable reproductive measures (i.e., all could be

potentially reproducing). In other nests, all the individuals were similar to each other but

the morphological values were comparable to non-reproductive individuals, suggesting

that none of them were reproducing yet (i.e., these nests did not have a primary

reproductive at the time of collection). This could be due to the recent loss of a

reproductive and the absence of a replacement individual.

Estimation of mating frequency: The effective mating frequency varied between

1.80-2.67 for the two populations of M. ecuadoria and between 1.00-3.00 for the three

populations of M. genalis (Table 1.7). Specifically, all M. ecuadoria brood in each nest

came from at least 2 patrilines (N = 7, range 2-4 patrilines); in addition, the effective

mating frequencies of all the females were above 2, except one that had an effective

mating frequency of 1.8. In M. genalis, only one nest out of eight seemed to have brood

from only a single patriline. Brood in all other M. genalis nests usually had 2 - 3

patrilines and most females had an effective mating frequency between 2 and 3. This

indicates that most of the reproductive females of both species were using the sperm of

several males. Because of our conservative approach in scoring the AFLP markers, it is

possible that actual mating frequencies are higher than the estimated values. On the other

15

hand, if our analyses failed to identify cases of supersedure of primary reproductives, our

results overestimate the actual mating frequencies.

Determination of potential intra-specific nest parasitism: The proportion of

markers shared among individuals of the same nest was above 87% for M. ecuadoria and

above 84% for M. genalis (Fig. 1.3 A&B). Based on the comparison with the distribution

of shared markers among unrelated individuals, there is no clear evidence for intra-

specific nest parasitism (egg dumping) except for one M. genalis’ nest from Bocas del

Drago where one individual shared only 70% of the markers compared to the rest of

individuals in that nest and the population in general (marker sharing of above 90%).

However, these estimates are based on only two nests from Bocas del Drago and hence

might not be representative of the population as a whole. The analysis of the shared

markers for all populations did not provide evidence of nest usurpation, because all brood

appeared to share a high proportion of markers with the primary reproductive, as

identified by the morphological and physiological data.

DISCUSSION

Our AFLP analyses indicate that cooperating nestmates of both M. ecuadoria and

M. genalis have a considerable degree of genetic relatedness (0.31-0.37 and 0.25-0.37,

respectively). Cooperating females are likely to be sisters or daughters of the primary

reproductive. Moreover, the relatedness estimates are within the range of other

primitively eusocial insects, including some Polistes wasps (Strassmann et al. 1989;

Francescato et al. 2002; Landi et al. 2003) and bees (Richards & Packer 1998; Stow et al.

2007). The fact that the relatedness estimate is lower than the 0.75 value expected for full

sisters in a single-mated single queen’s colony might be due to several factors, including

16

multiple mating, multiple reproductive females in a nest, and intraspecific nest

parasitism.

The estimated effective mating frequency and the existence of several patrilines in

most of our surveyed nests suggest that females are multiply mated, consistent with

studies of other halictid bees that also documented multiple mating (Page 1986; Kukuk

1989; Richards & Packer 1998). These studies were based on allozymes and behavioral

observations of mating. A more recent study using microsatellite markers in

Lasioglossum malachurum concluded that, even though most of the queens were singly-

mated, up to four patrilines could be detected in a single nest. The effective mating

frequency of L. malachurum varied between 1 and 3 with an average of 1.13 matings for

17 nests. L. malachurum therefore seems to be facultatively polyandrous or, the queens

seem to use sperm from only one male even if they had mated with several males (Paxton

et al. 2002). In another population of L. malachurum in Greece, eusocial colonies had

mostly single-mated queens (97 %; Richards et al. 2005). A very closely related bee,

Augochlorella striata, is the only halictid that seems to be clearly singly-mated, with

consequently high within-colony relatedness (Mueller et al. 1994).

Even though most nests had a female that dominated the reproductive output,

some nests (19%) of each species might be more egalitarian societies in which more than

one female reproduced in their nests. Based on the occurrence of aggressive interactions

among individuals of M. genalis, it seems that egalitarian societies are rare (Arneson &

Wcislo 2003); however, our morphological data suggest that they exist in a low

frequency. As in egalitarian societies of other halictid species, several matrilines might be

present in the nests (Wcislo 1997; Richards & Packer 1998; Richards 2000).

Although egg dumping has not been directly observed in Megalopta, it is a

common phenomenon among other bees (Kukuk et al. 1987; Wcislo 1987; Paxton et al.

17

2002), decreasing the relatedness between brood and adult females in affected nests. In

our study, we were unable to detect intra-specific nest parasitism, except possibly in a

single M. genalis nest in Bocas del Drago in which one of the individuals had a lower

proportion of shared markers and therefore may have been unrelated. This single

individual can also be the result of inter-specific parasitism by M. byroni (Biani & Wcislo

2007). Because we only found one such individual, it is impossible to reach any

conclusion about the existence of social parasitism in M. genalis.

Regarding the relationship between size and reproductive status of the bees, the

only significant associations were between ovary width and Dufour gland width with

head width in M. genalis, which is in agreement with a previous study on the same

species (Smith et al. 2008). The relationship between ovariole width and head width was

not significant in M. ecuadoria, perhaps because many large bees have small ovaries (the

so-called “pre-reproductive”) or alternatively, because smaller individuals become

dominant reproductives if they are the oldest (Arneson & Wcislo 2003; Smith et al. 2008;

Tierney et al. 2008). Other factors such as parasitism and cephalic allometry present in

these bees might affect the expression of body size and potentially the performance of

certain social roles (Sakagami & Moure 1965; Smith et al. 2008).

CONCLUSION

The evolution of social behavior is a very complex trait determined by intrinsic

and extrinsic factors (Evans 1977). Some of the relevant extrinsic factors driving the

evolution of sociality in Megalopta are predator pressure and risks of adult mortality

(Smith et al. 2003, 2007). Here we attempted to evaluate the importance of intrinsic

genetic factors in the formation of cooperative societies of M. genalis and M. ecuadoria

18

bees. Our results suggest that social cooperative groups are formed by related individuals

in M. genalis and M. ecuadoria; individual females of these species therefore may gain

inclusive fitness by staying in their natal nest rather than risking their fitness in an

inadequately defended nest of her own (Smith et al. 2003). Each social outcome might be

the product of a delicate combination of intrinsic and extrinsic factors mentioned above,

plus other suites of characteristics determined by the organisms themselves: intra-specific

interactions, the expression of dominance, their capability of nest mate recognition, and

competition for limited resources (Wcislo 1997). Because cooperative females are

related, this study provides additional support towards the possible importance of kin

selection in the evolution of sociality.

Even though environmental factors are known to play an important role in the

shape of the social structure of halictid nests (Eickwort 1966; Packer 1990; Soucy 2002;

Richards et al. 2005), most studies have been conducted on species from temperate

regions or at high altitudes (but see Wcislo et al. 1993; Soucy & Danforth 2002; Tierney

et al. 2008). In those species, the length of the reproductive season is determined mainly

by temperature, which constrains the existence of social groups. In Panama, the active

phase of nests is marked by the length of the dry season since there are fewer flower

resources to provision the nests as the rainy season progresses (Wright & Calderon 1995).

Furthermore, the rains affect the resilience of the nesting substrates specifically in

Megalopta, which in turn can also affect the expression of social behavior (Wcislo et al.

2004). The relatively long active season in central Panama generate opportunities for

several overlapping generations, which is a prerequisite for the formation of social

groups. This study constitutes another contribution to elucidate the social structure of

tropical halictids.

19

ACKNOWLEDGEMENTS

This research was funded by the Smithsonian Tropical Research Institute (STRI)

Short Term Fellowships and research grants from the Program of Ecology, Evolution and

Behavior of the University of Texas at Austin to NBB. We are extremely grateful to Juan

Carlos Santos for his statistical advice and help with the use of several software

programs; to Adam Smith, Hermógenes Fernández-Marín, Therany Gonzales and Juan

Carlos Santos for their invaluable help and company during nest collections; to Barrett

Klein, Sze Huei Yek, Ruchira Sen, Heather Ishak and Alexander Mikheyev for their

comments on earlier versions of the manuscript. We thank STRI staff for logistical

support. The Autoridad Nacional del Ambiente de la República de Panamá provided

collecting and export permits.

Table 2.1: AFLP scoring errors statistics.

Primer combination # of markers Average error Standard Deviation

Blue (EcoR1-TC/Msel CAA) 117 2.05 % 0.026 Yellow (EcoR1-AT/Msel

CTG ) 41 3.09 % 0.023

Green (EcoR1-AA/Msel CTC)

78 2.39 % 0.015

TOTAL 236 2.51 % 0.021 Error is estimated as the percentage of markers scored differently in blind repeat genotyping.

20

Table 2.2: Pairwise Fst values between populations of M. genalis

BCI B. Drago Pipeline BCI 0 Bocas Drago 0.1051 0 Pipeline 0.0544 0.1259 0

Table 2.3: Relatedness coefficient statistics for the M. genalis and M. ecuadoria populations studied.

Bootstrap average and standard errors of the relatedness coefficients estimated from 100 bootstrap subsamples.

Table 2.4: Bivariate Pearson correlation coefficients, Spearman’s correlation coefficients and significance level for several variables of M. ecuadoria and M. genalis.

Species Bivariate Pearson Correlations Coefficients Significance

M. ecuadoria Head width & Largest ovariole width 0.159 0.144 Head width & Dufour gland width 0.358 0.001

M. genalis Head width & Largest ovariole width 0.423 0.001 Head width & Dufour gland width 0.237 0.017

Non parametric Spearman

correlations Coefficients Significance

M. ecuadoria Head width & # yellow bodies 0.001 0.992 Largest ovariole width & # yellow bodies 0.480 0.001 Dufour gland width & # yellow bodies 0.478 0.001

M. genalis Head width & # yellow bodies 0.204 0.041 Largest ovariole width & # yellow bodies 0.458 0.001 Dufour gland width & # yellow bodies 0.318 0.015

Species Population # nests Real average

Bootstrap average

S.E.

M. ecuadoria Pipeline 13 0.369 0.340 0.002 M. ecuadoria BCI 3 0.309 0.314 0.003 M. genalis Pipeline 9 0.368 0.379 0.002 M. genalis BCI 10 0.348 0.339 0.002 M. genalis Bocas Drago 4 0.253 0.268 0.002

21

Table 2.5: Cluster analysis of the reproductive variables of M. genalis and M. ecuadoria.

The two clusters below are statistically significant for both species. M. genalis Intertegula

Distance Head Width Ovariole Width Dufour Width Mean SD Mean SD Mean SD Mean SD

Cluster 1 2.57 0.14 3.49 0.23 0.74 0.27 0.64 0.13

Cluster 2 2.22 0.22 2.82 0.28 0.36 0.19 0.45 0.17

Combined 2.36 0.26 3.09 0.42 0.51 0.29 0.52 0.18 M. ecuadoria

Intertegula Distance Head Width Ovariole Width Dufour Width

Mean SD Mean SD Mean SD Mean SD

Cluster 1 2.08 0.15 2.84 0.25 0.64 0.14 0.53 0.15

Cluster 2 1.93 0.13 2.51 0.16 0.35 0.13 0.33 0.14

Combined 1.98 0.15 2.63 0.25 0.45 0.19 0.40 0.17

Table 2.6: Means and standard error of morphological variables measured in females of M. ecuadoria and M. genalis.

Intertegula distance Head width

Largest ovariole width

Dufour gland width

mean SE mean SE mean SE mean SE M. ecuadoria (N=52) 1.98 0.02 2.63 0.04 0.45 0.03 0.40 0.02 M. genalis (N=58) 2.36 0.03 3.10 0.06 0.51 0.04 0.52 0.02

Measures are in mm.

22

Table 2.7: Mating frequency and number of patrilines estimated for each population of M. ecuadoria and M. genalis.

Species Population

Variables # Nests Min. Max. Mean Std. Error

M. ecuadoria

Pipeline Rd. Mating Freq. 6 1.80 2.67 2.12 0.13 # Patrilines 2 4 2.50 0.84

BCI Mating Freq. 1 N/A N/A 2.00* N/A # Patrilines N/A N/A 2* N/A

M. genalis

Pipeline Rd. Mating Freq. 4 1.60 2.00 1.85 0.10 # Patrilines 2 2 2.00 0.00

Bocas del Drago

Mating Freq. 2 1.00 3.00 2.00 1.00 # Patrilines 1 3 2.00 1.00

BCI Mating Freq. 2 1.80 2.00 1.90 0.10 # Patrilines 2 2 2.00 0.00

* estimate from a single M. ecuadoria nest in BCI.

23

Figure 2.1: Map of collection sites in Panama.

Bocas del Drago (Bocas del Toro Province), Pipeline Road, and Barro Colorado Natural Monument (BCI, Colon Province). Panama Canal Zone with details of collection locations on BCI and in Gamboa/Pipeline Road (modified from http://biogeodb.stri.si.edu/bioinformatics/maps).

24

Figure 2.2: Bivariate Pearson correlation between the head width and reproductive variables of M. ecuadoria and M. genalis and M. ecuadoria.

25

Figure 2.3: Frequency distribution of the proportion of shared markers between reproductive individuals and their nest-mates (brood and non-reproductive adults), and among unrelated individuals in each population of M. ecuadoria and M. genalis.

26

27

Chapter 3: Notes on the reproductive morphology of the parasitic bee Megalopta byroni (Hymenoptera: Halictidae), and a tentative new host

record

Brood parasitism has evolved repeatedly in sweat bees (Halictidae) (Michener

2000), and in other nest-making Hymenoptera (bees, wasps, and ants) (Choudhary et al.

1994; Savolainen and Vepsalainen 2003; Wcislo 1987). However, many parasitic species

are relatively rare and therefore little is known of their biology. Megalopta

(Augochlorini) is a Neotropical genus of sweat bees, most of which forage under

extremely dim light conditions and nest in dead wood (reviewed in Wcislo et al. 2004). It

also contains several parasitic species (see Michener 2000). Megalopta noctifurax Engel,

Brooks & Yanega and M. fununculosa Hinojosa-Díaz & Engel are putative obligate

parasites based on morphological features that are associated with parasitic behavior, but

there are no rearing records (Engel et al. 1997; Hinojosa-Díaz and Engel 2003). A third

parasitic species, M. byroni Engel, Brooks and Yanega, was reared from nests of M.

genalis Meade-Waldo, and is very rare: only three individuals have been found from

>300 nests between 1998 and 2001 (Wcislo et al. 2004). Here we provide notes on

reproductive structures that are relevant for understanding the evolution of parasitism,

and provide a tentative new host record for M. byroni in Panamá, which was collected

from a nest of M. ecuadoria Friese. Engel (2006) selected the synonymous name M.

centralis Friese as the name for this species, but a petition to conserve the commonly

28

used name M. ecuadoria will soon be before the International Commission on Zoological

Nomenclature for a decision; until a decision is rendered we use the name M. ecuadoria.

We collected 75 nests of M. genalis and M. ecuadoria Friese from June-August

2005, and 114 nests from March-July 2006, along Pipeline Road in Soberanía National

Park (Colón Province, Panamá, N 09.14778 W 079.72997), in nearby secondary forests

near Gamboa, and in the Barro Colorado Nature Monument. Nests were collected in the

field, and opened in the laboratory to collect resident adults and brood.

The metasoma of each parasite was preserved in formalin and later dissected

under light microscopy. The number of ovarioles per ovary was counted, as well as the

number of yellow bodies. The width of the largest ovariole and the width of the Dufour’s

gland were measured using an ocular micrometer. Means are given with their standard

deviations.

In 2005 only 2.7% of 75 nests contained an adult parasite, and in 2006 none of the

114 nests were parasitized (overall parasitism rate, ~1%). The first parasite was found in

late June (beginning of the wet season) in a nest with two M. ecuadoria females and three

open and empty brood cells. Thus, the new host record is tentative until it is confirmed

with rearing records. The second parasite was found in a nest of M. genalis in early

August. The nest was inhabited by a single M. genalis female and there were no cells in

the nest, so apparently it was in an early stage of construction. Almost nothing is known

on the biology of M. byroni, and we do not know if it is a cleptoparasite or a social

parasite. Given the social flexibility of the host bees (i.e., some bees are solitary and

some social) (Wcislo et al. 2004), this classification might not fit at all, and host-parasite

29

relationships might depend also on the social status of particular nests. These two

potential host species, M. genalis and M. ecuadoria, are sympatric in Panama, which is

the only locality record for M. byroni.

Both parasites had three ovarioles per ovary, which is the plesiomorphic number

for Halictidae, including parasitic forms (Rozen 2003), and the width of the largest

ovariole was 0.45 mm for both parasites. By comparison, the resident M. ecuadoria had a

mean ovariole width of 0.75 mm, slightly larger than the M. ecuadoria population mean

of 0.37 mm ± 0.16 (N = 54) and the M. genalis population mean of 0.41mm ± 0.18

(N=59). Furthermore, the first parasite, collected in June, had 12 yellow bodies, while the

second one, collected in August, had only one. In contrast, the nestmates had no yellow

bodies, and the M. ecuadoria population mean was 0.32 ± 0.96 (N=54) yellow bodies,

and 1.16 ± 2.65 (N=59) for M. genalis.

The Dufour’s gland, which produces secretions used in the internal lining of the

cell walls (Cane 1983), was 0.15 and 0.20 mm, respectively for each parasite, which is

much smaller than that of the resident host bees of M. ecuadoria (width, 0.37 mm ± 0.11,

N=2), implying that parasites do not use it in nest architecture or in antisocial

communication. The M. ecuadoria population mean (N=54) for Dufour’s gland width

was 0.41 mm ± 0.19, and 0.44 mm ± 0.19 for M. genalis (N=59).

In most bee taxa, ovariole number per ovary is constant among individuals and

within taxa, except for highly social bees (e.g., Apis) and some parasitic lineages that

exhibit an increased number of ovarioles with respect to the plesiomorphic number in

their free-living relatives (e.g., Alexander and Rozen 1987; Rozen and Ozbek 2003;

30

Serrão J. E. and Martins 2004). A larger number of ovarioles in parasitic species, or an

increased number of oocytes per ovariole, might increase the probability of laying eggs

both during a single nest usurpation event, as well as over the entire life of the parasite

(Ohl and D. 2003; Rozen 2003). However, other factors such as the number of mature

oocytes, the size and thickness of the chorion, may also influence the time taken to

produce eggs. Yellow bodies are remnants of re-absorbed oocytes (Bell and Bohm 1975).

The notable difference in the number of yellow bodies between the first and the second

parasite (12 vs. 1 yellow body, respectively) might reflect the number of failures when

trying to deposit an otherwise ready egg, given that the first parasite was found earlier in

the season and in a well established nest.

The apparent rarity of this parasite might be explained by three aspects of the

biology of hosts and parasite. First, the relative low degree of seasonality in tropical rain

forests might diminish the synchronization of life cycles between hosts and parasites

(reviewed in Wcislo 1987), though central Panama is characterized by strong seasonal

rhythms (Leigh 1999). Second, the lower probability of locating a suitable nest in a three

dimensional space (nests are hung in the vegetation) than in a two dimensional space

(ground nests) may make host searching more difficult (Wcislo 1996). Finally, the fact

that hosts fly at remarkably dim light levels and face severe challenges with respect to

visual ecology (Kelber et al. 2006; Warrant et al. 2004), implies that the parasites also

search for nests in the dark, which should be relatively more difficult. Alternatively, if

parasites try to enter nests during the day, then hosts would always be present as

functional guards.

31

ACKNOWLEDGMENTS

NBB is indebted for the training in finding and opening nests to Juan Carlos di Trani and

Fernando Soley, who participated in the research with WTW through STRI’s Behind-the-Scenes

Volunteer Program and STRI’s internship program, respectively. We thank STRI support staff

and financial support from STRI general research funds to WTW. Funds from the Frank and Fern

Blair Fellowship, University of Texas at Austin to NBB are gratefully acknowledged. We are

indebted to Ulrich Mueller, Michael Engel and an anonymous reviewer for their helpful

comments. The Autoridad Nacional del Ambiente de la República de Panamá provided collecting

and export permits.

32

Chapter 4: Cleaner mites: sanitary mutualism in the miniature ecosystem of Neotropical bee nests

ABSTRACT

Cleaning symbioses represent classic models of mutualism, and some bee-mites

are thought to perform cleaning services for their hosts in exchange for suitable

environments for reproduction and dispersal. These mutual benefits, however, have not

been rigorously demonstrated. We tested the sanitary role of bee-mites by correlating

mite loads with fungal contamination in natural nests of Megalopta genalis and

Megalopta ecuadoria, and by experimentally manipulating mite loads in artificial cells

with developing brood. Field observations revealed significant correlations between the

presence of mites and the absence of fungi inside the brood cells, as well as between the

absence of mites and increased bee mortality. Likewise, experimental brood cells with

mites have fewer fungal colonies than cells without mites. Field observations and

experimental manipulations therefore provide clear evidence of the sanitary effect of

mites in nests of Megalopta bees. This bee-mite association constitutes one of the few

examples of terrestrial cleaning mutualisms.

33

INTRODUCTION

Cleaning symbioses represent classic models of mutualism (Becker & Grutter

2004; Östlund-Nilsson et al. 2005) in which individuals of one species obtain food by

removing external parasites from those of another species. However, most examples are

aquatic and restricted to fishes and shrimps (Becker & Grutter 2005). A commonly cited

example of a terrestrial cleaning symbiosis involves birds that glean ectoparasites from

large mammals, but empirical tests have shown that this relationship is more complicated

and may not always be mutually beneficial (Weeks 2000). Mites associated with some

stingless bees (Trigona) are thought to decrease larval mortality due to fungi, but this has

not been rigorously demonstrated (Flechtmann & Camargo 1974). Parasitic mites

associated with the wasp Allodynerus delphinalis act as ‘bodyguards’ and defend juvenile

wasps from attacks by a hymenopteran parasitoid (Okabe & Makino 2008). Another

example of mutualism includes mites associated with burying beetles. These mites reduce

the number of fly larvae that compete with the beetles for food, and a lower number of

flies on the carcasses increases beetle fitness (Wilson 1983; Wilson & Knollenberg

1987). This symbiosis, however, fluctuates between commensalism and mutualism over

time. Here, we experimentally demonstrate that an association between bees and mites is

a cleaning mutualism, adding to only a handful of documented examples of this kind of

association.

Nests of Hymenoptera (ants, bees, wasps) are prone to attack by microorganisms

(Schmid-Hempel 1998). As a result, many species have evolved behavioral and chemical

34

traits to protect their brood and food. Examples include Philanthus wasps wrapping food

with postpharyngeal gland secretions that reduce the amount of mold growing on the

paralyzed prey (Herzner & Strohm 2007); wood ants furnishing nests with plant resins,

rich in secondary compounds (Chapuisat et al. 2007); and fungus-growing ants producing

metapleural gland secretions that decrease the amount of pathogenic fungi in the fungus

garden (Fernández-Marín et al. 2006). Some species have evolved symbiotic

relationships with other organisms that function to protect themselves from parasitic

attack, such as the antibiotic-producing bacteria associated with the beewolf Philanthus

triangulum (Kaltenpoth et al. 2005) or with fungus-growing ants (Currie et al. 1999; Mueller

et al. 2005).

Both solitary and social bees construct brood cells within their nests (Michener

1974; Wcislo 2000), which creates distinctive microenvironments (sensu Odling-Smee et

al. 2003). Each cell is a miniature ecosystem, and the stored food sustains not only the

development of the bee brood, but also an extensive array of commensal and parasitic

organisms, such as other arthropods, fungi, bacteria and nematodes. One of the more

common arthropod symbionts that live inside nests of bees, ants and wasps are mites

(Arachnida: Acari) (Eickwort 1990, 1993).

Bee-mite associations encompass a broad spectrum of interspecific interactions.

Some mites are parasites, some are predators on other arthropods, including other mites,

and others are hypothesized to be mutualists. One kind of mutualism may involve

cleaning services of mites by suppressing or curbing fungal growth within nests; hosts, in

exchange, provide a suitable environment for mite reproduction, as well as an effective

35

dispersal service (mites ride on the dorsal part of the bee’s metasoma or mesosoma)

(Ordway 1964; Eickwort 1979). The associated sanitary benefits, however, have not been

rigorously demonstrated until now (Flechtmann & Camargo 1974; see Discussion). This

mutualism hypothesis makes two critical predictions. First, if cells contain mites, the

incidence of contaminant fungal growth should be lower compared to cells without mites.

Second, bee brood should be healthier when mites are present in a cell, and brood

survivorship therefore should be higher if raised in the presence of mites, compared to

mite-free brood. We tested these two predictions by correlating natural mite loads with

fungal contamination of cells in natural nests of Megalopta genalis and M. ecuadoria,

and by experimentally manipulating mite load in artificial cells with developing brood.

Overview of bee-mite natural history

Megalopta Smith 1853, is a Neotropical genus of halictid bees (Moure & Hurd

1987), which exhibit a continuous spectrum from solitary to social life cycles, and forage

under dim light (Wcislo et al. 2004). Nests are usually constructed in dead wood and

consist of a single excavated tunnel with individual cells made of chewed wood (Fig. 4.1

A&B) (see Sakagami & Michener 1962; Wcislo et al. 2004). Cells are provisioned with

a mass formed from nectar and pollen. Following oviposition, the cell is sealed, and

brood development from egg to adult eclosion takes about 35 days. Fungi are common in

bee nests given the nutritious substrate available (Batra 1965; Batra et al. 1973; Inglis et

al. 1993; Hajek & Stleger 1994).

36

In Megalopta, a tight synchrony between bee and mite ontogeny exists: each

developmental stage (stadium) of the host bee is often associated with a specific mite

stage (Biani and Wcislo, unpublished data). When callow bees emerge, 91% of the mites

are deutonymphs, the dispersal stage of certain mites (NBB pers. obs.; see Fig. 4.2). A

preliminary identification of the tritonymphs (i.e., the penultimate mite stage before

adulthood) collected from Megalopta nests places them in the genus Laelaspoides

(Mesostigmata; H. Komplen, pers. com.), which are known to be obligate bee-mites

associated with other closely related halictid genera, Augochlorella and Caenaugochlora

(Eickwort 1966, 1979).

MATERIALS & METHODS

Nests collection and field surveys

Nests were collected in Soberanía National Park, Barro Colorado Natural

Monument, San Lorenzo Protected Area, and Bocas del Toro Province (Republic of

Panamá), during June through August of 2005, December 2005 to January 2006, March

through July 2006 and May to June of 2007. Throughout these field seasons, a total of

294 nests were collected, opened in the lab and scanned under a light microscope for the

presence of mites and fungi (39 nests were not included in this study because they were

either abandoned or without brood, leaving 255 nests for analysis).

37

Experimental brood cell procedures

Nests were opened in the lab and the contents of natural cells were transferred

into artificial brood cells made of paraffin, cover slips, and modeling clay (Fig. 4.3). In

order to empirically test the hypothesis that the presence of mites reduces fungal

infection, we experimentally manipulated the presence/absence of mites in artificial

brood cells. Mites were removed or added with a fine paintbrush and all cells were

opened and exposed for the same amount of time, even in control cells (sham treatment).

The paintbrush was disinfected before each manipulation. Between seven and ten

tritonymphs were transferred to the artificial cells. No cells were experimentally

inoculated with fungi, and we relied on natural rates of brood cell contamination.

For the experimental manipulations, we collected 27 nests that had mites and 23

nests that did not have mites. We then divided the group of nests with mites into two

experimental groups: one was used as control (N = 10, sham-opened but otherwise not

manipulated), and the other one had mites removed (N = 17). Similarly, the group of

nests that did not have mites was divided in two experimental groups: one was used as

control (N = 12, sham-opened but otherwise not manipulated), and one had mites added

(N = 11). One independent variable was the presence/absence of mites at the time of

collecting. The second, critical, independent variable was the experimentally controlled

presence/absence of mites at the end of the experiment.

38

Assessing levels of fungi contamination with colony-forming units procedure

Bees were reared in artificial brood cells and after 10 days (approximately a third

of the bee life cycle), fungal growth was quantified by plating diluted cell content and

counting the number of colony-forming units (CFU) (Smith 1980; Christe et al. 2003).

This method consists of diluting and homogenizing a sample (in this case 0.01g of pollen

mass or bee feces) in 10 ml of a solution with 0.85% sodium chloride, 0.2% peptone and

0.05% Tween 80 (a surfactant that facilitates the suspension of fungi spores in water,

Sigma, St. Louis, MO). Successive dilutions were made and then an aliquot of 100µl was

plated on a PDA (Potato Dextrose Agar) medium with 0.2g/L of chloramphenicol as

antibiotic. The optimal dilution is the one that allows the growth of 30 to 300 colonies per

plate; we used the dilution 1x10-4 for all the samples. All the cultures were done under a

sterile laminar-flow hood. Fungi were allowed to grow in a humidity and temperature-

controlled room and then the number of CFU was counted blindly with respect to

treatment. Two replicates were made for each dilution per sample and the CFU counts

were averaged. The CFU of cells that received the same treatment and that belonged to

the same nest were averaged in order to control for within-nest correlations.

Statistical analysis

All data analyses were performed using SPSS 15.0 for Windows (SPSS 2006).

Means are given with their standard errors. For the field survey, a Chi-square test (χ2)

was used. The effect of mites in the experimental brood cells was tested with a 2-way

39

ANOVA. The data met all the respective assumptions for these two tests. Voucher

specimens of the mites and bees are deposited in the Museo Fairchild, Universidad de

Panamá (MFUP), and the Dry Reference Collection of the Smithsonian Tropical

Research Institute (STRI).

RESULTS Field Observations

Nests of M. genalis and M. ecuadoria were often colonized by mites carried by

bees (M. genalis 63%, N=88; M. ecuadoria 17%, N=79). The number of mites per cell

depended on the instar of both bee brood and mites, but it ranged from 0 to 25, with an

average of 8.3 ± 5.8, N=35 tritonymphs per cell. Mites were more common in nests of M.

genalis than in nests of M. ecuadoria (Fig. 4.4 A). During four field surveys in 2005,

2006, and 2007, spanning both wet and dry seasons, we observed a significant correlation

between the presence of mites and the absence of fungi growing inside the brood cells of

both M. genalis and M. ecuadoria’s nests (Fig. 4.4 A, B &C). This provides support for

the first prediction that the incidence of contaminant fungal growth is lower in cells with

mites than in those without them.

There was no significant association between presence of mites and sex of brood

(χ2=0.838, p=0.6577, N=25; χ2=1.311, p=0.2522, N=13; χ2=0.009, p=0.9229, N=97;

χ2=1.543, p=0.2142, N=54, respectively, for the seasons mentioned in Fig. 4.4 A). Both

male and female brood therefore become equally infested by mites.

40

Although most brood are successfully reared in the absence of mites, there was a

significant association between healthy bees (i.e. bees that reach adulthood and are not

infected by fungi) and the presence of mites (χ2=5.665, p=0.0173, N=283). A total of 13

bees died out of 178 bees growing in the absence of mites, while only one pupa died out

of 105 bees in association with mites (the nest with the dead pupa had been abandoned by

the resident adults and the brood cell was broken, so it is likely that other factors were

responsible for the death of this pupa). This finding implies that the mites serve some

sanitary function, supporting the second prediction of healthier bees when mites are

present (Fig. 4.1 C&D).

Experimental brood cell test

The effect of mites on the incidence of fungal growth in cells was tested by

removing or adding mites to cells that naturally did or did not have mites, respectively.

The control cells that naturally had mites throughout the experiment yielded an average

of 130.6 ± 26.5, N=10 fungal colonies (CFU). Cells that naturally did not have mites but

had them added yielded an average of 190.6 ± 26.4 CFU, N=11. In contrast, cells that

never had mites had an average of 224.4 ± 25.6 CFU, N=12, and cells that naturally had

mites but had them removed had an average of 230.1 ± 16.3 CFU, N=17 (Fig. 4.5).

A two-way ANOVA assessed the effect of the presence of mites before and after

experimental manipulation on the number of fungal colonies. Cells that always had mites

or had mites added later, have a significantly lower number of CFU than cells that never

41

had mites or had mites removed (F1,46=8.227, p=0.006). There was no significant effect

on the CFU counts due to the handling of the cells (F1,46=1.366, p=0.249), and no

interaction between the presence/absence of mites at the beginning of the experiment and

at the end (F1,46=1.994, p=0.165).

The effect size (η2p) is 0.152, which means that most of the variation can be

explained by the treatments, and that the size of the difference between means is

moderate to large. The power is about 80% and so it is reasonable to conclude that there

are biologically real differences between the means.

DISCUSSION

Our study provides the first rigorous demonstration of the repeatedly

hypothesized cleaning mutualism between mites and bees. A single natural history

account suggested that stingless bees benefit from the presence of mites inhabiting their

nests because the latter clean the nest of fungal contaminants, but this account presented

neither methodological details nor statistical analyses (Flechtmann & Camargo 1974). In

contrast, our tests provide clear evidence of the cleaning effect of mites in nests of

Megalopta bees. Laboratory results demonstrate that there are fewer fungal colonies

(CFU) in cells with mites than in cells without mites. This outcome mirrors the

observational data from natural nests of both M. genalis and M. ecuadoria. Experimental

manipulation of mite load further confirms that the absence or presence of mites is a

causal factor in the growth of fungi (or not) in brood cells. It is unclear why M. genalis

has generally higher mite loads compared to M. ecuadoria. This difference might be due

42

to the incipience of the mutualism or perhaps M. genalis spreads mites more effectively

between nests. Future studies should address the ecological and species-specific reasons

for the species differences in mite load between M. genalis and M. ecuadoria.

Symbiotic relationships range through a wide continuum from parasitic to

mutualistic, as well as from diffuse to highly coevolved associations between pairs of

species (Herre et al. 1999). Of these relationships, mutualisms are widespread, and some

are thought to be one of the driving forces of diversification and stability in complex

ecological systems (Boucher et al. 1982; Leigh 1991; Bronstein et al. 2006). The fact that

not all the nests were colonized by mites, and that brood can be reared successfully even

in the absence of mites, may indicate that this bee-mite relationship is a facultative

mutualism, at least from the bee perspective. However, as far as it is known, these mites

are obligate bee-associates (Eickwort 1966), and it remains unclear if they can survive in

favorable habitats outside of bee nests, such as flowers.

Host sex preference has evolved in some mites associated with hymenopterans

(Cowan 1984; Hunter & Rosario 1988; Dawicke et al. 1992). For instance, some parasitic

mites are found preferentially on female hosts, presumably to disperse to new nests,

while others are found in males, and are transferred to females during copulation

(OConnor 1982; Cowan 1984). Other bee mites do not exhibit such sex-based

associations (Cross & Bohart 1969; OConnor 1982; Houston 1987; Cross & Bohart 1992;

Okabe & Makino 2003). Our results show that cells with male brood are as equally likely

to have mites as are those cells that produce females. This prompts the question of the

mode of transmission of the mites associated with Megalopta. Based on the ontogeny of

43

the mites relative to that of the bees (Biani et al., unpublished data), we assume that bees

are infested by mites in their natal nests during development. Nothing is known about

how mites are transferred from males to females. Since only females potentially found

new nests, mites developing in males brood cells would constitute an evolutionary dead

end unless there is a mechanism that allows their transference to females. The mating

behavior of Megalopta is unknown, but venereal transmission during mating is possible,

as occurs in mites associated with an eumenine wasp (Cooper 1955; Cowan 1984).

Alternatively, both male and female Megalopta remain in the natal nest for up to 10 days

after they eclose (Wcislo & Gonzalez 2006), a period that could constitute an opportunity

for mites to transfer from males to females. Furthermore, M. genalis and M. ecuadoria

are socially flexible species, meaning that some newly emerged females might disperse to

found new nests, while others will remain in their natal nest to become part of a social

unit. Hence, it is possible that mites attached to dispersing females will be developing a

new generation in a brand new nest, whereas mites attached to daughter bees remaining

in their natal nest might colonize new brood cells in the same nest.

Mites are present in the nests of bees from diverse families, including Halictidae,

Megachilidae, and Apidae; many taxa possess pouch-like structures (acarinaria) that

localize mites to certain body regions, and museum specimens often house numerous

mites (Batra 1965; Eickwort 1966; Cross & Bohart 1969; Eickwort 1979; Mc Ginley

1986; Houston 1987; Eickwort 1990; Cross & Bohart 1992; Eickwort 1993; Okabe &

Makino 2002, 2005). Yet ecological and behavioral studies of bee-mite symbioses are

scarce (Woodring 1973; Rack & Eickwort 1979; Houston 1987; Cross & Bohart 1992;

44

Walter et al. 2002; Okabe & Makino 2008). Acarinaria were interpreted as structures that

evolved to transport beneficial mites. However, due to the dearth of examples of such

beneficial interactions, this notion has recently been challenged by an alternative

hypothesis that the acarinarium is a defense mechanism against harmful mites because it

decreases the likelihood that mites will leave any given host and thus infest other cells in

the same nests (Klimov et al. 2007). The evolution of these morphological structures

needs to be examined in more depth, for example by exploring the percentage of species

where acarinaria are exclusively a female attribute (Okabe & Makino 2002). If bees are

carrying harmful mites in the acarinaria, there would be strong sexual selection against

males with these structures.

Coevolutionary interactions between wild bees and mites have been largely

neglected, except for a few notable examples (Rack & Eickwort 1979; Houston 1987;

Cross & Bohart 1992). Future studies need to investigate experimentally the nature of the

relationships based on fitness advantages for each of the players; identify the mode of

transmission and behavioral repertoires leading towards the association; and examine

phylogenetic convergences or ecological constraints that may be affecting the system.

The mechanisms that curb fungal growth in brood cells need to be further investigated, as

our study did not address whether the mites are eating fungal spores or hyphae, or if they

might be secreting biochemical substances with fungistatic activity. This study is the first

experimental demonstration of the mutualistic nature of a bee-mite association, and one

of the few examples of terrestrial cleaning mutualisms.

45

ACKNOWLEDGMENTS

This work was supported by the Smithsonian Tropical Research Institute (STRI)

Short Term Fellowships as well as research grants from the Program of Ecology,

Evolution and Behavior of the University of Texas at Austin to NBB. Additional support

came from general research funds from STRI to WTW; support for an intern came from

STRI’s Office of Education. We are extremely grateful to Juan Carlos Santos for his

statistical advice and help with figures; Andre Rodrígues for his suggestions on

microbiology methods; Adam Smith, Simon Tierney and Therany Gonzales for their

invaluable help and company during nest collections; Hermógenes Fernández-Marín for

logistic help in the lab; Hans Klompen and Jon Shik for their help in the identification of

the mite species; Barrett Klein, Christian Rabeling, Rachelle Adams, Alexander

Mikheyev and Heather Ishak for their comments on earlier versions of the manuscript.

Dr. Gard Otis and an anonymous reviewer offered very helpful suggestions. We thank

STRI staff for logistical support. The Autoridad Nacional del Ambiente de la República

de Panamá provided collecting and export permits. And of course, we thank the bees and

the mites.

46

Figure 4.1: Megalopta nest structure and cleaning effects of the mites.

A. Section through nest in dead branch. Brood cells are constructed out of chewed wood. B.

Nest entrance. C. Brood cell completely overgrown with fungi. D. A dead pre-pupa with

fungi growing on the bee feces.

47

Figure 4.2: Deutonymphs (dipersal stage of the mites) on thorax and metasoma of Megalopta pupa.

Figure 4.3: Artificial cell with a bee larva.

Artificial cells are made of paraffin, a cover slip, and modeling clay. Natural cells can be

transferred into these artificial cells to allow for observation and rearing of bee brood.

48

Figure 4.4 A. Frequency distribution of brood cells with or without mites (M) and/or

fungi (F) in M. genalis and M. ecuadoria for all four field seasons.

(X2* p<0.05, ** p<0.001). In M. genalis mites are more abundant and fungi are scarce, while in M. ecuadoria, mites are less abundant and fungi are more pervasive. The negative relationship between these organisms holds for within species associations. B. & C. Brood cells with and without mites, respectively. Both cells are of approximately the same age. The pellets in the cells are feces deposited by the bee larvae. Cells with mites tend to be fungus-free (B), whereas cells without mites show significant fungal growth (C) that may harm the developing bee brood. Cell diameters are on average 8 mm.

49

Figure 4.5. Means ± 2 Standard Errors of Colony Forming Units in the four experimental treatments.

“Control with Mites” are cells that naturally had mites. Cells that did not have mites received them experimentally. “Control without Mites” are cells that naturally did not have mites throughout the experiment, and other cells had mites experimentally removed. (Two-way ANOVA, F1,46=8.227, p=0.006).

50

References

Alexander, B., and J. G. Rozen. 1987. Ovaries, ovarioles and oocytes in parasitic bees

(Hymenoptera: Apoidea). Pan-Pacific Entomologist 63:155-164. Arneson, L., and W. T. Wcislo. 2003. Dominant-subordinate relationships in a

facultatively social, nocturnal bee, Megalopta genalis (Hymenoptera : Halictidae). Journal of the Kansas Entomological Society 76:183-193.

Bargum, K., H. Helantera, and L. Sundstrom. 2007. Genetic population structure, queen

supersedure and social polymorphism in a social Hymenoptera. Journal of Evolutionary Biology 20:1351-1360.

Batra, L. R., S. W. T. Batra, and G. E. Bohart. 1973. The mycoflora of domesticated and

wild bees (Apoidea). Mycopathologia 49:13-44. Batra, S. W. T. 1965. Organisms associated with Lasioglossum zephyrum

(Hymenoptera: Halictidae). J. Kans.Entomol. Soc. 38:367-389. Becker, J. H. A., and A. S. Grutter. 2004. Cleaner shrimp do clean. Coral Reefs 23:515-

520. Becker, J. H. A., and A. S. Grutter. 2005. Client fish ectoparasite loads and cleaner

shrimp Urocaridella sp. c hunger levels affect cleaning behaviour. Anim. Behav. 70:991-996.

Bell, W. J., and M. K. Bohm. 1975. Oosorption in insects. Biological Reviews of the

Cambridge Philosophical Society 50:373-396. Bernal, X. E., R. A. Page, A. S. Rand, and M. J. Ryan. 2007. Cues for eavesdroppers:

Do frog calls indicate prey density and quality? Am Nat 169:409-415. Biani, Natalia B., Ulrich G. Mueller, and William T. Wcislo. 2009. Cleaner Mites:

Sanitary Mutualism in the Miniature Ecosystem of Neotropical Bee Nests. The American Naturalist 173:841-847.

Biani, N. B., and W. T. Wcislo. 2007. Notes on the reproductive morphology of the

parasitic bee Megalopta byroni (Hymenoptera : Halictidae), and a tentative new host record. Journal of the Kansas Entomological Society 80:392-394.

51

Bonin, A., D. Ehrich, and S. Manel. 2007. Statistical analysis of amplified fragment

length polymorphism data: a toolbox for molecular ecologists and evolutionists. Molecular Ecology 16:3737-3758.

Boucher, D. H., S. James, and K. H. Keeler. 1982. The Ecology of Mutualism. Annu.

Rev. Ecol. Syst. 13:315-347. Bourke, A. F. G. 2005. Genetics, relatedness and social behaviour in insect societies.

Pp. 1-30 in M. Fellowes, et al., ed. Insect Evolutionary Ecology. CABI Publishing.

Bronstein, J. L. 2001. The exploitation of mutualisms. Ecology Letters 4:277-287. Bronstein, J. L., R. Alarcón, and M. Geber. 2006. The evolution of plant-insect

mutualisms. New Phytol. 172:412-428. Cane, J. H. 1983. Chemical evolution and chemosystematics of the Dufour's gland

secretions of lactone-producing bees (Hymenoptera: Colletidae, Halictidae, and Oxaeidae). Evolution 37:657-674.

Chapuisat, M., A. Oppliger, P. Magliano, and P. Christe. 2007. Wood ants use resin to

protect themselves against pathogens. Proc. R. Soc. Lond. B 274:2013-2017. Choudhary, M., J. E. Strassmann, D. C. Queller, S. Turillazzi, and R. Cervo. 1994.

Social parasites in polistine wasps are monophyletic implications for sympatric speciation. Proceedings of the Royal Society B: Biological Sciences 257:31-35.

Christe, P., A. Oppliger, F. Bancala, G. Castella, and M. Chapuisat. 2003. Evidence for

collective medication in ants. Ecol. Lett. 6:19-22. Cooper, K. W. 1955. Venereal transmission of mites by wasps, and some evolutionary

problems arising from the remarkable association of Ensliniella trisetosa with the wasp Ancistrocerus antilope. Biology of eumenine wasps II. Trans. Am. Entomol. Soc. 80:119-174.

Cowan, D. P. 1984. Life history and male dimorphism in the mite Kennethiella trisetosa

(Acarina: Winterschmidtiidae), and its symbiotic relationship with the wasp Ancistrocerus antilope (Hymenoptera: Eumenidae). Ann. Entomol. Soc. Am. 77:725-732.

Cross, E. A., and G. E. Bohart. 1969. Phoretic behavior of four species of alkali bee

mites as influenced by season and host sex. J. Kans. Entomol. Soc. 42:195-219.

52

Cross, E. A., and G. E. Bohart. 1992. The biology of Imparipes apicola (Acari:

Scutacaridae) and its relationships to the alkali bee, Nomia melanderi (Hymenoptera: Halictidae), and to certain fungi in the bee cell ecosystem. J. Kans. Entomol. Soc. 65:157-173.

Currie, C. R., U. G. Mueller, and D. Malloch. 1999. The agricultural pathology of ant

fungus gardens. Proc. Natl. Acad. Sci. U.S.A. 96:7998-8002. Danforth, B. N. 2002. Evolution of sociality in a primitively eusocial lineage of bees.

Proceedings of the National Academy of Sciences of the United States of America 99:286-290.

Dawicke, B. L., G. W. Otis, C. Scott-Dupree, and M. Nasr. 1992. Host preference of the

honey bee tracheal mite (Acarapis woodi Rennie). Exp. Appl. Acarol. 15:83-98. Eickwort, G. C. 1966. A new genus and species of mite associate with the green bee

Augochlorella (Hymenoptera:Halictidae) in Kansas (Acarina: Laelaptidae: Hypoaspidinae) J. Kans. Entomol. Soc. 39:410-429.

Eickwort, G. C. 1979. Mites associated with sweat bees (Halictidae) in J. G. Rodriguez,

ed. Recent advances in Acarology. Academic Press, New York: 575-581. Eickwort, G. C. 1990. Associations of mites with social insects. Annu. Rev. Entomol.

35:469-488. Eickwort, G. C. 1993. Evolution and life-history patterns of mites associated with bees.

Pp. 218-251 in M. A. Houck, ed. Mites: Ecological and Evolutionary Analyses of Life-History Patterns. Chapman and Hall, New York.

Eickwort, G. C., K. R. Eickwort, J. M. Eickwort, J. Gordon, and M. A. Eickwort. 1996.

Solitary behaviour in a high altitude population of the social sweat bee Halictus rubicundus (Hymenoptera: Halictidae). Behavioral Ecology and Sociobiology 38:227–233.

Engel, M. S. 2000. Classification of the bee tribe Augochlorini (Hymenoptera :

Halictidae). Bulletin of the American Museum of Natural History:4-89. Engel, M. S. 2006. A nocturnal bee of the genus Megalopta, with notes on other Central

American species. Mitteilungen des Internationalen Entomologischen Vereins 31:37-49.

53

Engel, M. S., R. Brooks, W., and D. Yanega. 1997. New genera and subgenera of augochlorine bees (Hymenoptera: Halictidae). Bulletin of the Natural History Museum The University of Kansas 5:1-21.

Evans, H. E. 1977. Commentary: Extrinsic versus Intrinsic Factors in the Evolution of

Insect Sociality. BioScience 27:613-617. Excoffier, L., P. E. Smouse, and J. M. Quattro. 1992. Analysis of Molecular Variance

Inferred from Metric Distances among DNA Haplotypes - Application to Human Mitochondrial-DNA Restriction Data. Genetics 131:479-491.

Fægri, K., and L. van der Pijl. 1979. The Principles of Pollination Ecology. Pergamon

Press, New York. Faulkes, C. G., and N. C. Bennett. 2001. Family values: group dynamics and social

control of reproduction in African mole-rats. Trends Ecol Evol 16:184–190. Fernández-Marín, H., J. K. Zimmerman, S. A. Rehner, and W. T. Wcislo. 2006. Active

use of the metapleural glands by ants in controlling fungal infection. Proc. R. Soc. Lond. B 273:1689-1695.

Flechtmann, C. H. W., and C. A. Camargo. 1974. Acari associated with stingless bees

(Meliponidae, Hymenoptera) from Brazil. International Congress of Acarology, Akadémiai Kiadó, Budapest. 4:315-319.

Foster, K. R., T. Wenseleers, F. L. W. Ratnieks, and D. C. Queller. 2006. There is

nothing wrong with inclusive fitness. Trends in Ecology & Evolution 21:599-600. Francescato, E., A. Massolo, M. Landi, L. Gerace, R. Hashim, and S. Turillazzi. 2002.

Colony membership, division of labor, and genetic relatedness among females of colonies of Eustenogaster fraterna (Hymenoptera, Vespidae, Stenogastrinae). Journal of Insect Behavior 15:153-170.

Gadagkar, R. 1990. Origin and Evolution of Eusociality - A Perspective from Studying

Primitively Eusocial Wasps. Journal of Genetics 69:113-125. Gemeno, C., K. V. Yeargan, and K. F. Haynes. 2000. Aggressive Chemical Mimicry by

the Bolas Spider Mastophora hutchinsoni: Identification and Quantification of a Major Prey's Sex Pheromone Components in the Spider's Volatile Emissions. Journal of Chemical Ecology 26:1235-1243.

Gilbert, O. M., K. R. Foster, N. J. Mehdiabadi, J. E. Strassmann, and D. C. Queller.

2007. High relatedness maintains multicellular cooperation in a social amoeba by

54

controlling cheater mutants. Proceedings of the National Academy of Sciences of the United States of America 104:8913-8917.

Griffin, A. S., and S. A. West. 2003. Kin discrimination and the benefit of helping in

cooperatively breeding vertebrates. Science 302:634–636. Hajek, A. E., and R. J. Stleger. 1994. Interactions between fungal pathogens and insect

hosts. Annu. Rev. Entomol. 39:293-322. Hamilton, W. D. 1964. The genetical evolution of social behaviour. I & II. Journal of

Theoretical Biology 7:1-52. Hauber, M. E., P. J. Yeh, and J. O. L. Roberts. 2004. Patterns and coevolutionary

consequences of repeated brood parasitism. Proceedings of the Royal Society of London Series B-Biological Sciences 271:S317-S320.

Herre, E. A., N. Knowlton, U. G. Mueller, and S. A. Rehner. 1999. The evolution of

mutualisms: exploring the paths between conflict and cooperation. Trends Ecol. Evol. 14:49-53.

Herzner, G., and E. Strohm. 2007. Fighting fungi with physics: Food wrapping by a

solitary wasp prevents water condensation. Curr. Biol. 17:R 46-47. Hinojosa-Díaz, I. A., and M. S. Engel. 2003. Megalopta (Noctoraptor) furunculosa

sp.n., a new nocturnal, cleptoparasitic bee from Guyana (Hymenoptera: Halictidae). Folia Heyrovskiana 11:137-141.

Houston, T. F. 1987. The symbiosis of acarid mites, genus Ctenocolletacarus (Acarina:

Acariformes) and stenotritid bees, genus Ctenocolletes (Insecta: Hymenoptera). Aust. J. Zool. 35:459-468.

Hughes, W. O. H., B. P. Oldroyd, M. Beekman, and F. L. W. Ratnieks. 2008. Ancestral

monogamy shows kin selection is key to the evolution of eusociality. Science 320:1213-1216.

Hunter, P. E., and R. M. T. Rosario. 1988. Associations of Mesostigmata with other

arthropods. Annu. Rev. Entomol. 33:393-417. Inglis, G. D., L. Sigler, and M. S. Goettel. 1993. Aerobic microorganisms associated

with alfalfa leafcutter bees (Megachile rotundata). Microb. Ecol. 26:125-143.

55

Jarvis, J. U. M. 1981. Eusociality in a mammal: cooperative breeding in naked mole-rat colonies. Science 212:571–573.

Jones, C. B. 1997. Social parasitism in the mantled howler monkey, Alouatta palliata

gray (Primates: Cebidae): Involuntary altruism in a mammal? Sociobiology 30:51-61.

Kaltenpoth, M., W. Göttler, G. Herzner, and E. Strohm. 2005. Symbiotic bacteria

protect wasp larvae from fungal infestation. Curr. Biol. 15:475-479. Kelber, A., E. J. Warrant, M. Pfaff, R. Wallen, J. C. Theobald, W. T. Wcislo, and R. A.

Raguso. 2006. Light intensity limits foraging activity in nocturnal and crepuscular bees. Behavioral Ecology 17:63-72.

Klimov, P. B., S. B. Vinson, and B. M. OConnor. 2007. Acarinaria in associations of

apid bees (Hymenoptera) and chaetodactylid mites (Acari). Invertebr. Syst. 21:109-136.

Kukuk, P. F. 1989. Evolutionary genetics of a primitively eusocial halictine bee,

Dialictus zephyrus. Pp. 183-202 in M. D. Breed, Page, R.E., ed. Genetics of social evolution. Westview Press Boulder, CO.

Kukuk, P. F., G. C. Eickwort, and B. May. 1987. Multiple maternity and multiple

paternity in first generation brood from single foundress colonies of the sweat bee Dialictus zephyrus (Hymenoptera, Halictidae). Insectes Sociaux 34:131-135.

Kukuk, P. F., S. A. Ward, and A. Jozwiak. 1998. Mutualistic benefits generate an

unequal distribution of risky activities among unrelated group members. Naturwissenschaften 85:445-449.

Landi, M., D. C. Queller, S. Turillazzi, and J. E. Strassmann. 2003. Low relatedness and

frequent queen turnover in the stenogastrine wasp Eustenogaster fraterna favor the life insurance over the haplodiploid hypothesis for the origin of eusociality. Insectes Sociaux 50:262-267.

Leigh, E. G. 1991. Genes, bees and ecosystems: The evolution of a common interest

among individuals. Trends Ecol. Evol. 6:257-262. Leigh, E. G., Jr. 1999. Tropical Forest Ecology: A View from Barro Colorado Island.

Oxford University Press, New York. Maddison, W. P., and D. R. Maddison. 2009. Mesquite: a modular system for

evolutionary analysis. http://mesquiteproject.org.

56

Maynard Smith, J. 1989. Generating novelty by symbiosis. Nature 341:284-285. Maynard Smith, J., and E. Szathmary. 1997. The Majors Transitions in Evolution

Oxford University Press, Oxford, UK. Mc Ginley, R. J. 1986. Studies of Halictinae (Apoidea: Halictidae), I: Revision of the

New World Lasioglossum. Smithson. Contrib. Zool. 429:1-294. Michener, C. D. 1974. The Social Behavior of the Bees: A Comparative Study. Harvard

University Press, Cambridge, MA. Michener, C. D. 2000. The Bees of the World. Johns Hopkins University Press,

Baltimore. Morin, P. A., R. G. LeDuc, E. Archer, K. K. Martien, B. L. Taylor, R. Huebinger, and J.

W. Bickham. 2007. Estimated genotype error rates from bowhead whale microsatellite data. Scientific Committee of the International Whaling Commission, 1-10.

Moure, J. S., and P. D. Hurd. 1987. An Annotated Catalogue of the Halictidae Bees of

the Western Hemisphere. Smithsonian Institution Press., Washington, D.C. Mueller, U. G. 1996. Life history and social evolution of the primitively eusocial bee

Augochlorella striata (Hymenoptera: Halictidae). Journal of the Kansas Entomological Society 69:116-138.

Mueller, U. G., G. C. Eickwort, and C. F. Aquadro. 1994. DNA-Fingerprinting Analysis

of Parent-Offspring Conflict in a Bee. Proceedings of the National Academy of Sciences of the United States of America 91:5143-5147.

Mueller, U. G., N. Gerardo, T. R. Schultz, D. Aanen, and D. Six. 2005. The evolution of

agriculture in insects. Annu. Rev. Ecol. Evol. Syst. 36:563-569. OConnor, B. M. 1982. Evolutionary ecology of astigmatid mites. Annu. Rev. Entomol.

27:385-409. Odling-Smee, F. J., K. N. Laland, and M. W. Feldman. 2003. Niche Construction.

Princeton University Press., Princeton. Ohl, M., and L. D. 2003. Ovaries, ovarioles, and oocytes in apoid wasps, with special

reference to cleptoparasitic species (Hymenoptera: Apoidea: "Sphecidae"). Journal of the Kansas Entomological Society 76:147-159.

57

Okabe, K., and S. Makino. 2002. Phoretic mite fauna on the large carpenter bee

Xylocopa appendiculata circumvolans (Hymenoptera: Apidae) with descriptions of its acarinaria on both sexes. Nihon Dani Gakkai Shi 11:73-84.

Okabe, K., and S. Makino. 2003. Life history of Kurosaia jiju (Acari:

Winterschmidtiidae) symbiotic with a mason wasp, Anterhynchium flavomarginatum micado (Hymenoptera: Eumenidae). Ann. Entomol. Soc. Am. 96:652-659.

Okabe, K., and S. Makino. 2005. Mite faunas and morphology of acarinaria on japanese

and taiwanese large carpenter bees (Hymenoptera: Apidae). Nihon Dani Gakkai Shi 14:105-115.

Okabe, K., and S. Makino. 2008. Parasitic mites as part-time bodyguards of a host wasp.

Proc. R. Soc. Lond. B 275:2293-2297. Okello, J. B. A., G. Wittemyer, H. B. Rasmussen, I. Douglas-Hamilton, S. Nyakaana, P.

Arctander, and H. R. Siegismund. 2005. Noninvasive Genotyping and Mendelian Analysis of Microsatellites in African Savannah Elephants. J Hered 96:679-687.

Ordway, E. 1964. Sphecodes pimpinellae and other enemies of Augochlorella

(Hymenoptera: Halictidae). J. Kans. Entomol. Soc. 37:139-152. Östlund-Nilsson, S., J. H. A. Becker, and G. E. Nilsson. 2005. Shrimps remove

ectoparasites from fishes in temperate waters. Biol. Lett. 1:454-456. Packer, L. 1990. Solitary and eusocial nests in a population of Augochlorella striata

(Provancher) (Hymenopter, Halictidae) at the northern edge of its range. Behavioral Ecology and Sociobiology 27:339–344.

Page, R. E. 1986. Sperm Utilization in Social Insects. Annual Review of Entomology

31:297-320. Paxton, R. J., M. Ayasse, J. Field, and A. Soro. 2002. Complex sociogenetic

organization and reproductive skew in a primitively eusocial sweat bee, Lasioglossum malachurum, as revealed by microsatellites. Molecular Ecology 11:2405-2416.

Paxton, R. J., L. R. Bego, M. M. Shah, and S. Mateus. 2003. Low mating frequency of

queens in the stingless bee Scaptotrigona postica and worker maternity of males. Behavioral Ecology and Sociobiology 53:174-181.

58

Queller, D. C. 1989. The evolution of eusociality: Reproductive head starts of workers. .

Proc Natl Acad Sci U S A 86:3224-3226. Queller, D. C., and J. E. Strassmann. 1998. Kin selection and social insects. Bioscience

48:165-175. Rack, G., and G. C. Eickwort. 1979. Biology and description of a new pygmephorid

mite (Acarina: Tarsonemida) associated with the soil-nesting bee Agapostemon nasutus (Hymenoptera: Halictidae). Acarologia 21:267-278.

Ratnieks, F. L. W., K. R. Foster, and T. Wenseleers. 2006. Conflict resolution in insect

societies. Annu. Rev. Entomol. 51:581–608. Reeve, H. K., D. F. Westneat, and D. C. Queller. 1992. Estimating average within-group

relatedness from DNA fingerprints. Molecular Ecology 1:223-232. Richards, M. H. 2000. Evidence for geographic variation in colony social organization

in an obligately social sweat bee, Lasioglossum malachurum Kirby (Hymenoptera; Halictidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie 78:1259-1266.

Richards, M. H., D. French, and R. J. Paxton. 2005. It's good to be queen: classically

eusocial colony structure and low worker fitness in an obligately social sweat bee. Molecular Ecology 14:4123-4133.

Richards, M. H., and L. Packer. 1998. Demography and relatedness in multiple-

foundress nests of the social sweat bee, Halictus ligatus. Insectes Sociaux 45:97-109.

Rozen, J. G. 2003. Eggs, ovariole numbers, and modes of parasitism of cleptoparasitic

bees, with emphasis on neotropical species. American Museum Novitates 3413:1-36.

Rozen, J. G., and H. Ozbek. 2003. Oocytes, eggs, and ovarioles of some long-tongued

bees (Hymenoptera : Apoidea). American Museum Novitates:3-35. Sachs, J. L., U. G. Mueller, T. P. Wilcox, and J. J. Bull. 2004. The evolution of

cooperation. Quarterly Review of Biology 79:135-160. Sakagami, S. F., and C. D. Michener. 1962. The nest architecture of the sweat bees

(Halictinae). A comparative study of behavior. The University of Kansas Press.

59

Sakagami, S. F., and J. S. Moure. 1965. Cephalic polymorphism in some neotropical halictine bees. An. Acad. Brasil. Ciencias 37:303–313.

Savolainen, R., and K. Vepsalainen. 2003. Sympatric speciation through intraspecific

social parasitism. Proc Natl Acad Sci U S A 100:7169-7174. Schmid-Hempel, P. 1998. Parasites in Social Insects. Princeton University Press,

Princeton. Selosse, M. A., F. Richard, X. He, and S. W. Simard. 2006. Mycorrhizal networks: des

liaisons dangereuses? Trends Ecol Evol 21:621-628 Serrão J. E., and G. Martins. 2004. A comparative study of the ovaries in some Brazilian

bees (Hymenoptera; Apoidea). Papéis Avulsos de Zoologia Museu de Zoologia da Universidade de São Paulo 44:45-53.

Smith, A., W. Wcislo, and S. O’Donnell. 2008. Body Size Shapes Caste Expression, and

Cleptoparasitism Reduces Body Size in the Facultatively Eusocial Bees Megalopta (Hymenoptera: Halictidae). Journal of Insect Behavior 21:394-406.

Smith, A. L. 1980. Microbiology and Pathology, S. Louis Missouri. Smith, A. R., W. T. Wcislo, and S. O'Donnell. 2003. Assured fitness returns favor

sociality in a mass-provisioning sweat bee, Megalopta genalis (Hymenoptera : Halictidae). Behavioral Ecology and Sociobiology 54:14-21.

Smith, A. R., W. T. Wcislo, and S. O'Donnell. 2007. Survival and productivity benefits

to social nesting in the sweat bee Megalopta genalis (Hymenoptera : Halictidae). Behavioral Ecology and Sociobiology 61:1111-1120.

SoftGenetics, L. L. C. 2006. Gene Marker. Operation Manual. www.softgentics.com,

State College, PA. Soucy, S. L. 2002. Nesting biology and socially polymorphic behaviour of the sweat bee

Halictus rubicundus (Hymenoptera: Halictidae). Ann Entomol Soc Am 95:57–65. Soucy, S. L., and B. N. Danforth. 2002. Phylogeography of the socially polymorphic

sweat bee Halictus rubicundus (Hymenoptera: Halictidae). Evolution 56:330–341.

SPSS. 2006. SPSS 15.0 for Windows Version 15.0, Chicago, IL: SPSS Inc.

60

Starr, C. K. 1984. Sperm competition, kinship, and sociality in the aculeate Hymenoptera. Pp. 427–464 in R. L. Smith, ed. Sperm Competition and the Evolution of Animal Mating Systems Academic Press, Orlando, FL.

Stow, A., L. Silberbauer, A. J. Beattie, and D. A. Briscoe. 2007. Fine-scale genetic

structure and fire-created habitat patchiness in the Australian allodapine bee, Exoneura nigrescens (Hymenoptera : Apidae). Journal of Heredity 98:60-66.

Strassmann, J. E., C. R. Hughes, D. C. Queller, S. Turillazzi, R. Cervo, S. K. Davis, and

K. F. Goodnight. 1989. Genetic Relatedness in Primitively Eusocial Wasps. Nature 342:268-270.

Stuart, R. J. 2002. The Behavioral Ecology of Social Parasitism in Ants in J. F. C. E.E.

Lewis, & M.V.K. Sukhdeo, ed. The Behavioral Ecology of Parasites. CAB International.

Tierney, S. M., T. Gonzales-Ojeda, and W. T. Wcislo. 2008. Biology of a nocturnal bee,

Megalopta atra (Hymenoptera : Halictidae; Augochlorini), from the Panamanian highlands. Journal of Natural History 42:1841-1847.

Vekemans, X., T. Beauwens, M. Lemaire, and I. Roldan-Ruiz. 2002. Data from

amplified fragment length polymorphism (AFLP) markers show indication of size homoplasy and of a relationship between degree of homoplasy and fragment size. Molecular Ecology 11:139-151.

Walter, D. E., J. J. Beard, K. L. Walker, and K. Sparks. 2002. Of mites and bees: A

review of mite-bee associations in Australia and a revision of Raymentia womersley (Acari: Mesostigmata: Laelapidae), with the description of two new species of mites from Lasioglossum (Parasphecodes) spp. (Hymenoptera: Halictidae). Aust. J. Entomol. 41:128-148.

Warrant, E. J., A. Kelber, A. Gislen, B. Greiner, W. Ribi, and W. T. Wcislo. 2004.

Nocturnal vision and landmark orientation in a tropical halictid bee. Current Biology 14:1309-1318.

Wcislo, W. T. 1987. The roles of seasonality, host synchrony, and behaviour in the

evolutions and distributions of nest parasites in Hymenoptera (Insecta), with special reference to bees (Apoidea). Biological Reviews 62:515-543.

Wcislo, W. T. 1996. Parasitism rates in relation to nest site in bees and wasps

(Hymenoptera: Apoidea). Journal of Insect Behavior 9:643-656.

61

Wcislo, W. T. 1997. Behavioral environments of sweat bees (Halictinae) in relation to variability in social organization in B. J. C. J.C.Choe, ed. The Evolution of Social Behavior in Insects and Arachnids. Cambridge University Press, Cambridge, UK.

Wcislo, W. T. 2000. Environmental hierarchy, behavioral contexts, and social evolution

in insects. Oecologia Brasiliensis supplement 8:49-84. Wcislo, W. T., L. Arneson, K. Roesch, V. Gonzalez, A. Smith, and H. Fernandez. 2004.

The evolution of nocturnal behaviour in sweat bees, Megalopta genalis and M. ecuadoria (Hymenoptera: Halictidae): an escape from competitors and enemies? Biol. J. Linn. Soc. Lond. 83:377-387.

Wcislo, W. T., and B. N. Danforth. 1997. Secondarily solitary: the evolutionary loss of

social behavior. Trends Ecol Evol 12:468-474. Wcislo, W. T., and V. H. Gonzalez. 2006. Social and ecological contexts of trophallaxis

in facultatively social sweat bees, Megalopta genalis and M. ecuadoria (Hymenoptera, Halictidae). Insectes Soc. 53:220-225.

Wcislo, W. T., A. Wille, and E. Orozco. 1993. Nesting biology of tropical solitary and

social sweat bees, Lasioglossum (Dialictus) figueresi Wcislo and L. (D.) aeneiventre (Friese) (Hymenoptera: Halictidae). Insectes Sociaux 40:21–40.

Weeks, P. 2000. Red-billed oxpeckers: vampires or tickbirds? Behav. Ecol. 11:154-160. West, S. A., A. S. Griffin, and A. Gardner. 2007. Social semantics: altruism,

cooperation, mutualism, strong reciprocity and group selection. Journal of Evolutionary Biology 20:415-432.

Wilson, D. S. 1983. The effect of population structure on the evolution of mutualism: a

field test involving burying beetles and their phoretic mites. Am. Nat. 121:851-870.

Wilson, D. S. 2008a. Social semantics: toward a genuine pluralism in the study of social

behaviour. Journal of Evolutionary Biology 21:368-373. Wilson, D. S., and W. G. Knollenberg. 1987. Adaptive indirect effects: the fitness of

burying beetles with and without their phoretic mites. Evol. Ecol. 1:139-159. Wilson, E. O. 1971. The insect societies. Harvard University Press, Cambridge, MA. Wilson, E. O. 2005. Kin selection as the key to altruism: its rise and fall. Soc. Res.

72:159–168.

62

Wilson, E. O. 2008b. One giant leap: How insects achieved altruism and colonial life.

Bioscience 58:17-25. Wilson, E. O., and B. Hölldobler. 2005. Eusociality: origin and consequences. Proc Natl

Acad Sci U S A 102:13367–13371. Woodring, J. P. 1973. Four new anoetid mites associated with halictid bees (Acarina:

Anoetidae - Hymenoptera: Halictidae). J. Kans. Entomol. Soc. 46:310-327. Wright, S. J., and O. Calderon. 1995. Phylogenetic patterns among tropical flowering

phenologies. Journal of Ecology 83:937–948. Zhivotovsky, L. A. 1999. Estimating population structure in diploids with multilocus

dominant DNA markers. Molecular Ecology 8:907-913.

63

Vita

Natalia Beatriz Biani was born in Rosario, Argentina, on September 28th 1972. She is one

of the daughters of Ines Cavo and Ricardo Biani. She has three wonderful siblings.

During her childhood she lived in several countryside places of Brazil and Northern

Argentina, which together with her parents’ influence, inspired her love for nature. She is

married to Juan Carlos Santos and the fruit of their love gave origin to Ignacio, their

adorable son. She graduated from the National University of Rosario where she earned an

Agricultural Engineering degree in 1997. She worked as a teaching assistant in the

Botany department at that university and as a technical promoter in small family farms in

Rosario. Later she earned a Master in Conservation Biology in the International

University of Andalucia in Spain. While doing her master’s field work in Argentina, she

was marveled by the amazing interactions among ants and plants, and decided to pursue a

PhD. In the fall of 2002 she entered the Program of Ecology, Evolution and Behavior at

the University of Texas at Austin under the co-supervision of Dr. Ulrich Mueller and Dr.

Larry Gilbert. She was a teaching assistant for many courses such as Principles of Animal

Behavior, Biostatistics, Genetics and Genetics Lab, and Introductory Biology courses.

She also participated in the program of Supplemental Instruction which is a nationally

recognized, assistance program aimed at improving student performance and increasing

retention. She conducted her doctorate field work in Panama, but Ecuadorian and Costa

Rican tropical forests were stirring sources of wonder.

Permanent address: San Juan 2906, Rosario, Santa Fe, Argentina. This dissertation was typed by the author.